Mexico Public Finance Review - PFR - To Client División

MEXICO
Public Finance Review

Draft
December 2022
Contents
Executive summary.....................................................................................................................................5
Towards higher and more efficient revenues.........................................................................................7
Enhancing the fiscal performance of SNGs.............................................................................................9
Policy Options.......................................................................................................................................11
Chapter 1: Macro-Fiscal Overview............................................................................................................13
Macro-fiscal context.............................................................................................................................13
Introduction.......................................................................................................................................13
Slow growth undermined by structural weaknesses.........................................................................14
Evolution of public finances...............................................................................................................16
Expenditures......................................................................................................................................21
Fiscal balances and debt....................................................................................................................33
Additional fiscal space is needed to address spending pressures......................................................39
Appendix 1.1: Administrative Units of Government Spending in BOOST..............................................41
Appendix 1.2: Federal Spending by Budgetary Unit..............................................................................42
Appendix 1.3: Spending by Economic Classification based on Authorities’ Categorization...................43
Chapter 2: Revenues.................................................................................................................................44
Introduction...............................................................................................................................................44
Main characteristics of Mexico’s tax system.............................................................................................46
Direct taxes............................................................................................................................................49
Personal income tax (PIT)..................................................................................................................49
Simulations – Fiscal and Distributive impact......................................................................................52
Corporate Income Taxes (CIT)...............................................................................................................55
Property taxes.......................................................................................................................................57
Indirect taxes.............................................................................................................................................59
VAT........................................................................................................................................................60
Simulations – Fiscal and distributive impacts....................................................................................63
Excise taxes............................................................................................................................................66
Simulations........................................................................................................................................67
Environmental taxes..............................................................................................................................71
Align the vehicles tax with the emissions..........................................................................................74
Replacing current tax exemptions with Output-Based Rebates........................................................74
2
The new Simplified Regime for Small and Medium Enterprises........................................................75
Final remarks.............................................................................................................................................77
References.................................................................................................................................................82
Appendix 2.1: Methodological Appendix...............................................................................................85
Treatment of informality.......................................................................................................................85
Direct taxes............................................................................................................................................86
Simulation results (PIT)......................................................................................................................86
Indirect taxes.........................................................................................................................................90
Simulation results (VAT)....................................................................................................................90
Appendix 2.2: Tax schedules..................................................................................................................94
Personal Income Tax Rates....................................................................................................................94
Payroll taxes..........................................................................................................................................95
Value added taxes.................................................................................................................................97
Excise Taxes...........................................................................................................................................99
Appendix 2.3: RIF vs. RESICO...............................................................................................................100
Appendix 2.4: Distributional Impact of the Introduction of the Simplified Regime for Small and
Medium Enterprises............................................................................................................................103
Chapter 3: Intergovernmental Transfers................................................................................................107
Introduction........................................................................................................................................107
Mexico’s Intergovernmental Transfer System at a Glance................................................................109
Institutional framework..................................................................................................................109
The transfer system........................................................................................................................110
Main stylized facts of the current system of transfers.......................................................................114
Large vertical fiscal imbalances......................................................................................................114
Equalization Impact of Intergovernmental Transfers....................................................................114
Low levels of tax effort at the subnational level............................................................................117
Low levels of spending efficiency...................................................................................................119
Subnational borrowing.......................................................................................................................120
The Fiscal Responsibility Law for Subnational Governments.........................................................122
Main objectives of the reform of the transfer system and other considerations.........................125
Policy options......................................................................................................................................125
Transfer system...............................................................................................................................125
3
Main building blocks of the new system........................................................................................128
Local government revenue collection: property taxes...................................................................132
Subnational borrowing...................................................................................................................135
Final remarks...........................................................................................................................................138
4
Executive summary
1. Mexico’s economy is gradually bouncing back from one of its deepest recessions in decades.
Real GDP contracted by 8.0 percent in 2020, owning to pandemic-related mobility restrictions, supply
chain disruptions, and reduced demand in services sectors. Partly due to base effects, the economy grew
4.7 percent in 2021, still somewhat short from fully offsetting the sharp contraction seen in 2020. By
end-2022, economic activity and employment have surpassed pre-pandemic levels, with real GDP
reaching its Q42019 level. Labor market indicators continue to show signs of improvement, with jobs
and participation rates exceeding pre-pandemic levels, as well as a gradual improvement in job quality.
Figure ES1: Impact of the COVID-19 pandemic on real GDP vs. the Global Financial Crisis
105
100
95
90
85
80
t =0 t =1 t =2 t =3 t =4 t =5 t =6 t =7 t =8 t =9
GFC (t = 0 2008 q3) COVID (t = 0 2020 q1)
Source: INEGI
2. The COVID-19 pandemic has had significant human and social costs, with over 300,000 deaths
recorded in Mexico. The pandemic is expected to leave long-lasting effects on fiscal accounts and
potential growth. Long-term impacts of the pandemic in human capital accumulation, together with
persistent and rising inflation, and a deteriorating policy and institutional framework are expected to
continue to weigh down on the medium-term growth. Mexico entered the pandemic with a solid fiscal
position, underpinned by prudent policies. Government’s support in response to the pandemic was
relatively restrained compared to other countries, despite available fiscal space to support the economy.
As a result, the increase in public debt was also relatively small. The government aimed to balance fiscal
responsibility and curbing public debt with growing spending pressures amid a subdued economic
outlook and tighter conditions in financial markets.
3. Mexico’s fiscal stance remains solid, and debt is comparatively lower than its peers. However,
it is crucial to manage public spending pressures over the medium term to preserve fiscal sustainability.
There are various factors that can contribute to an increase in expenditures, including key infrastructure
projects going over budget, rising pension and social spending, and growing debt servicing costs.
Projected revenues would not suffice to fully cover the additional planned spending to implement
priority programs and infrastructure projects, which may require reallocating resources from other areas
or tolerating a higher deficit. Furthermore, the financial position of Pemex and CFE, and political
economy issues negatively impacting private investment represent important headwinds to government
finances and growth prospects. Pemex’s ongoing challenges imply that the state-owned company will
likely continue to rely on financial support from the federal government.
5
Figure ES2: Fiscal support and COVID Figure ES3: Change in general government gross debt
16
∆ in gross debt (t-1, t+1), t = 2020
40.1
14
12
23.3
20.7
16.5
10
12.3
12.2
11.9
11.2
10.8
9.4
8.8
8.0
7.1
7.0
5.6
5.1
4.4
4.3
8
0.2
% of GDP
-1.6
-2.8
-8.2
4
u
ile
G u ti
H o az il
Su ivia
na
ma
ica
er
Pa dor
y
i
e
y
Sa bia
Ja a
Ur ua
N i u ras
Ha
Gu xic
ala
Co ica
ua
Ch
Pa r
C o gu a
am
Bo .
n
P
ya
do
b
na
Br
ma
l
g
Ec p .
n ti
ua
ug
m
e
To
em
aR
ra
rin
lva
ra
M
nd
Re
lo
ge
ca
at
.&
st
Ar
n
0
ca
in
El
Tr
ni
mi
Do
ZAF
Y
Z
ALB
IDN
PER
SRB
AEs
A
A
IND
S
R
SAU
L
L
L
K
O
R
N
G
U
N
X
Es
EM s
KA
AR
EG
CO
PO
RU
PH
CH
PA
BR
TU
ME
TH
BG
GE
AR
TU
RO
C
CH
MK
GT
LID
Additional spending and forgone revenue Equity, loans, and guarantees
Source: Database of Country Fiscal Measures in Response to the COVID-19 Pandemic; WEO.
Note: Estimates as of September 27, 2021. Numbers in U.S. dollar and percent of GDP are based on October 2021 World
Economic Outlook unless otherwise stated. Country group averages are weighted by GDP in US dollars adjusted by
purchasing power parity.
4. A more growth-friendly fiscal policy becomes increasingly important to support inclusive

medium- and long-term growth in Mexico. Mexico’s growth performance was already weak prior to the
COVID-19 pandemic. Between 1990 and 2019, GDP per worker trailed at an annual rate of just 0.2
percent. Real GDP per capita relative to the US fell from close to 19 percent in 1990 to less than 15.3
percent in 2021. In the short term, new supply chain disruptions, Russia’s invasion of Ukraine, rising
inflation and tightening domestic and external conditions represent important headwinds to growth
moving forward. An aging population means that pressure on public services is expected to rise. There is
an urgent need to accelerate medium-term growth, for which additional resources to human capital
development and higher and better public investment are central.
5. To support higher growth and address pressing spending needs, Mexico needs to generate
additional fiscal space, especially from the revenue side, and strengthen its subnational fiscal
framework. Additional resources will be needed to avoid further increases in the level of indebtedness
(Chapter 2 discusses potential areas to increase revenue collection and potential distributional
implications of such measures). The expansion of key social programs (health care for people without
social security, financial assistance for people with permanent disabilities, universal pension for people
68+ and scholarships for students in public education levels) included in the recent constitutional
reform, rising expenditures from pensions and health care in the context of an aging population, and
improvements in education to foster the preparedness of the workforce will continue to exert pressure
on public finances.
6. In addition, significant and sustained efforts must be made to increase public infrastructure
investment to resume convergence towards OECD income levels. Mexico’s public investment is low
relative to peers, with a large geographical divide between the North-Center and South of Mexico in the
provision of such public infrastructure, and fiscal consolidation largely came at the expense of public
investment. Over the past decade, public investment and capital transfers have acted as a buffer, partly
due to a high degree of budgetary rigidities. Recent fiscal adjustments, which relied on reducing growth-
enhancing expenditures as the primary target for fiscal consolidation, have led to a significant decline in
public investment. This over-reliance on reducing capital spending could have negative impacts on
medium-term growth, especially considering that Mexico’s public investment is already low compared
6
to regional and structural peers. Higher growth rates over the medium term would require more and
better public investment. Moreover, key features of Mexico’s subnational fiscal framework undermine
efficiency of public spending of SNGs, depressing growth while cementing further wide-spread regional
socio-economic disparities.
7. Addressing these pressing spending needs without increasing the country’s debt burden
cannot be done without more revenues. Despite the 2013 and 2021 reforms and recent adjustments,
including strengthening tax administration and promoting formalization, tax collection remains below
the average of the LAC region and the lowest among OECD countries. Incentives for increase revenue
collection at the subnational level are limited. Revenue generation is not enough to fulfil critical needs to
build human capital, finance much needed infrastructure, and strengthen public service provision.
8. The focus of this Public Finance Review is on revenue mobilization and Mexico’s subnational
transfer system. The analysis presented in this report complements the World Bank’s Public
Expenditure Review (2016), which focused on subnational finances, the budget process, performance
and evaluation system, human resources management, health, education, social protection programs,
subsidies for productive inclusion, water and sanitation, and public security. The analysis presented in
this report relies on benchmarking exercises with regional peers and OECD countries, fiscal incidence
analysis using the 2018 ENIGH survey and the Commitment to Equity Approach (Lustig 2018), and a
microsimulation model to project the incidence of alternative policy reforms going forward.
Towards higher and more efficient revenues
9. Improvements in the design of Mexico’s tax system could help financing extra spending
without increasing economic distortions. Mexico’s tax structure offers opportunities for a reform that
can improve equity (enhancing progressivity in the system), simplicity (reducing taxpayer compliance
costs), and efficiency (including fostering— not harming— the competitiveness of the economy).
Mexico’s current tax system is not efficient, it raises relatively low revenues and creates distortions that
affect production and consumption decisions. The tax system also does not contribute enough to reduce
income inequality. Though Mexico’s personal income tax reduces gross income inequality more than any
other personal income tax system in Latin America, the overall tax system contributes little towards the
reduction of inequality given the low collection.
Figure ES4 Tax Revenues as a percentage of GDP
A. Evolution of Tax Revenues as % GDP B. Latin America and the Caribean (2019)
Source: SHCP and OECD Revenues Statistics.
7
Figure ES5. Mexico. The redistributive impact of taxes and transfers, 2020
(Cumulative interventions as a share of market income)
70%
60%
50%
Benefit as a share of market income plus pensions
40%
30%
20%
10%
0%
-10%
-20%
Poorest 2 3 4 5 6 7 8 9 Richest
Market income plus pension deciles
PIT State payroll tax Benito Juárez Educación Básica
Benito Juárez Educación Superior Pension Adulto Mayor Jovenes Construyendo Futuro
Jovenes Escribiendo Futuro Transferencia discapacidad Transferencia Madres Trabajadoras
Transferencia Precios Garantias Transferencias Ganadero Palabra Transferencias Fertilizantes
Tandas para el Bienestar Sembrando Vida Produccion para bienestar
Electricity subsidies VAT Excises
Netcash position
Source: World Bank Fiscal microsimulation model based on 2020 ENIGH.
10. While collection of income taxes – both PIT and CIT – is high compared to LAC, preferential
regimes across a large range of activities drag down revenues:
 A large informal sector weighs on Mexico’s PIT collection, which at around 5 percent of
GDP, is high compared to LAC but below the OECD average. There is limited scope for
increasing PIT revenue through rate increases. The income threshold for the top PIT rate
in Mexico is high, whereas the bottom statutory rate is low and could be raised.
However, the tax base could be expanded. Generous tax allowances and exemptions,
many of which are regressive, result in in PIT collection losses of an estimate 1.1 percent
of GDP (SHCP, 2020). Also, the complexity of the system could be reduced. For example,
the number of PIT tax brackets with 11 tax brackets is excessive.
 CIT collection is at 3.6 percent of GDP high compared to OECD and LAC countries. While
the statutory tax rate of 30 percent is high, effective tax rates are closer to 20 percent
due to accelerated depreciation on investments in buildings and machinery, below the
median of OECD countries. Also, revenue performance is affected by holes in the tax
system that provide ample opportunities to evade taxation. There is space to broaden
the tax base of the tax by eliminating some of the numerous deductions, exemptions,
special regimes, deferrals and administrative facilities.
8
11. VAT collection in Mexico is –at about 3.9 percent of GDP— low and inefficient. The standard
VAT rate is 16 percent, which is below the OECD average (19.2 percent). Furthermore, the base of the
VAT in Mexico is low due to tax expenditures on final products, such as exemptions of education and
housing (including rent and mortgage interest) which represent 0.24 percent of GDP as well as reduced
rate of 0 percent applied to a large number of goods and services. According to the SHCP, tax
expenditures from VAT represent around 1.43 percent of GDP. In addition, the VAT rate in border areas
is only 8 percent, which introduces significant economic distortions and revenues losses (SHCP, 2020).
The VAT-C efficiency - the ratio of actual VAT collections to the theoretical revenues under a perfectly
enforced tax levied at the standard rate on all final consumption without any exemptions – was at 29
percent in 2019, one of the lowest among OECD and LAC countries. Also, VAT compliance gap is
relatively high at 45.8 percent of potential revenue in 2016 (or 2.41 percent of GDP) as estimated by IMF
(2018). VAT tax expenditures are also inefficient in alleviating poverty as higher-income households
capture most of the benefits and due to high informality.
12. Sub-national governments in Mexico only collect less than 1 percent of GDP in taxes compared
to an OECD average of 5 percent of GDP. States and municipalities collect only a limited number of
taxes, including related to payrolls, motor vehicle registration, lodging taxes, and property taxes. Local
or municipal taxes only represent 0.3 percent of GDP in Mexico compared to 3.9 percent of GDP in the
OECD average. Collection from property taxes (“predial”) in Mexico is extremely low both in absolute
and relative terms. In 2019, Mexico collected 0.3 percent of GDP in revenues from property taxes, only a
fraction compared to OECD and other LAC countries. A reform of property taxes could be a simple and
efficient way to increase tax collection for local governments, but would require improvements in
cadastral systems, including updates to the valuation of existing properties in each jurisdiction and
keeping up with new construction, while building local capacity on tax administration.
13. Mexico recently introduced a new simplified regime for small and medium taxpayers
(Régimen Simplificado de Confianza, RESICO) which is expected to slightly increase formality 1 and
revenue collection over the medium term. The reform aims to reduce the time and number of
procedures for business and professionals required to pay taxes in Mexico with the objective of
increasing tax compliance and formalization. This new regime replaces the Fiscal Incorporation Regime
(Régimen de Incorporación Fiscal, RIF) which granted deductions to individual taxpayers with new
entrepreneurial activities. Most contributors expected to switch to the RESICO regime are likely to
belong to the top quintile of the pre-tax distribution and have higher profit margins. In the short term,
this could lead to a decline in collections and a negative distributional impact, which is expected to be
small because wage workers, who contribute 90 percent of all personal income taxes, do not have
access to the new regime. In the medium term, preliminary estimates suggest that 5.5 percent of
previously informal self-employed workers would become formal, increasing tax collections.
Enhancing the fiscal performance of SNGs
14. Mexico has one of the largest vertical fiscal imbalances (VFI) among OECD countries,
undermining incentives for SNGs to raise revenues. Revenue collection at the SNG level is very low, but
their spending responsibilities are much larger (see Figure ES1). In 2020, SNGs collected less than 1
percent in tax revenues, of which state and municipal taxes totaled 0.7 percent and 0.3 percent,
respectively. There are also large disparities in own-source revenue collection between SNGs in the
1
Loayza (2018) argues for a well-conceived formalization strategy aimed at reducing the costs and increasing the benefits of
formalization that would consist of four pillars: (i) increasing labor productivity; (ii) making labor markets more flexible; (iii)
making the regulatory framework more efficient; and (iv) reforming social protection. Simplification of the tax regime may
foster formalization, but it should be complemented with other reforms.
9
north and center-north and SNGs in the south. A large share of transfers in state revenue weakens
incentives for SNGs to boost own-source revenue collection.
Subnational revenue in % of total GG revenue
Figure ES6: Decentralization ratios in OECD countries, 2020

70
60
Tax decentralization > CAN
50
spending decentralization
CHE
AUS
USA
40
DEU
SWE
30 ICL FIN ESP
DNK
BEL
20 CZE MEX
FRA LVA NOR
POL
ITA
SLV AUT
10 PRT GBR
ISR
HUN NLD
CRI GRCIRLLUX SLK EST
LIT
0
0 10 20 30 40 50 60 70
Subnational spending in % of total GG spending
Source: OECD Fiscal Decentralization database

15. High dependence on federal transfers translates into lack of accountability and responsibility
for much of the funds spent and results in low levels of spending efficiency. The lack of spending
efficiency is also reinforced by the design of several participaciones and aportaciones transfers, which
introduce perverse incentives. For instance, among the Participaciones, the “Fondo de Participaciones
por el 100% de la Recaudación del ISR” is a matching grant from the federal government to the states to
hire and expand the number of their public employees (the more public employees are hired, the larger
the transfer). In the case of the Aportaciones, the current conditional grants for health, education and
other sectors are distributed with formulas based on capacity (supply side) measures (number of clinics,
hospital beds, number of teachers, etc.) as opposed to being distributed in terms of needs, such as the
size of the client base. Such distribution rules provide clear perverse incentives to overspending and
building excess capacity.
16. Horizontal gaps across SNGs are also large. Socioeconomic conditions vary significantly across
SNGs in Mexico, in particular the industrialized north and center-north and the less developed south. In
2020, the country’s wealthiest federative entity, Mexico City, had a GDP per capita over six times that of
the poorest state, Chiapas. Poverty and other socioeconomic indicators follow a similar pattern. Per
capita spending levels of state and local governments vary greatly, leading to disparities in access to
basic services. Also, wealthier states tend to have larger tax bases and more efficient tax
administrations. Intergovernmental transfers do very little to attenuate existing regional socioeconomic
disparities across SNGs.
17. Overall, the adoption LDFEFM and ancillary regulations have strengthened the framework for
controlling and monitoring subnational indebtedness. The fiscal rule that underpins the LDFEFM is
simple and transparent, and it provides clear operational guidance for fiscal policy. A “traffic light”
system defines the financing ceilings for the fiscal rule and provides information to creditors and
taxpayers on the fiscal solvency of SNGs. Likewise, the public debt record system is designed to offer
reliable, regularly updated information to market participants. Regulations requiring that debt be
contracted under the best market conditions help improve transparency and lower financing costs, and
the rules regarding federal guarantees for debt-restructuring operations are expected to enhance the
10
federal government’s leverage over the fiscal performance of highly indebted SNGs and reduce the
frequency of debt restructuring. Finally, the LDFEFM also contains important budget-preparation and
execution rules, such as the mandatory inclusion of medium-term fiscal frameworks in SNG budget laws,
which is expected to improve fiscal planning and enhance expenditure efficiency.
18. The high thresholds used by the “traffic light” system limit its ability to differentiate between
the level of indebtedness of SNGs. One of the goals of the LDFEFM was to control the growth of
borrowing and mitigate financial stress among subnational governments. High thresholds create large
intervals, and states with dissimilar levels of debt obligations and liquidity indicators are often grouped
together. The current thresholds are too high and the “traffic light” system may not convey the
appropriate signals. States are not classified as highly indebted are not constrained by the fiscal rule,
and thus, may continue to contract debt until hitting the threshold of 200 percent of non-earmarked
revenues (NER). This would undermine the core objective of the LDFEFM.
Policy Options
19. A well-designed reform of Mexico’s tax system and federal fiscal framework could lead to an
increase in revenues while reducing economic distortions and income inequality. The analysis
presented in this report does not provide a definitive proposal for tax system reform. Instead, it
presents a menu of potential options that could aid in the design of a reform that is both equity and
growth friendly. Some options to consider that could expand the tax base with limited impact on
poverty or inequality are given by:
 Direct taxes:
(i) PIT: removing deductions for medical expenses, mortgage interest and school tuition, which
would increase PIT revenue by 0.05pp of GDP. Most of these deductions largely benefit
individuals with higher incomes, thus increasing progressivity, reducing post-tax inequality,
and no impact in poverty.
(ii) PIT: reducing the exemption on pensions to less than 7 minimum wages (from 15 times). PIT
would become slightly more progressive, increasing revenue by 0.02pp of GDP, without an
impact on poverty or inequality.
(iii) CIT: eliminating tax expenditures from the incentives granted to the taxpayers of the
northern border region (0.3pp of GDP).
 Indirect taxes:
(i) VAT: Eliminating exemptions and reduced rates except for a basic food basket and
public transportation. Depending on the scenario, the additional VAT collection would
amount to an increase of between 1.4 pp or 1.6 pp of GDP. Both inequality and poverty
would increase relative to the baseline. Given these distributional impacts, eliminating
some exemptions and zero-rates for food and public transport would require additional
compensation for the poor to ensure poverty reduction and net fiscal savings.
(ii) VAT: If exemptions to educational services were eliminated, VAT revenue would
increase by 0.5pp. of GDP. There would be no significant impact on inequality. This
change would mostly affect better-off households who tend to be the ones using private
schools and other private services.
(iii) VAT: If exemptions to entertainment services and social services were eliminated, VAT
revenue would increase by 0.04pp. of GDP. There would be no significant impact on
11
inequality. This change would mostly affect better-off households who tend to use these
services.
(iv) VAT: Equalizing the standard rate at 16 percent (eliminating the reduced border rate)
across the country would improve both horizontal and vertical equity. This would
increase the progressivity of VAT, while having an insignificant impact on inequality and
a very small impact on poverty. This change would increase VAT revenue by 0.2pp. of
GDP.
(v) Tobacco taxes: Increase the specific component to MX$0.54, following the WHO
recommendation of a 75 percent tax share of the retail price of tobacco, resulting in
additional revenues of 0.01pp. of GDP. Short-term negative distributive impacts to be
offset by positive long-term distributive impact from health costs.
20. Mexico’s intergovernmental transfer system is complex, and its objectives are inconsistent,
resulting in a system that does not promote devolution or equalization. The system results in large
vertical fiscal imbalances and does little to address existing regional socioeconomic disparities, while
creating weak incentives for SNGs to support local economic growth, spend efficiently, or strengthen
own-source revenue collection. The chapter discusses some preliminary proposals to address these
issues, which would require a more in-depth examination. Some reform options to consider include:
(i) Property taxes: Mexico’s property tax collection is relatively low, presenting an
opportunity to increase revenue collection by improving coordination among different
levels of government and enhancing institutional and technical capacity. International
experience offers some practical solutions, including a more centralized state
administration of the property tax system, while preserving municipal autonomy for the
tax. State level collection on behalf of municipalities may help improve collections,
reduce collection costs, overcome limited capacity at the local level, and improve
efficiency, effectiveness, and sophistication in terms of fiscal cadasters, all due to
economies of scale and capacity gains.
(ii) Traffic light system: reducing the thresholds could enhance the signaling of the “traffic
light” system and strengthen the effectiveness of the fiscal rule in preventing excessive
indebtedness among SNGs. Halving the debt-to-NER thresholds and adjusting the debt-
service and short-term debt indicators accordingly could significantly enhance the
warning system and strengthen its impact on SNGs’ debt dynamics. Under the lower
proposed thresholds, fewer states would be classified as having sustainable debt levels,
enhancing the system’s ability to discriminate between different fiscal situations. The
proposed adjustment would enhance the impact of the LDFEFM’s fiscal rule on SNG
debt trajectories, as a larger number of SNGs would face lower financing ceilings. The
lower proposed debt thresholds would also reduce aggregate debt at the state level.
Figure ES7: Current (LHS) and proposed (RHS) thresholds for the “Traffic Light” system
12
Chapter 1: Macro-Fiscal Overview
Macro-fiscal context
Introduction
1. Mexico’s economy is gradually bouncing back from its deepest recessions in decades (Figure
1.1). Real GDP contracted by 8.0 percent in 2020, owning to pandemic-related mobility restrictions,
unprecedented disruptions in supply chains, and reduced demand in services sectors. The contraction in
output was among the largest observed in G20 countries. The COVID-19 pandemic resulted in high
human and social costs, with more than 300,000 excess deaths in 2020.
Figure 1.1: Impact of the COVID-19 pandemic on real GDP vs. the Global Financial Crisis
105
100
95
90
85
80
t =0 t =1 t =2 t =3 t =4 t =5 t =6 t =7 t =8 t =9
GFC (t = 0 2008 q3) COVID (t = 0 2020 q1)
Source: INEGI
2. Mexico’s fiscal support in response to the pandemic has been relatively muted compared to
both advanced and emerging economies. Additional non-financial public sector spending and forgone
revenue amounted to about 0.7 percent of 2020 GDP, of which 0.4 percentage points represented
additional spending and forgone revenue in the health sector. Additional support was provided by off-
budget measures (about 1.2 percent of GDP), which included loans to state workers with low interest
rates and personal loans, among other measures. Mexico’s response was one of the lowest among
emerging and advanced economies (Figure 1.2). The average of additional spending and forgone
revenue amounted to 11.7 percent and 5.7 percent of GDP for advanced economies and emerging
market economies, respectively. As such, the increase in Mexico’s general government gross debt
(Figure 1.3) between 2019 and 2021 was relatively small compared to other countries in the region
(+4.3pp of GDP).
Figure 1.2: Fiscal support and COVID Figure 1.3: Change in general government gross debt
13
16
∆ in gross debt (t-1, t+1), t = 2020
40.1
14
12
23.3
20.7
16.5
10
12.3
12.2
11.9
11.2
10.8
9.4
8.8
8.0
7.1
7.0
5.6
5.1
4.4
4.3
8
0.2
% of GDP
-1.6
-2.8
-8.2
4
u
ile
G u ti
H o az il
Su ivia
na
ma
ica
er
Pa dor
y
i
e
y
Sa bia
Ja a
Ur ua
N i u ras
Ha
Gu xic
ala
Co ica
ua
Ch
Pa r
C o gu a
am
Bo .
n
P
ya
do
b
na
Br
ma
l
g
Ec p .
n ti
ua
ug
m
e
To
em
aR
ra
rin
lva
ra
M
nd
Re
lo
ge
ca
at
.&
st
Ar
n
0
ca
in
El
Tr
ni
mi
Do
ZAF
Y
Z
ALB
IDN
PER
SRB
AEs
A
A
IND
S
R
SAU
L
L
L
K
O
R
N
G
U
N
X
Es
EM s
KA
AR
EG
CO
PO
RU
PH
CH
PA
BR
TU
ME
TH
BG
GE
AR
TU
RO
C
CH
MK
GT
LID
Additional spending and forgone revenue Equity, loans, and guarantees
Source: Database of Country Fiscal Measures in Response to the COVID-19 Pandemic; WEO.
Note: Estimates as of September 27, 2021. Numbers in U.S. dollar and percent of GDP are based on October 2021 World
Economic Outlook unless otherwise stated. Country group averages are weighted by GDP in US dollars adjusted by
purchasing power parity.
3. By late 2022, Mexico’s GDP has returned to its pre-pandemic level. The rebound in economic
activity gained pace in the second half of 2020, as mobility restrictions were eased domestically, US
demand accelerated, the reopening of the economy, and a steady vaccination pace. Economic activity
and employment have bounced back, surpassing pre-pandemic levels in 2022. Real GDP grew 4.7
percent in 2021 largely due to base effects, somewhat short from fully offsetting the sharp contraction
seen in 2020. Mexico’s GDP in Q42022 reached its Q42019 level. Labor market indicators continue to
show signs of improvement, with both jobs and participation rates already surpassing Q42019 levels,
together with a gradual improvement in job quality. Long-term impacts of the COVID-19 pandemic in
human capital accumulation, persistent and rising inflation, and a weak domestic demand associated
with structural factors, including a deteriorating policy and institutional framework, are expected to
continue to weigh down on the Mexican economy. Tightening financial conditions in the US and the
world represent additional headwinds to growth via a lower external demand and higher domestic
borrowing costs, among other channels.
Slow growth undermined by structural weaknesses2
4. The pandemic is expected to leave persistent effects on potential output growth. Yet Mexico’s
long-term economic performance has been weak relative to other emerging market economies. Over
the past few years, a stable macroeconomic environment was not sufficient to spur economic growth
over time. Real GDP growth declined from above 8 percent in the early 1980s to settle around 2 percent
in the early 1990s and remained around that level since then (Figure 1.4). Between 1990 and 2020, real
GDP growth averaged 2.2 percent per annum. The Tequila crisis of 1994, the Global Financial crisis, and
the most recent COVID-19 pandemic erased some of the gains in real output. Relative to the United
States, real GDP per capita fell from close to 19 percent in 1990 to less than 15.3 percent in 2021 (Figure
1.5). Mexico’s lackluster economic growth can be traced to declining productivity.
Figure 1.4: Mexico’s Declining GDP Growth Figure 1.5: GDP per Capita Falling Relative to
Trend the United States (real GDP per capita,
constant USD)
2
Based on World Bank (2022)
14
15 60
50
10
40
5
30
Percent
Percent
0 20
10
-5
19 0
19 1
19 2
19 3
19 4
19 5
19 6
19 7
19 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20
9
9
9
9
9
9
9
9
9
9
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
-10
19
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
10
12
14
16
18
20
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
Mexico Chile Poland Malaysia Korea
Real GDP 5-yr moving average
Source: World Development Indicators. Source: World Development Indicators
5. Poor aggregate productivity has been the main factor behind Mexico’s poor long-term
economic performance (World Bank 2022). Decomposing growth into contributions of factors of
production and total factor productivity suggests that factor accumulation has been the main drive of
growth in Mexico. Since 1990, contributions from labor (measured in efficiency units) were most
significant due to the demographic dividend, partly offset by migration, informality, and in particular low
female labor force participation. Only 45 percent of Mexican women of working age are part of the
workforce, well below countries in Latin America, the OECD, and peers with similar levels of economic
development as Mexico. Educational attainment and quality of education requires improvements, given
the increasing importance of more complex sectors in the economy (World Bank 2019). Insecurity and
crime are reported as key constraints by businesses in Mexico, contributing to misallocation of labor,
while also impeding investment.
6. The contribution of physical capital accumulation to growth has also been insufficient to
propel economic growth to the rates of peer countries. Investment has been broadly comparable to
Latin America and the Caribbean, averaging about 19 percent of GDP. Capital’s contribution to economic
growth has been slightly declining during the past decades, at around 1 percentage point of GDP, with
private investment somewhat compensating the decline in public investment since 1990. Yet,
investment has been lower than in fast-growing economies (i.e., Korea, Malaysia), where capital
represents a larger contribution to growth. Low investment, particularly in public infrastructure in areas
such as transport, electricity, water, and telecommunications, is hampering economic growth and
creating bottlenecks in key sectors of the economy and is not conducive to economic growth. By
improving public investment efficiency, Mexico could generate more and better infrastructure (IMF
2019). The efficiency gap between Mexico and the most efficient countries with similar levels of public
capital stock per capita is about 40 percent, larger than the averages for OECD countries (13 percent),
emerging market economies (27 percent) and Latin American countries (29 percent).
Figure 1.6 Labor Is the Main Driver of GDP Growth Figure 1.7 Mexico’s Low Investment
in Mexico Compared with Peers
(factor contribution to growth) (gross fixed capital formation, % of GDP)
15
5.0% 45
4.0% 40
35
3.0%
30 29
2.0%
24
25 22 22 21 21
1.0% 19 19
20 18
16 15
0.0% 15
-1.0% 10
-2.0% 5
1990-1999 2000-2009 2010-2017 Total
0
Mexico Korea Malaysia Thailand Brazil Chile Uruguay Peru Turkey Poland Argentina
TFP Capital Human Capital Labor Quantity GDP Total
1990-1999 2000-2009 2010-2015
Source: Penn World Table 9.1, World Bank (2022) Source: Data from the International Monetary Fund in
Note: GDP = gross domestic product; TFP = total factor productivity. purchasing power parity international dollars, World Bank (2022)
Figure 1.8 Efficiency frontier Figure 1.9 Public Investment Efficiency
Source: IMF (2019)

Note: Hybrid Indicator – Benchmark based on perceived quality and physical access to infrastructure
7. Positive contributions from the accumulation of factors of production have been offset by
negative productivity growth. TFP growth stagnated between 1990 and 2019. Similarly, GDP per worker
stagnated between 1990 and 2019, growing at an annual rate of 0.2 percent, well below the growth
seen in countries such as Poland or Korea, where GDP per worker increased at annual average rates of
3.2 and 3.7 percent, respectively. Large heterogeneity in terms of productivity can be observed between
formal and informal firms in terms of integration to global value chains, access to finance, and
managerial capabilities. There is also large divide between the North-Center and South of Mexico, which
has not narrowed over time.
Evolution of public finances
8. Fiscal policy has been a core pillar of macroeconomic stability in Mexico in the recent past.
Mexico entered the pandemic with a solid fiscal position, underpinned by prudent fiscal policies. The
government adopted a conservative fiscal stance, which helped contain public debt. Debt-to-GDP has
increased in the aftermath of COVID-19, but it is expected to stabilize around 60 percent in the medium
term. In the near term, risks to the outlook are tilted to the downside, with deteriorating prospects for
16
the economy, tightening conditions in financial markets, the financial position of Pemex and CFE, and
political economy issues that may negatively impact private investment and further impact growth
prospects. In the medium-term, additional resources will be required to address pressing spending
needs, including an aging population and pressures arising from pensions and health care, the expansion
of key social programs included in the recent constitutional reform, and improvements in human capital
to foster the preparedness of the workforce. Mexico’s public investment is low relative to peers, with a
large geographical divide between the North-Center and South of Mexico in the provision of such public
infrastructure. Further investments in public infrastructure, complemented with reforms aimed at
improving public investment management, are also needed to foster growth. Mexico’s budgetary
expenditures show a high degree of rigidity, undermining the quality of fiscal adjustment. In recent fiscal
consolidation episodes, public investment has taken the brunt of the adjustment. To address these
pressing spending needs without increasing the debt burden would require concerted efforts to raise
revenue collection. Mexico’s revenue (tax) mobilization is low compared to both regional peers and
OECD countries.
Revenues
9. Revenue collection in Mexico has improved over time, yet there is room for improvement.
International benchmarking suggests that Mexico’s revenue mobilization remains below the average for
OECD and LAC countries (Figure 1.10). In 2021, general government revenues represented a little over
23 percent of GDP. Successive reforms have improved tax collections and modernized the tax
administration. The 2013 Tax Reform3 helped raise additional non-oil tax revenue and thus reduce the
dependency from oil related revenues (Figure 1.11). The reform increased revenues from 9.6 percent of
GDP in 2013 to 13.0 percent by 2018, mainly through measures that raised income tax revenues, but
progress has stalled since then. Mexico’s tax-to-GDP ratio is relatively low when compared to other
emerging market and advanced economies, as well as other countries in the region.
Figure 1.10: General government revenues vs. Figure 1.11: Budgetary revenues (% of GDP)
LAC (% of GDP)
40
24.1
25
23.3
23.2
23.2
35
22.9
22.8
22.7
22.6
22.5
23
22.3
22.2
22.1
21.7
21.6
22
21.3
30
20.4
20.3
20.1
20
18.6
25
17.6
18
20
15
15
10
10
5
0
5
Do uat aiti
Me ru
alv e
Arg razil
Su l i v i a
Pa ana
n. c o
Gu a
or
J am ay
El S Chil
y
Co ame
Ecu na
s
Ni c m b i a
la
U r gu a
Pa ica
Pe
an a
Bo r
m
aic
Ho Tob.
ua
u ra
H
o
ni c al
Tri xi
ad
zue
u
sta .
ti
na
ad
B
y
Co Rep
ug
R
rag
mi em
ara
en
ri n
nd
ne
lo
&
Ve
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022e
Tax Revenue Non Tax Revenue Oil Revenue (Pemex) Oil Revenue (FG) IMSS+ISSSTE+CFE
2021 2020
Source: WEO April 2022 Source: SHCP
Figure 1.12: Tax revenues 2019 (% of GDP) Figure 1.13: Tax revenues in Mexico (% of GDP)
3
These reforms include the elimination of exemptions, measures to broaden the income tax base, the harmonization of the VAT
rate in border areas with the prevailing rate in the rest of the country, and the introduction of new excise taxes on high-calorie
foods and soft drinks (World Bank 2016)
17
40
16
34.3
35
14.3
13.6
13.5
30 14
13.1
12.8
12.7
13
13
24.9
24.6
24.6
24.2
23.4
25
22.6
22.4
22.2
20.7
12
19.2
18.2
18.2
20
18.1
17.8
17.5
10.3
15.5
15.1
14.5
9.8
9.8
13.7
13.6
13.3
9.6
13.1
15 10
9.4
9.3
9.3
10.5
8.8
10.0
8.7
8.7
8.7
8.5
8.4
8.3
10
8.1
8
5
6
0
Italy
Chile
Peru
Israel
Brazil
OECD
France
Canada
Finland
Mexico
Belgium
Uruguay
Portugal
Australia
Denmark
Euro area
Colombia
Argentina
Nicaragua
El Salvador
South Africa
United States
United Kingdom
LATAM & Caribbean

Africa Eastern and Southern 2
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022e
Source: World Bank. Excludes social security contributions. Source: SHCP
10. Mexico’s tax structure relies heavily on direct taxes relative to OECD and LAC countries.
Income tax revenue makes more than 40 percent of Mexico’s total tax revenue compared to 33 percent
in OECD. In fact, Mexico’s personal income tax reduces gross income inequality more than any other
personal income tax system in Latin America. Indirect taxes in Mexico are underperforming due to large
inefficiencies/tax expenditures in VAT collections. Due to broad exemptions and zero-rating in the VAT
regime, Mexico collects only 31.5 percent of the revenue that it could theoretically collect if VAT was
applied at the standard rate to all goods and services. Similarly, social contributions and property taxes
are well below regional peers and OECD countries. In addition, subnational tax collection as a share of
total public revenue is the lowest in Latin America and the OECD
Figure 1.14: Mexico’s tax structure relies heavily on income taxes (% of GDP) - Mexico vs
OECD countries
Source: OECD Revenue Statistics.
11. Oil revenues4 have lost importance as a source of revenue for the public sector, with the
decline in Pemex’s oil/gas/fuels production and, more recently, as the SOE loses market share in fuel
wholesale, which opened up to competition with the 2013 energy reform. In recent years, oil revenues
fell from about 39.4 percent of total public revenues in 2012 (or 8.6 percent of GDP) to a low-point of
just 11.3 percent in 2020 (2.6 percent of GDP) driven by the impact of the COVID-19 pandemic on oil
markets and the tensions within OPEC+ during the first half of the year. Oil revenues seemingly
4
Oil revenues as reported by SHCP include those received by Pemex from all its business units, net of imports and taxes paid.
18
recovered pre-pandemic levels, reaching 19.4 percent of total revenues (or 3.8 percent of GDP) in 2021.
However, Pemex’s revenues include federal government transfers for MXN317.2bn: MXN202.6bn for
debt amortizations and MXN114.6bn for the Dos Bocas Refinery. 5 Excluding those, this share drops to
14.1 percent. SHCP estimates that total oil revenues will reach MXN1.26 trillion in 2022, or 25 percent of
the total, mainly due to oil prices being 37.8 dpb higher than expected (92.9 vs 55.1 used in the 2022
Economic Package estimates).
Figure 1.15: Oil revenues as share of GDP Figure 1.16: Oil revenues as share of total
(percentage) revenues (percentage)
12 50
44.3
10.3
45
39.8
39.4
10
38.0
40
37.2
36.4
8.8
35.4
34.7
8.5
8.5
8.3
33.0
35
8
7.7
30.9
30.7
7.4
7.2
27.8
30
7
27.1
6.6
25.0
23.5
6 25
5.6
5.6
21.5
19.8
19.4
4.8
19.1
4.5
17.7
20
4.4
16.7
4.2
4.2
16.3
3.9
3.9
3.8
4
4
15
11.3
2.6
10
2
0 0
2022e
2022e
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: SHCP. Source: SHCP.
12. Starting in January 2022, the government introduced a new simplified tax regime (Régimen
Simplificado de Confianza, or RESICO), to foster formalization.6 The reform seeks to reduce the
procedures and time allocated to declare and pay taxes in Mexico to increase tax compliance and
facilitate formalization. The reform is also calibrated to ensure revenue-neutrality in the short-term –
i.e., maintaining similar effective rates for income tax for individuals that meet the requirements. The
reform also aims to lower the barriers to formalization, by simplifying the calculation, reducing the
amount of taxes, and fighting the use of invoices—both false and real—currently used to deduct taxes.
The reform simplifies the tax regimes for MSMEs and self-employed, and grants, accelerated
depreciation to support investment. The new simplified system for income tax extends to a broader set
of taxpayers (i.e., professionals) for individuals with a gross income of up to MXN 3.5 million per year
and legal entities with an income of up to MXN 35 million annual gross income. The new regime
simplifies compliance with tax obligations, preventing the taxpayer from keeping all the digital tax
receipts or invoices, since the tax collection will be calculated directly on the total income. The new
regime allows for the auto-population of the monthly tax returns based on e-invoices, significantly
simplifying tax filing for small and micro enterprises. While implementing measures to promote
formalization has become particularly urgent as the share of informal workers has increased post-
pandemic, it is rather premature to evaluate the impact of the reform in terms of revenue and
formalization resulting from decreasing compliance costs. In addition, a number of enhancements to the
legal tax framework7 approved and implemented between 2019 and 2022 have offset the impact on
revenues of a weaker economy, but sustained and strong growth remains elusive and tax revenues as a
share of GDP have not consistently surpassed the 14 percent ceiling.
13. In the absence of a new tax reform in the short-term, there are still some options that may
contribute to increasing revenue collection in the short-term. First, reducing policy and compliance
5
SHCP’s 4Q21 report
6
President Lopez Obrador promised no new taxes or hikes to the existing ones would be imposed, relying instead mainly on
administrative measures and ending tax breaks for the largest companies to increase revenues.
7
This included the elimination of universal netting of favorable tax balances, the reduction of tax expenditures related to fuel
consumption and the introduction of several measures against BEPS recommended by the OECD.
19
gaps could potentially add 2 percentage points in VAT collection over the medium term (IMF). The SAT
has much progress to do in terms of technology solutions and analytic tools to identify and prevent tax
fraud and evasion in income tax, VAT and excise taxes, and fully exploit the cfdi (digital invoicing). Their
VAT refunds’ risk model hasn’t changed significantly since it was developed in the 1990s, and general
audit selection relies largely on whistleblowers. Second, reinstating the federal vehicle property tax
could generate 0.2 percent of GDP in revenues, according to SHCP estimates. Another reference is that
average state collection in 2018 was 0.05 percent of GDP, whereas the maximum (State of Mexico) was
0.34 percent.8 Another potential source of additional revenues is property tax. Improving cadaster using
digital technologies could help increase average state collection, 0.17 percent of GDP in 2018, to Mexico
City’s figure of 0.57 percent.9
14. Tax exemptions continue to represent a significant amount of forgone revenue. Tax
exemptions on gasoline and diesel, which have become subsidies in 2022, are large and increasing. Last
year the gap versus the original estimate in fuel excise tax revenues amounted to MXN128.8bn, and this
year SHCP estimates it at MXN257.5bn or 0.9 percent of GDP in Precriterios 2023, an increase of 87
percent in real terms. Though in principle oil revenues surpluses should more than compensate for the
loss, Pemex’s difficult financial situation make it unlikely that they will pay the federal government the
royalties they should (see section on Pemex). SHCP estimates tax expenditures in 2022 at 2.7 percent of
GDP,10 even excluding those related to the VAT (1.5 percent of GDP). These include a number of
exemptions, deductions, special treatments and tax benefits (“estímulos fiscales”) on the personal
income tax, corporate income tax and excise taxes that on top of eroding the tax base and distorting
resource allocation are often regressive and/or have considerable social costs, the most salient example
being fossil-fuel exemptions or subsidies. Replacing the latter with direct cash transfers could increase
revenue, protect the income of disadvantaged groups and reduce inefficiencies. Limiting or eliminating
certain deductions (e. g. on gas-powered car acquisition for firms and tuition spending for individuals)
could boost collection and make the tax structure more progressive.
15. There are also some untapped opportunities in non-tax revenues. Although comparatively
smaller, non-oil non-tax revenues (royalties, fees for mining rights, payments in exchange for
infrastructure/transport/telecommunication concessions, user charges and payments for goods and
services) represent a potential source of additional resources, while improving their design could also
help achieve environmental and efficiency goals. Mining royalties (general: 7.5 percent of net revenues;
special for gold, silver and platinum: 0.5 percent of revenues from those minerals), introduced as early
as 2014, may need a revision. In terms of road concessions, SHCP has complete discretion over
payments (aprovechamientos) from new concessions and extensions of existing ones. The latter are
always granted by SCT to the original concessionaire, without undertaking a new tendering process.
Fees for the use of water and waste discharge may be reviewed and updated to consider social
externalities. Exemptions and special treatments for certain sectors can also be reviewed and limited.
16. Chapter 2 analyzes Mexico’s tax structure and compares it with other OECD and Latin
American countries to later present a menu of potential reform options that would help design an
equity-and-growth friendly reform. The current tax system is not efficient, it raises relatively low
revenues and creates distortions that affect production and consumption decisions. There is, therefore,
scope for a revenue enhancement reform supporting the government’s commitment to not raise tax
rates, by incorporating best practices to improve efficiency and simplicity of the system. The chapter
8
See CIEP estimates at https://ciep.mx/el-potencial-recaudatorio-en-las-entidades-federativas-impuesto-a-la-tenencia-
vehicular/#:~:text=La%20recaudaci%C3%B3n%20adicional%20de%20tenencia%20representar%C3%ADa%2058%20%25%20de
%20la%20reducci%C3%B3n,otras%20(SHCP%2C%202021).
9
See CIEP estimates at https://ciep.mx/el-potencial-recaudatorio-del-impuesto-predial-en-las-entidades-federativas/.
10
See SHCP’s Renuncias Recaudatorias 2021 at https://www.gob.mx/cms/uploads/attachment/file/649263/DRR_2021.pdf.
20
looks separately at Personal and Corporate Income Taxes (PIT and CIT), Value Added Taxes (VAT),
selected Excise taxes, and the property tax. The options considered here follow a clear principle: raise
revenues in a growth-and-equity-friendly way by broadening tax bases, strengthening the overall
progressivity of the fiscal system, enhancing tax policy and administration, and encouraging
formalization. In other words, the chapter identifies reforms that align inclusiveness and growth goals to
contribute to significantly reducing efficiency-equity trade-offs.
Expenditures
Structure of government spending
17. Based on Mexico’s federal system, spending is organized by different levels of government. 11
There is “public sector” spending, which consists of spending by the central government and by para-
statals. In addition, there is “subnational” spending, which consists of spending by states and by
municipalities. In Mexico, federal revenues are transferred from the central government to states and
municipalities, which are responsible for the provision of key services such as education, health, and
infrastructure (see Chapter 3).12 Historically, a considerable proportion of spending is executed at the
subnational level by states and municipalities.13, 14 Unless otherwise stated, the “federal” government is
defined as the central government and contributions to social security. Federal government excludes the
parastatals IMS, ISSSTE, PEMEX, CFE. The “public sector” is the aggregation of the federal government
and parastatals. A consolidated public sector would aggregate the public sector and sub-nationals. In the
period 2017-2020, 37 percent, 25.9 percent, 30.7 percent, and 6.5 percent of spending were executed
by the central government, parastatals, states, and municipalities, respectively. This amounted to just
over 32 percent of GDP in the period 2017-2020 (Figures 1.17-1.18). 15
Figure 1.17 Cumulative Spending by Level, 2017- Figure 1.18: Consolidated Spending by Level,
2020 2017-2020
(In percent of total spending, current pesos) (In percent of GDP)
40
Municipalities; 6.5 35
1.8
2.4 2.2 2.1
30
10.2
25 10.2 10.2 9.7
Central /1; 37.0
20
States ; 30.7 9.3
15 8.0 8.3 8.5
10
11.9 11.9 11.7 13.0
5
0
2017 2018 2019 2020
Parastatales /2; 25.9

Central /1 Parastatales /2 Estados Municipalidades
11
This chapter uses data provided by the World Bank BOOST team on Mexican expenditures. Where indicated, data is used
directly from the Ministry of Finance (SHCP), INEGI, IMF Government Financial Statistics, and IMF WEO.
12
For a list of entities captured by BOOST, please refer to Annex I: Administrative Units of Government Spending.
13
For more details on the institution structure of government spending see United Mexican States: Mexico Public Expenditure
Review. World Bank. March 30,2016. https://openknowledge.worldbank.org/handle/10986/25062
14
See Ministry of Finance Budgetary Transparency Portal for graphics and explanation of federal transfers to subnational
entities https://www.transparenciapresupuestaria.gob.mx/es/PTP/Cuenta_Publica_2020#slide07
15
Denominator effect pushed the size of government up in 2020 during the pandemic when nominal GDP contracted by 4.5
percent.
21
/1 Includes transfers to social security "Aportaciones a Seguridad Social". However, transfers to subnationals are excluded
from central spending to prevent double counting at the state and municpal levels. Specifically, excluded transfers are
transfers to federal entitites (states) and municipalities "Participaciones a Entidades Federativas y Municipios", participations
and contributions "Participaciones y aportaciones", and subsidies to federal entities (states) and municipalities "Subsidios a
entidades federativas y municipio".
/2 Parastatales are IMSS, ISSSTE, PEMEX, and CFE.
Sources: World Bank Mexico BOOST and Cross-Country Interface (CCI). Nominal GDP from IMF WEO, April 2022.
18. At 26.6 percent of GDP, total general government spending in the period 2017-2021 was
slightly below the Latin American average, and considerably lower than the OECD average (WEO April
2022). Moreover, at 22 percent of GDP, Mexico’s central government current spending is slightly below
Latin America and almost half the average size for central government spending in OECD economies. 16
Compared to Latin American, Mexico has the lowest spending in the categories of Compensation of
Employees (5.1 percent of GDP), Use of Goods and Services (2.2 percent of GDP), and spends below
average on the categories of Social Benefits (4.8 percent of GDP). Compared to OECD comparators,
Mexico’s spending gaps are particularly large in the category of Social Benefits (15.8 percent of GDP for
OECD and 4.8 percent of GDP for Mexico). More generally, in terms of relative compositions, compared
to both Latin American and OECD peers, central government spending in Mexico is light on
Compensation of Employees, Use of Goods and Services, and Social Benefits and heavy on Interest
Expenses and Other Expenses (Figures 1.19-1.21).
Figure 1.19: General Government Total Expenditure
(In percent of GDP)
Average, 2017-2021 Average, OECD (excluding LA)

60.0
50.0
44.0
40.0
26.6 28.5
30.0
20.0
10.0
0.0
GTM
SWE
AUT
PAN
DNK
DEU
MEX
ARG
AUS
SVN
FRA
TUR
LUX
POL
BOL
COL
CHL
PER
SLV
EST
SVK
NZL
ESP
BEL
IRL
Source: WEO April 2022
16
For Latin American and OECD countries for which the breakdown of central government current spending data is available.
22
Figure 1.20: Central Government Current Spending by Economic Classification (% of GDP) Figure 1.21: Central Government Current Spending by Economic Classification (% of current spend-
Compensation of employees Use of goods and services Interest expense Subsidies expense Grants expense ing) Compensation of employees Use of goods and services Interest expense Subsidies expense Grants expense
Social benefits expense Other expense Social benefits expense Other expense
Average OECD /1 9.9 6.2 1.9 15.8 2.6 Average OECD /1 25.4 15.9 4.9 4.8 40.5 6.6
Korea 6.0 6.1 1.1 7.4 2.5 Korea 21.4 21.7 4.0 17.9 26.2 8.8
7.4 3.6 0.4 11.6 3.1 24.8 12.0 1.2 10.6 39.0 10.4
Turkey 8.2 4.8 2.6 12.0 1.3 Turkey 26.7 15.7 8.5 5.8 38.8 4.1
9.7 6.7 1.3 10.8 3.4 28.7 19.7 4.0 5.2 31.8 10.0
Estonia 10.7 6.3 0.1 14.6 1.5
8.9 5.3 4.0 15.4 0.3 Estonia 30.6 18.0 0.2.30 41.8 4.3
Japan 5.2 3.4 1.7 21.2 3.5 25.3 14.9 11.5 2.8 43.7 1.0
9.2 5.7 0.8 15.8 1.8 Japan 14.6 9.5 4.91.6 59.3 9.8
Israel 10.0 8.4 2.2 12.7 2.7 25.1 15.5 2.1 6.4 43.1 4.9
12.2 8.8 3.0 11.2 2.1 Israel 26.5 22.3 5.8 4.0 33.8 7.3
Netherlands 8.1 5.6 1.1 21.3 0.8 31.2 22.4 7.5 4.5 28.5 5.3
8.9 8.1 2.3 13.6 4.7 Netherlands 20.2 14.1 2.7 4.7 53.5 2.1
Spain 10.7 5.1 2.4 19.2 1.6 22.1 20.0 5.7 4.6 33.6 11.5
10.4 7.9 2.6 13.1 4.7 Spain 25.9 12.4 5.9 2.9 46.6 3.9
Iceland 14.4 10.9 5.1 7.3 3.2 25.0 19.0 6.2 4.0 31.4 11.3
10.7 5.1 3.5 18.8 2.7 Iceland 34.0 25.7 12.0 3.1 17.2 7.5
Germany 7.6 5.2 0.9 24.7 2.2 25.3 12.0 8.3 1.7 44.1 6.4
12.7 6.9 0.5 16.4 4.6 Germany 17.7 12.1 2.22.6 57.5 5.2
Norway 14.8 7.1 0.7 17.5 2.6 28.6 15.6 1.04.2 36.9 10.3
10.7 6.2 1.8 22.6 3.0 Norway 32.4 15.6 1.54.3 38.3 5.7
Belgium 12.1 3.7 2.4 25.4 1.8 22.5 13.0 3.7 4.5 47.6 6.4
Belgium 23.9 7.3 4.8 7.8 50.2 3.6
Average LA /1 8.6 3.9 2.7 6.1 3.3
Average LA /1 34.2 15.3 10.8 1.6 24.3 13.1
Guatemala 7.0 2.7 1.3 2.00.8 4.3
8.7 2.30.7 0.9 Guatemala 50.2 19.2 9.5 1.0 13.9 5.9
Peru 6.8 6.1 1.22.1 2.2 51.5 13.5 4.40.0 25.3 5.3
5.1 2.2 2.9 4.8 6.0 Peru 36.6 33.1 6.7 0.0 11.6 12.1
Chile 7.0 2.9 0.9 6.1 5.5 22.9 10.1 13.1 3.1 21.7 27.3
11.2 4.3 3.0 2.5 3.5 Chile 29.1 12.2 3.5 6.8 25.4 22.8
Costa Rica 12.6 3.3 3.3 6.0 3.0 43.3 16.6 11.6 1.5 9.7 13.6
6.0 5.7 3.0 8.5 6.2 Costa Rica 44.8 11.6 11.8 0.0 21.1 10.5
Brazil 13.1 5.1 8.1 18.7 1.6 20.0 19.1 10.0 1.7 28.5 20.7
0 5 10 15 20 25 30 35 40 45 50 Brazil 27.8 10.9 17.3 0.7 39.8 3.4
Source: IMF GFS. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Source: IMF GFS.
/1 Average for countries displayed for which there is data.
Composition and drivers
19. Following the global financial crisis, government spending increased 3.8 percent per year in
real terms between 2009 and 2016 until the federal government reduced programmable spending 17 in
2017 to curb a growing stock of debt. Since 2017, programmable spending remained basically constant
through 2020. In 2021, it spiked 7.5 percent in real terms, mostly explained by transfers to Pemex for
debt amortizations and the Dos Bocas Refinery in the amount of MXN272.2bn that were not originally
included in the 2021 Budget. Estimates in SHCP’s Precriterios 2023 envisages programmable expenditure
to remain broadly constant in real terms in 2022 (or approximately at 94.5 percent of its 2016 level),
provided that Pemex and/or other priorities do not require any additional resources. Pemex already
received MXN45.1bn from the federal government in Q12022 that were not included in the 2022
17
Expenditures that can be determined and changed in a more or less discretionary way, such as investment, social programs,
current expenses. Non-programmable spending includes financial cost, which can be moved by changes in interest and
exchange rates; participaciones to states, which move with observed tax and oil revenues; and arrears, which can vary
significantly between the moment the Federal Budget is presented (September of a given year) and the moment payments has
to be made (1st quarter of the following year).
23
Budget. Non-programmable expenditure, on the other hand, has broadly shown an upward trajectory
over time. It is projected to increase 12.5 percent in 2022, driven by debt servicing costs. The latter went
from 1.8 percent of GDP in 2008 to 2.6 percent in 2018, with the increase in total debt. Debt service
payments remained broadly constant during 2019-2021, however, they are projected to reach 3 percent
of GDP in 2022 and 3.2 percent of GDP in 2023 due to tightening global financial conditions (SHCP’s
Precriterios 2023).
Figure 1.22: Expenditures (trillion 2022 pesos) Figure 1.23: Variation in expenditures (annual
real growth rate)
Programmable Non programmable Total Programmable Non programmable
8 15
7
10
1.65
1.93
1.56
1.72
1.43
6
1.89
1.77
1.85
1.35
1.72
1.31
5
1.28
1.26
5
1.20
1.27
1.19
1.28
4 0
3
5.77
5.60
-5
5.47
5.45
5.39
5.21
5.09
5.03
5.03
5.01
4.95
4.75
4.60
4.48
4.22
2
3.84
3.56
-10
1
0 -15
2022e
2022e
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
20. The composition of government expenditures has gradually shifted towards current expenses
in detriment of public investment. Physical investment18 as a share of the public sector’s primary
spending went from 20.2 percent in 2010 to 11.4 percent in 2021. As a share of GDP, physical
investment declined 4.7 percent in 2014 to 2.6 percent in 2021. Most of the reduction took place
between 2014 and 2017 (40.1 percent in real terms), first due to Pemex’s increasing financial problems
forcing it to curb spending, and later due to the fiscal consolidation that was implemented to reduce the
debt level of indebtedness.19 These measures included a containment of the public-sector wage bill and
a consolidation in federal programs for subsidies and transfers. In addition, the consolidation also relied
heavily on cuts to public investment (-1.5 percent of GDP between 2015 and 2018), with potential
negative impacts on long-term growth. Public physical investment averaged just 2.6 percent of GDP over
2017-2021, of which two-thirds were Pemex’s investments (or 1.7 percent of GDP).
Figure 1.24: Physical investment as share of Figure 1.25: Physical investment as share of GDP
primary spending (percentage) (percentage)
18
Spending directly related to fixed capital formation.
19
Debt-to-GDP reached 50.5 percent by end-2016 (SHCP's 4Q2016 report on public finances, page 71, defined as SHRFSP,
Hacienda's broadest measure of public net debt). Excluding Banxico's surplus of MXN239.1bn (or 1.4 percent of GDP) that the
federal government received in April, net debt was 51.9 percent of GDP. The increase in debt during the 2013-2018
administration was the result of a policy decision taken in 2013 to run temporarily higher deficits than in previous years (see
page 171 on Criterios 2014), with actual deficits being higher than planned. This was mainly the result of economic growth and
oil prices being lower than anticipated and exchange rate shocks that affected outstanding debt.
https://www.finanzaspublicas.hacienda.gob.mx/work/models/Finanzas_Publicas/docs/congreso/infotrim/2016/ivt/01inf/
itinfp_201604.pdf
https://www.finanzaspublicas.hacienda.gob.mx/work/models/Finanzas_Publicas/docs/paquete_economico/cgpe/
cgpe_2014.pdf
24
25 5.0
4.7
4.7
4.5
4.5
4.4
4.5
4.3
4.2
20 4.0
3.6
3.5
3.0
3.0
15 3.0
2.8
2.8
2.6
2.6
2.6
2.4
2.5
2.3
10 2.0
1.5
5 1.0
0.5
0 -
2022e
2022e
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
21. The approved 2022 Budget envisaged an increase in physical investment. Its share of primary
expenditures is expected to reach 14.0 percent and 3.0 percent of GDP. Close to 30 percent of the
increase is attributed to the intended acceleration of the Tren Maya (Train Maya) and the Interoceanic
Corridor projects, with the expectation that these mega projects may propel growth higher, especially in
the less developed southern states. About 54 percent of the increase is associated to Pemex’s growing
costs in E&P due to its aging oil fields. However, budget execution has been a recurring issue: as of
Q12022, physical investment was MXN75.7bn (or 33.1 percent) below program. In 2021, physical
investment ended up MXN29.8bn below program, 1.2 percent lower in real terms than in 2020 and 3.4
percent lower in real terms than in 2018, at a time when government spending was most needed to
boost the economic recovery.
22. The main drivers of growth in federal spending (economic classification) in recent years have
been the categories of financial investment, transfers to sub-nationals, and transfers other than to
sub-nationals.20 Financial investment is almost exclusively composed of spending on the purchases of
titles and securities and purchases of financial investments and other provisions, and these have seen a
sharp increase in 2021. These are likely to be large capital injections into state owned enterprises. Direct
transfers to sub-nationals (states and municipalities) 21 have steadily grown in recent years in line with
participaciones and aportaciones (except for 2020, the first year of the pandemic, when participaciones
were slashed). Finally, growth of transfers other than to sub-nationals has principally been driven by
transfers to executive branch trusts, pensions and retirement payments, and grants and subsidies.
During the first year of the pandemic, these categories caused non-subnational transfers to grow by 11.9
percent y-o-y. In addition, pensions and retirement payments have consistently grown during the period
2017-2021.
Figure 1.26: Composition of Federal Spending, Figure 1.27: Contribution to Growth in Federal
by Economic Classification Spending, by Economic Classification
BOOST Presentation BOOST Presentation
20
World Bank Mexico BOOST
21
It is composed of “aportaciones”, “participaciones”, and “convenios” (see Chapter 3). Aportaciones are resources that are
transferred by the federal government to states and municipalities through special funds to achieve specific objectives in
health, education, social development, security, amongst others. Participaciones are federal resources transferred to states and
municipalities, and these resources can be used to meet whatever objectives that are established by subnational governments
directly. Convenios are a small proportion of federal resources that are managed either directly by the states and/or in
coordination with the Federal Public Administration. See SCHP
https://www.transparenciapresupuestaria.gob.mx/es/PTP/Cuenta_Publica_2020#slide07
25
(In billions of pesos) (In percent, y/y)
6000 20
Other
Other 15
5000
Financial investment
Financial investment 9.6
10
7.0 3.2 3.8
4000 Capital investment
Capital investment
5
Transfers to subnationals
3000 Transfers to subnationals

0 Transfers other than to sub-nationals
Transfers other than to sub-nationals

2000 Interest
-5
Interest
Use of goods and services
1000 -10
Compensation of employees
Compensation of employees -15

0 Total
2017 2018 2019 2020 2021
2017 2018 2019 2020 2021
Transfers to Sub-nationals by Federal Government Contribution to Growth Federal Transfers to Sub-nationals

2000000000000
1800000000000 8.0
8500 Convenios 7.4
1600000000000 8500 Convenios
6.0 5.9
1400000000000
5.0
1200000000000 4.0 8300 Aportaciones
1000000000000 8300 Aportaciones
2.0
800000000000
8100 Participaciones
600000000000 0.0
-0.5
400000000000 8100 Participaciones -2.0
Total transfers
200000000000
-4.0
0
2017 2018 2019 2020 2021
2017 2018 2019 2020 2021
Transfers other than to Sub-nationals Contribution to Growth, Transfers other than to Sub-nationals
4900 Transfers abroad

2000000000000 4900 Transfers abroad
1800000000000 15.0 4800 donations

4800 donations
1600000000000 11.9
4700 Transfers to social security
1400000000000 4700 Transfers to social security 10.0
4600 Transfers to trusts, mandates and other similar

1200000000000
4600 Transfers to trusts, mandates and other similar 5.8
5.0
1000000000000 4500 Pensions and retirement
3.2
4500 Pensions and retirement
800000000000
0.0 4400 Social assistance
600000000000 -1.9
4400 Social assistance
4300 Grants and subsidies
400000000000
-5.0
4300 Grants and subsidies
200000000000 4100 Internal transfers and allocations to the public sector
0 4100 Internal transfers and allocations to the public sector -10.0 Total transfers other than to subnationals
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
26
Subsidies and Grants Contribution to Growth, Subsidies and Grants
439 Other Subsidies

800000000000 25.0
439 Other Subsidies
20.0
700000000000
438 Subsidies to Federal Entities and Municipali- 15.0
ties 438 Subsidies to Federal Entities and Municipalities
600000000000 10.0
7.9
5.0 4.7
500000000000 437 Consumer subsidies
437 Consumer subsidies 2.0
0.0 0.0
400000000000 -5.0 434 Subsidies for the provision of public services
434 Subsidies for the provision of public services -10.0
300000000000
-15.0 433 Investment subsidies
200000000000 -20.0
433 Investment subsidies
-25.0 431 Production subsidies
100000000000 2017 2018 2019 2020 2021
431 Production subsidies
0 Total Grants and subsidies
2017 2018 2019 2020 2021
Source: World Bank Mexico BOOST and WB staff calculations.

23. Spending on Social Protection and Education has consistently been a positive driver of federal
spending. In response to the pandemic, spending on Health increased in 2020 and 2021. Also, four
priority social programs were included with the reforms to the Article 4 of the Constitution in May 2020:
health care for people without social security, financial assistance for people with permanent
disabilities, universal pension for people 68+ and scholarships for students in public education levels (all
levels). The Constitution also states that the budget for these programs must at least remain constant in
real terms over time.22 For instance, the universal pension for the elderly program more than doubled
from MXN100bn in 2019 to MXN238bn in 2022. SHCP’s Precriterios 2023 anticipates a MXN303.7bn
budget for the program next year (i.e., an additional real increase of close to 23 percent), bring the
budget of the program to 1 percent of GDP. The figure is expected to continue growing, since the
administration’s goal is to give each beneficiary MXN6,000 every other month by 2024, a 56 percent
nominal increase from the current MXN3,850. With the old-age dependency ratio expected to increase
from 11.5 in 2020 to 25.7 by 2050,23 even maintaining transfers per beneficiary constant after 2025, the
program will absorb increasing public resources. In addition, the regular federal public sector’s pension
obligations (Federal Government, IMSS, ISSSTE, PEMEX, CFE and others) have been growing at 4.5
percent y-o-y on average between 2016 and 2021, above the economy or tax revenues growth rates and
is expected to reach close to MXN1.2 trillion (4.1 percent of GDP) in 2022.
24. In 2021, spending in the category of Economic Affairs was the largest contributor to growth in
federal spending, with fuels and energy being a key driver of growth in recent years. In 2021, it was
the single largest driver of growth in this sector, contributing more than 60 percent of the final 53.3
percentage growth in sector spending. Spending on fuels and energy has exclusively been in
consumption subsidies and financial investment or capital injections in energy companies. In the last
three years, most of this spending has been in financial investments in the form of purchases of titles
and securities/ capital injections, with a considerable spike of MX270bn pesos in 2021 in this category
(Figures 1.28-1.29).
Figure 1.28: Composition of Federal Spending by Figure 1.29: Contribution to Growth in Federal
COFOG Classification Spending
22
Estimates provided in SHCP’s Precriterios 2023 (Annex IV) for each program for 2023 (2022) are as follows: health care –
MX80.6bn (MX77.6bn), disabilities – MX30.1bn (MX20bn), universal pension – MX303.7bn (MX238bn), and scholarships –
MX34.5bn (MX33.2bn).
23
See http://www.conapo.gob.mx/work/models/CONAPO/Cuadernillos/33_Republica_Mexicana/33_RMEX.pdf.
27
6000000000000 12
Other not classified /1
Other not classified /1 9.6
10 7.0
Social protection
5000000000000
Social protection
8 Education
Education
3.2 3.8 Recreation, culture, religion
4000000000000
Recreation, culture, religion 6
Health
Health
3000000000000 4 Housing and community ammenities
Housing and community ammenities

Environmental protection
2
2000000000000 Environmental protection
Economic affairs
Economic affairs
0
Public order and safety
1000000000000 Public order and safety -2 Defense
Defense General public services

-4
0
General public services 2017 2018 2019 2020 2021 Total
2017 2018 2019 2020 2021
Spending in Economic Affairs Contribution to Growth, Economic Affairs

(In billions of pesos) 9 Other Industries and Other Economic Matters
(In percent, y/y) 9 Other Industries and Other Economic Matters
8 Science, Technology and Innovation

700000000000 8 Science, Technology and Innovation
80.0 7 Tourism
600000000000 7 Tourism
53.3
60.0 6 Communications
500000000000 6 Communications -11.9
40.0 5 Transportation
-5.3
400000000000 5 Transportation 18.5
20.0 4 Mining, Manufacturing and Construction
300000000000 4 Mining, Manufacturing and Construction

0.0 3 Fuels and Energy
200000000000 3 Fuels and Energy -20.0 2 Agriculture, Forestry, Fishing and Hunting
100000000000
2 Agriculture, Forestry, Fishing and Hunting -40.0 1 Economic, Commercial and Labor Affairs in General
0 -60.0
1 Economic, Commercial and Labor Affairs in General Total economic affairs
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Federal Spending on Fuels and Energy

(In billions of pesos)
450000000000
400000000000
350000000000
Other
300000000000
250000000000 Purchase of titles and securities
200000000000
150000000000 Subsidies
100000000000
50000000000
0
2017 2018 2019 2020 2021
/1 Includes transactions related to public debt, transfers between different levels of the government, financial system
recovery costs, and arrears from previous fiscal years.
Source: World Bank BOOST. Government presentation reclassified to IMF GFS COFOG classifications.
Figure 1.30: Pension spending growth (real Figure 1.31: Pension spending as share of GDP
variation) (percentage)
28
16 4.5
4.1
4.1
4.0
14 4.0
3.6
3.5
3.4
12
3.2
3.2
3.2
3.0
3.0
2.9
10
2.7
2.6
2.6
2.5
2.4
8
2.1
2.0
2.0
1.9
6
1.5
4
1.0
2 0.5
0 0.0
2022e
2022e
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
25. The federal government tends to over-execute its budget. Excluding 2019 and 2020, the
federal budget has been over-executed by on average 7-8 percent. Over-execution of the budget has
been driven by non-subnational transfers and financial investments. Within the category of non-
subnational transfers, transfers to funds and trusts that are controlled by the executive branch and the
category of grants and subsidies are the principal areas where the government was spending more than
was allocated. In the last 5 years, compensation of employees has been consistently under-executed by
an average of 13.3 percent every year over the period 2017-2021. From the functional perspective, the
Economic Affairs sector has been the main driver of growth in the period 2017-2020. Within Economic
Affairs, the category of Fuels and Energy has been an area in which the government has over-executed
the allocated budget, and principally on subsidies and the purchase of shares/capital injections.
Figure 1.32: Allocated Budget and Executed Figure 1.33: Budget Over-Execution
(In billions of pesos) (Percent deviation of executed over allocated)
8.0
Budget amount Executed amount
7.1
5,500 6.6
5,000
4,500
4,000
3,500
1.0
0.8
3,000
2017 2018 2019 2020 2021
2017 2018 2019 2020 2021
Budget over-execution
Figure 1.34: Economic Classification: Figure 1.35: Contributions to Over-Execution in

Contributions to Over-Execution of Total Budget “Transfers other than to Sub-Nationals” Budget
(In percent, deviation of executed from (In percent, deviation of executed from
budgeted) budgeted)
29
12 24
8.0 6.6 21.6
7.1 22
Other 4900 Transfers abroad
10
20
8 Financial investment 18 4800 donations
16.7
16
6 Capital investment 4700 Transfers to social security
1.0 14
0.8
4 Transfers to subnationals 12 8.0 4600 Transfers to trusts, mandates and other similar
9.5
10
2 Transfers other than to sub-nationals 4.0
4500 Pensions and retirement
8
0 Interest 6
4400 Social assistance
4
-2 Use of goods and services
2 4300 Grants and subsidies
-4 0
4100 Internal transfers and allocations to the public sector
-2
-6
Total -4
2017 2018 2019 2020 2021 Total transfers other than to subnationals
1 2 3 4 5
Figure 1.36: Functional Classification Figure 1.37: Functional Classification

Contributions to Over-Execution of Total Budget Contributions to Over-Execution in “Economic
(In percent, deviation of executed from Affairs” Budget
budgeted) (In percent, deviation of executed from
budgeted)
12 80
Other not classified /1 70.8 9 Other Industries and Other Economic Matters
10 70
Social protection
8.0 6.6 8 Science, Technology and Innovation
60
8 7.1 Education
7 Tourism
Recreation, culture, religion
50
6 19.5
37.9 21.3 6 Communications
Health 40
4 1.0 0.8
30 5 Transportation
Housing and community ammenities 25.2
2
Environmental protection 20 4 Mining, Manufacturing and Construction
0 Economic affairs
10 3 Fuels and Energy
-2 Public order and safety

0 2 Agriculture, Forestry, Fishing and Hunting
Defense
-4
-10 1 Economic, Commercial and Labor Affairs in General
General public services
-6 -20
Economic affairs total
2017 2018 2019 2020 2021 Total 2017 2018 2019 2020 2021
Source: WB BOOST.
/1 Includes transactions related to public debt, transfers between different levels of the government, financial system
recovery costs, and arrears from previous fiscal years.
Cyclicality of Spending in Mexico

26. Mexican government spending has been slightly procyclical. With a coefficient close to zero,
general government spending in Mexico has been relatively neutral to the business cycle. Adopting a
more countercyclical fiscal policy may help counteract the business cycle and reduce overall
macroeconomic volatility. In general, advanced economies tend to have counter-cyclical fiscal policies –
most OECD countries tend to also have counter-cyclical fiscal policies (Figure 1.38).
30
Figure 1.38: Cross-Country Spending Cyclicality and Income per Capita, 2000-
1.0
2021
SLBTLS
LBRAll countries AZE
Polynomial
LBN LVA(All countries)
0.8 NIC
COG MDA UKR IRN URY TTO QAT
CAF OECD FSM FJI ROU
Linear (OECD) GRC
HUN CYP
COD NPL ECU
BIH
0.6 Latin America NAM
CIVSDN AGO GEO SRB ATG ISL
spending
GHA
Procycli -
KIR CMR PAK BOL JOR BRA BLRSUR HRV EST
LTU
MDG COL PAN GNQ
TKM LBY RUS
cal
0.4 LSO VUT KEN ALBVCT KNA KWT
RWA
ETH TZA
GMB TON
LAO
BFAMLI COM DJI GAB
Correlation Coefficient of Cyclical Spending and
JAM BGR POL OMNSAU

NER TCD BEN BTNLKA
CHN SWE BRN
0.2 BGD
MMR CPV
WSM
IDN
GRDPLW SYC
BDI MOZ SLE UGA VNM BLZ MNGGUY
SWZ MDV BRB SVK PRT CZEBHS
PER LCA THA
SEN
HTI KGZ MEX ESP NLDIRL
0.0
Cyclical Output
TGO ZMB IND HND

UZBPHL DMA MKD MLT MAC
SLV EGY ZAF TUR BHR
PNG SVN
-0.2 NGA DZABWACRI
TUN
MUS ITA
NZL
STP PRY FIN
UVK DOM MYS ISR
Counter -
spending
GNB MARGTM NOR
cyclical
-0.4 KHM
GINTJK CHL KOR
ABW
DEUAUT UAE LUX
GBR
-0.6 AUS BELUSASwitzerland
SGP
FRA DNK HKG
-0.8 JPNCAN
-1.0
2.50 3.00 3.50 4.00 4.50 5.00 5.50
Log (GDP per capita, PPP (constant 2017 international $), Average 2000-2021)
Sources: IMF WEO database (October 2021 Vintage) World Bank WDI and staff calculations. Methodology based on Abdih, Y.,
Lopez-Murphy, P., Roitman, A., and R. Sahay (2010), "The Cyclicality of Fiscal Policy in the Middle East and Central Asia: Is the
Current Crisis Different", IMF Working Paper. Spending deflated by country's respective CPI. Correlation coefficient is between
cyclical components of real spending and real output for the years 2000-2021. Cyclical component of real spending and real
output generated using HP filter lambda=100. Forecast data from IMF for 2021-2026. Future years estimated at constant
growth rates of end of medium-term horizon.
27. While on average, over the period 1995-2021, total spending has been moderately procyclical
(cyclicality coefficient of 0.3, Figure 1.39), it is slightly countercyclical during periods of positive output
gaps (-0.1) and strongly procyclical during periods of negative output gaps (0.6). These cyclicality
coefficients are a product of different fiscal policies in current spending and in capital spending. On the
one hand, current spending exhibits moderately strong procyclicality over the period (0.5) that is
particularly prevalent during periods of negative output gaps (0.5). On the other hand, capital spending
exhibits slightly weak counter-cyclicality over the entire period (-0.2), much stronger counter-cyclicality
during periods of positive output gaps (-0.5) and is neutral during periods of negative output gaps (0.0).
This suggests that the government has been increasing current spending during periods of booms and
has been cutting it during periods of economic bust, reinforcing or accentuating the business cycle. The
results for capital cyclicality suggest that the government has been cutting capital expenditure during
31
periods of economic boom, perhaps to meet its fiscal constraints, while, on average, has been reluctant
to use capital spending as a stabilizing lever during periods of economic bust. Mexico’s government
might benefit from coordinating it fiscal policy in both current and capital spending such that they aim at
stabilizing the business cycle to reduce its volatility, while at the same time sustaining sufficient capital
investment to meet its development and medium- to long-term growth objectives.
Figure 1.39: Public Sector: Cyclicality of Spending by Economic Classification, 1995-2021
(Correlation coefficient of cyclical components of spending and real GDP)
All time Good times Bad times
Procyclical spend - Countercyclical

1.0 0.8
0.8 0.7 0.7
0.6 0.5 0.5
0.6 0.4 0.4 0.5 0.5 0.5
0.4 0.3
ing
0.4 0.3 0.3 0.3 0.2
0.2 0.1
0.1
0.2 0.0 0.0 0.0
0.0
-0.2 -0.1 0.0 0.0 -0.1
-0.2 -0.1
-0.4 -0.2 -0.3
-0.6 -0.5 -0.5-0.5 -0.5 -0.5
spending
-0.8
-1.0
Grants
Total
Pensions
Capital
Current
Interest
Personnel
Other current
Financial investment
Transfers: participaciones
Transfers: aportaciones + subsidies
Sources: SHCP, INEGI, IMF WEO (April, 2022 Vintage), and WB staff calculations.
Note: Good times are defined as periods of positive output gaps. Bad times are defined as periods of negative output gaps.
28. Current spending become more procyclical and capital spending become countercyclical after
the financial crisis of 2008. As a result, in recent years, current spending has been accentuating the
business cycle while capital spending has been used to counter-act the business cycle (see chart).
Depending on the size of the fiscal multipliers, current and capital spending policies are likely to be
delivering opposing forces on medium and long-term growth. Current spending policy might be
accentuating the business cycle and capital spending policy, while countercyclical, might not be
maximizing, and in fact, could be jeopardizing, the long-term growth prospects of the economy if it is
the only fiscal tool relied on to stabilize output. Since current spending tends to be rigid and politically
costly to adjust downwards, the government might be over-relying on less politically costly adjustments
in capital spending.
32
Figure 1.40: Public Sector Cyclicality of Spending, 2002-2021
(10-year rolling window correlation coefficients of cyclical com ponents of spending and GDP, + =procyclical, - =counter -
cyclical)
1.0
Total spending Current Capital Financial investment
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Sources: SHCP, INEGI, IMF WEO (April 2022 Vintage), and WB staff calculations.
Notes: SHCP economic classification rearranged to include more categories in current spending. Spending deflated by GDP
deflator from IMF WEO. Cyclical component of real spending and real output generated using HP filter lambda=100. Forecast
data from IMF for 2022-2027. GDP growth in 2022-2027 set to GDP growth projections in WEO. All other spending growth in
2022-2027 set to real expenditures growth projections in IMF WEO.
29. High levels of budget rigidity will likely impact on the quality of fiscal adjustment. Previous
studies suggest that the federal budget in Mexico contains a moderate-to-high degree of expenditure
rigidity, which may undermine the quality of fiscal adjustment efforts, reduce the incentives to improve
allocative efficiency, and likely result in public investment bearing the blunt of the expenditure cuts
during fiscal consolidation episodes.24 Herrera and Olaberria (2020) find that high levels of budget
rigidities are associated with: (i) higher financing needs; (ii) increase likelihood of a country getting into
fiscal distress and reduced ability to start fiscal adjustment; and (iii) more inefficient levels of public
spending.25
Fiscal balances and debt
30. Net public debt increased steadily in the aftermath of the 2008 global financial crisis. Net
public sector debt increased 21.1 pp of GDP between 2007 and 2016, first because of a fiscal stimulus
aimed at jump-starting the economy in the aftermath of the global financial crisis, and second, due to a
temporary increase in the deficit aimed at supporting the implementation of the constitutional reforms
of 2013 –with the expectation that those reforms would ultimately accelerate growth, boost tax
revenues, and eventually decrease deficit and debt. 26 By end-2016, net public debt reached 49.9 percent
of GDP. The government started to implement fiscal consolidation efforts to contain the growing level of
indebtedness.27 The original medium-term fiscal forecast for 2014-2019 envisaged a peak of 41 percent
of GDP by 2025 and a decline thereafter. However, actual fiscal deficits for 2014-2016 were higher than
anticipated. Fiscal consolidation was largely achieved by cutting public investment, with primary
surpluses in 2017-2019 that contributed to reduce net debt to 44.5 percent by 2019.
Figure 1.41: Historical Balance of the Public Sector Borrowing
24
World Bank (2015) finds that about 81 percent of Mexico’s federal budget is somewhat rigid. Echeverry, Bonilla, and Moya
(2006) puts rigid spending in Mexico at 76 percent of primary spending.
25
Herrera and Olaberria (2020) estimate the rigid component of the wage bill to be around 70 percent.
26
See pages 161 and171-172 on Criterios Generales de Política Económica 2014 at
https://www.finanzaspublicas.hacienda.gob.mx/work/models/Finanzas_Publicas/docs/paquete_economico/cgpe/
cgpe_2014.pdf.
27
Net debt had reached 51.4 percent of GDP excluding Banxico’s surplus according to SHCP’s calculations at the time. INEGI’s
updates to the GDP estimate later reduced that ratio to current 49.9 percent.
33
Requirements/Net debt (percentage of GDP)
HBPSBR with Banxico's surplus
55
51.7
50.0
49.9
49.6
50
47.2
46.7
44.9
44.5
45
42.6
40.0
40
37.2
37.2
36.8
36.0
35.0
33.2
35
32.9
32.5
31.1
30.9
30.6
29.5
28.8
30
25
20
e
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: SHCP.
31. Mexico entered the pandemic with fiscal space to support the economy, yet the policy
response was rather limited, especially compared to other emerging and advanced economies.
Striking the right balance between the government’s pledges of fiscal responsibility and curbing public
debt and growing spending pressures, amid a subdued economic outlook and tighter financial
conditions, has proven to be an increasing difficult task. By year-end 2022 the public deficit would have
grown 44.2 percent in real terms during the administration or 9.6 percent on average every year,
reaching 3.1 percent of GDP or 12 percent of total federal public spending, up from the 8.9 percent
recorded in 2018. Mexico’s debt level itself remains manageable and it is low compared to peers.
34
Figure 1.42: Primary balance (% of GDP) Figure 1.43: Public Sector Borrowing
28
Requirements (% of GDP)
Primary balance with Banxico's surplus PSBR with Banxico's surplus
2 0
1.8
-1
-0.8
1.1
1
0.6 -2
0.4
-2.2
-2.3
0.1
-2.5
-
0 -3
-0.3
-3.3
-0.4
-3.7
-4
-3.7
-3.7
-0.6
-3.8
-0.6
-3.9
-4.0
-4.0
-0.8 -0.9
-4.2
-1
-1.0
-4.5
-5
-1.1
-4.9
-1.2
-2 -6
2022e
2022e
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Figure 1.44: Public deficit (% of GDP) Figure 1.45: HBPSBR (MXN 2022 trillion)
15
-
14
-1.0
13
12
-2.0
-2.1 -2.1
-2.3 11
-2.5 -2.6
-3.0 -2.9
-3.1 -3.1 10
-3.6 -3.6 9
-4.0
-3.7
2022e
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022e
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: SHCP. Excludes Banxico’s surplus (2015, 2016 and 2017) and Source: SHCP.
the use of the Stabilization Fund (2019 and 2020), which are
revenues that are neither recurrent nor the result of fiscal policy.
Figure 1.46: General Government Gross Debt (% Figure 1.47: Expected change in GG Gross Debt
of GDP) 2021-2027 (percentage points)
100 20
93
16.7
15.9
90
15
80.6
80 11.1
72.9
72.4
10
68.3
67.5
66.1
70
64.6
5
57.6
60 2.5
1.3 1.6
50 0
-0.5
-1.5
36.3
35.9
40 -5 -3.0
30
-10
20 -11.7
-15
10
-20
0 -17.6
Chile
Chile
Asia
Asia
Argentina
Colombia
Argentina
Colombia
Brazil
Uruguay
Brazil
Uruguay
Peru
Peru
LATAM
LATAM
Mexico
Mexico
G-20 EMEs
G-20 EMEs
EMEs and MICs
EMEs and MICs
Source: IMF, Fiscal Monitor April 2022. Source: IMF, Fiscal Monitor April 2022.
28
Broadest measure of public deficit. Includes non-budget items, the effect of inflation and corrections to the Public Deficit to
take into account the economic effect of the formation or spending of financial assets. The difference between the public deficit
and PSBR is explained by extra-budgetary financing, such as that of development banks and the Investment Infrastructure Fund
(FONADIN), the use/spending of financial assets (such as those in the Stabilization Fund, trust funds or the Treasury) and the
effects of inflation on indexed debt.
35
Figure 1.48: Level of Indebtedness and gross financing needs (2022)
General government gross debt, 2022 (% of GDP)
120
Sri Lanka
Morocco
100
Dominican Republic Brazil Egypt
India
80 Croatia Hungary Argentina
Malaysia South Africa Pakistan
Uruguay
Colombia Thailand
60
Qatar Poland Philippines
Turkey
40 Chile
Oman Mexico
Peru
Indonesia Saudi Arabia
20 Romania
Russia
Kuwait
0
0 5 10 15 20 25 30 35 40
Gross financing needs, 2022 (% of GDP)
Source: IMF, Fiscal Monitor April 2022.
32. According to the World Bank’s debt sustainability analysis (March 2022), public gross debt is
forecast to hover around 60 percent of GDP over the medium term. Real GDP growth is forecast at
about 2 percent over the medium term. The primary balance is projected to gradually improve from -0.7
percent of GDP in 2022 to 0.6 percent in 2027. Public gross financing needs are projected at 10.9-11.8
percent of GDP between 2022 and 2027. Under the baseline, gross public debt is projected to remain
under 70 percent over the medium term (59.3 percent by 2027). This is aligned with the DSA in the
IMF’s Article IV of 2021. SHCP’s Debt Sustainability Analysis presented in the 2022 Annual Borrowing
Plan envisages a less optimistic picture than the IMF’s Article IV analysis. With 75 percent probability,
gross HBPSBR remains below 72 percent over the medium term, even assuming a real GDP growth rate
of 3 percent and without testing for shocks to the primary balance (assumed to reach 0.8 percent by
2026).
Figure 1.49: Contribution to Changes in Public Debt
36
Source: World Bank (March 2022)
Box: Fiscal risks deriving from PEMEX and CFE

Petróleos Mexicanos
Pemex’s financial standing has not materially improved over time. In 2021, the federal government
transferred Pemex MXN460.6bn, comprised of MXN202.6bn for debt amortizations, around
MXN143.3bn in tax benefits to face regular spending, and MXN114.7bn for the Dos Bocas Refinery 29.
Excluding the MXN345.9bn, Pemex’s real financial balance in 2021 would amount to -MXN280.6bn
instead of the official MXN65.3bn, close to 300 percent higher in real terms than the deficit recorded
in 2018.
Figure 1.50: Crude production (million barrels per Figure 1.51: Financial Balance (MXN 2021
day) billion)
2.58
2.55
2.55
2.52
Official Excl. Transfers and Tax Benefits

2.43
2.27
65.3
2.15
30.3
1.95
1.81
-71.7 -71.7
1.69
1.68
1.66
-107.2
-118.8
-188.5
-280.6
2018
2019
2020
2021
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: Pemex, Anuario Estadístico 2020 and Estadísticas Petroleras Source: SHCP, Estadísticas Oportunas de Finanzas Públicas and
Febrero 2022. 4th Quarterly Reports of each year.
Further reflecting the fragility of Pemex’s finances, its balance in 2021 should have been a
MXN125.5bn surplus, since: (i) the approved goal was a MXN92.7bn deficit, (ii) the government’s
transfer of MXN202.6 bn transfer, which should have resulted in an improvement of the financial
balance in the same amount, and (iii) Article 1 of the Revenues Law states that Pemex’s revenues
29
See page 42 on SHCP’s 4 th Quarterly Report. Transfers from SHCP to PEMEX are made via SENER. The latter’s spending was
MXN 272.2 bn above program. Excluding the MXN 202.6 bn for debt amortizations leaves MXN 69.6 bn, which must have been
for the Dos Bocas Refinery. The original budget for the project in 2021 was MXN45.1bn, according to the PEF.
37
from the first two dollars above the US$42.1dpb oil price employed in the 2021 Economic Package
estimates (equivalent to MXN15.6bn, from Criterios 2021 sensitivities) had to be used to improve the
balance. An observed average oil price of US$63.5dpb (some 50.8 percent higher than the program),
ceteris paribus, should have meant additional revenues for Pemex of MXN130.1 bn. Pemex, however,
had to use MXN60.2bn from the government’s transfers to face payment obligations.30
Pemex’s debt continues to grow. The government transferred MXN202.6bn in 2021 and MXN97.1bn
in 2019, yet PEMEX’s debt has increased from US$105.8 bn in 2018 to US$109.3 bn in 2021. The pace
of increase has moderated after the US$45bn increase between 2012 and 2018 but the large amount
of debt and recurrent financial losses have placed increasing spending pressure. Upcoming debt
amortizations and mounting pressures to face long-delayed payments to contractors are expected to
continue requiring support from the federal government.
Figure 1.52: Pemex total debt (USD billion) Figure 1.53: Pemex debt maturity profile (USD
billion)
120
113.2
24.6
109.3
25
105.8
105.2
103.0
96.0
100
86.8
20
77.7
80
64.3
15
60.5
56.0
60
53.8
52.3
49.9
48.4
46.1
43.6
43.3
10
8.8
37.1
8.4
40
7.6
7.5
7.4
6.2
5.0
4.9
5
4.0
20
3.2
2.9
2.3
0 0
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: Pemex, Annexes to Consolidated Financial Statements. Source: Pemex. March 2022 Investor’s presentation
Comisión Federal de Electricidad (CFE)
Mexico’s public utility financial balance has been deteriorating since 2016. CFE managed to keep its
balance basically constant in 2019-2020 with respect to 2018, despite declining transfers in real terms
from the federal government and the sharp drop in sales that came with the COVID-19 pandemic.
Notwithstanding, the natural gas shortage of February 2021, associated to the extreme winter
weather in Texas, and the generalized increase in the price of natural gas and other fossil fuels used
for power generation in 2021 increased CFE’s costs. This, combined with continued weak sales
associated to the Covid-19 pandemic, translated into MXN60.6bn losses in its 2021 financial balance
compared to the MXN28.5bn surplus originally approved by Congress. CFE did not require additional
transfers from the government beyond the MXN70.3bn approved in the 2021 Budget.
CFE Financial Balance 2013-2021 (MXN billion)

A: Official Balance | B: Balance excluding government’s transfers | C: B in 2021 bn pesos
2013 2014 2015 2016 2017 2018 2019 2020 2021
A 5.86 31.75 28.83 161.10 32.26 18.14 5.98 0.00 -60.54
B -1.86 28.36 28.51 131.10 -33.66 -63.27 -69.20 -70.00 -130.82
30
In 2021, the federal government’s oil revenues surplus amounted to only MXN21.7bn, even with the observed average oil
price of US$63.5dpb (or 50.8 percent higher than the one used for the budget). This is mainly explained by a MXN73.3bn tax
benefit granted by SHCP to Pemex in February and numerous tax deferrals granted by the SAT to Pemex along the year (see
https://dof.gob.mx/nota_detalle.php?codigo=5639465&fecha=27/12/2021). The latter allowed Pemex to pay November and
December 2021 royalties (around MXN70bn combined) until January and February 2022, respectively.
38
C -2.74 40.01 39.11 170.35 -40.97 -73.38 -77.07 -75.01 -130.82
Source: SHCP, Estadísticas Oportunas de Finanzas Públicas.
CFE revenues continue to underperform in 2022. SHCP’s first quarterly report for 2022 shows that
the firm is already short on this year’s financial goal by MXN18.4bn 31, due to lower-than-expected
revenues by MXN20.7bn. Together with the suspension in 2019 of the Long-Term Power Auctions that
were helping reduce average generation costs, new investments in current and additional
infrastructure, the reversal of improvements to the pension scheme made in 2016, and the risks from
its 21 ongoing arbitration disputes32, are expected to put additional strain on the public sector’s
finances and increase the potential need for extra transfers in 2022 and onwards.
33. Deteriorating prospects for the economy, tightening conditions in financial markets, the
financial position of Pemex and CFE, and political economy issues, all else equal, are likely to result in
a widening deficit over the short- and medium-term and will continue adding pressure to public
finances. SHCP’s Criterios 2022 medium-term fiscal forecasts33 suggest that, in order to stabilize debt-to-
GDP, a reduction of the fiscal deficit from 3.1 percent of GDP in 2022 to 2.3 percent by 2025 would be
required, assuming real GDP grows 2.5 percent from 2025 onwards. However, SHCP’s preliminary
estimates for 2023 presented in April 2022 assume a fiscal deficit of 3.1 percent of GDP in 2022, given
revenue and spending trends. Since then, the short- and medium-term economic outlook has worsened.
Stagnant private investment and scarring from the COVID-19 pandemic weigh on potential output, on
top of long-standing factors that have hindered productivity growth such as informality, financial
exclusion, and market concentration. Weakening growth prospects represent a significant downside risk
for tax revenues. Tighter-than-expected financial conditions will further debilitate the recovery and
public revenues and will also reflect into higher borrowing costs for the public sector.
Additional fiscal space is needed to address spending pressures
34. Key infrastructure projects going over budget, increasing pension and social spending, rising
debt servicing cost are likely to put upward pressure on expenditures. Projected increases in revenues
in 2023 would not suffice to fully cover the additional planned spending to implement the
administration’s priority programs and infrastructure projects in 2023, which may require reassigning
resources from other areas (i.e., capital spending) or tolerating a higher deficit. Aversion to private
investment in the energy sector will likely continue to take an increasing toll on Pemex’s and CFE’s
finances. While the 2021 constitutional reform in the energy sector was halted, other legal, regulatory,
and administrative measures will continue rendering the 2013 reform ineffective, as has been the case
in 2019-2021, hurting the business environment. Pemex’s challenges in terms of rising E&P costs,
increasing pressure to invest downstream, diminishing fuel sales and escalating debt service mean most
likely it will continue relying on financial support from the federal government. On the other hand,
higher oil prices should somewhat compensate (at least temporarily) for sluggish tax revenues.
Figure 1.54: Special spending items (% of GDP)
31
See page 32 at https://www.finanzaspublicas.hacienda.gob.mx/work/models/Finanzas_Publicas/docs/congreso/infotrim/
2022/it/01inf/itinfp_202201.pdf.
32
See page 55 at https://www.cfe.mx/finanzas/reportes-financieros/Documents/2021/4to%20Trimestre%202021-1.pdf.
33
See pages 100-102 at https://www.finanzaspublicas.hacienda.gob.mx/work/models/Finanzas_Publicas/docs/
paquete_economico/cgpe/cgpe_2022.pdf.
39
Pensions Programs in Constitution Pemex's investment Financial cost
12
10 10.3
9.1 9.5
3.2
8 2.7
2.8
1.4
6 1.3
1.2
1.4 1.6
1.1
4
2 4.1 4.1 4.1
0
2021 2022 2023
Source: PEF 2021, PEF 2022 and Precriterios 2023. Assumed real growth rate of 5.0% (5-
year average 2017-2022) to estimate pension spending in 2023, and of 10.8% to
estimate Pemex’s physical investment (average 2018-2022). The “Programs in
Constitution” items includes the pension for the elderly, transfers for disabled, health
insurance for informal workers and scholarships for grades 1 to 12.
35. Significant and more sustained efforts will also be required on public infrastructure
investment to support higher growth rates over the medium term and to reinitiate the converge
towards the OECD. Higher growth rates over the medium term require a higher level of public
infrastructure investment. Over the past decade, public investment and capital transfers have acted as a
buffer, hinting at very high levels of budgetary rigidities. The recent fiscal adjustment shows a
concerning trend in terms of squeezing growth-enhancing expenditure as the primary tool to achieve
fiscal consolidation, with a large decline in public investment. This reliance on reducing capital spending
can have negative medium- to long-term effects, particularly as public investment is already very low in
Mexico compared to its regional and structural peers.
36. In addition, pressures on public service delivery spending are building up and would need
fiscal space from the revenue side to avoid increasing debt in the post-COVID19 period . Additional
resources will be needed to address pressing spending pressures without further increases in the level
of indebtedness (Chapter 2 discusses potential areas to increase revenue collection and potential
distributional implications of such measures). Resources are needed to support human capital through
quality education, continue improving the health system that needs to attend a changing burden of
disease and likely future shocks such as the one prompted by the recent pandemic. Resources are
needed to maintain the existing social programs while expanding social assistance in critical areas to
tend to the most vulnerable groups. The impact of an aging society and demands of an upper middle-
income country are expected to exert additional fiscal pressure. Aging of the population will continue to
place pressures on pension spending and Mexico’s public healthcare system. Further investments in
education are required to foster the potential of the workforce, increase export competitiveness, and
shift towards more complex sectors.
40
Appendix 1.1: Administrative Units of Government Spending in BOOST
Administrative Spending Units by Level of Government
A. Central C. Estados
1 Poder Legislativo 01 Aguascalientes
10 Economía 02 Baja California
11 Educación Pública 03 Baja California Sur
12 Salud 04 Campeche
13 Marina 05 Coahuila de Zaragoza
14 Trabajo y Previsión Social 06 Colima
15 Desarrollo Agrario, Territorial y Urbano 07 Chiapas
16 Medio Ambiente y Recursos Naturales 08 chihuahua
17 Procuraduría General de la República 09 Distrito Federal
18 Energía 10 Durango
19 Aportaciones a Seguridad Social 11 Guanajuato
2 Oficina de la Presidencia de la República 12 Guerrero
20 Bienestar 13 Hidalgo
20 Desarrollo Social 14 Jalisco
21 Turismo 15 México
22 Instituto Nacional Electoral 16 Michoacán de Ocampo
23 Provisiones Salariales y Económicas 17 Morelos
24 Deuda Pública 18 Nayarit
25 Previsiones y Aportaciones para los Sistemas de Educación Básica, Normal, Tecnológica y de Adultos 19 Nuevo León
27 Función Pública 20 Oaxaca
28 Participaciones a Entidades Federativas y Municipios 21 Puebla
3 Poder Judicial 22 Querétaro
30 Adeudos de Ejercicios Fiscales Anteriores 23 Quintana Roo
31 Tribunales Agrarios 24 San Luis Potosí
32 Tribunal Federal de Justicia Administrativa 25 Sinaloa
32 Tribunal Federal de JusticiaÂ Administrativa 26 Sonora
33 Aportaciones Federales para Entidades Federativas y Municipios 27 Tabasco
34 Erogaciones para los Programas de Apoyo a Ahorradores y Deudores de la Banca 28 Tamaulipas
35 Comisión Nacional de los Derechos Humanos 29 Tlaxcala
36 Seguridad y Protección Ciudadana 30 Veracruz de Ignacio de la Llave
37 Consejería Jurídica del Ejecutivo Federal 31 Yucatán
38 Consejo Nacional de Ciencia y Tecnología 32 Zacatecas
4 Gobernación D. Municipalidades
40 Información Nacional Estadística y Geográfica Include 2465 municipalies.
41 Comisión Federal de Competencia Económica
42 Instituto Nacional para la Evaluación de la Educación
43 Instituto Federal de Telecomunicaciones
44 Instituto Nacional de Transparencia, Acceso a la Información y Protección de Datos Personales
45 Comisión Reguladora de Energía
46 Comisión Nacional de Hidrocarburos
47 Entidades no Sectorizadas
48 Cultura
49 Fiscalía General de la República
5 Relaciones Exteriores
6 Hacienda y Crédito Público
7 Defensa Nacional
8 Agricultura y Desarrollo Rural
8 Agricultura y Desarrollo Rural
8 Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación
9 Comunicaciones y Transportes
B. Parastatales
19 Aportaciones a Seguridad Social
50 Instituto Mexicano del Seguro Social
51 Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado
52 Petróleos Mexicanos
53 Comisión Federal de Electricidad
41
Appendix 1.2: Federal Spending by Budgetary Unit
Federal Spending, by Budgetary Administrative Unit
(Admins ordered by cumulative absolute contribution to growth in total spending, 2018-2021
Cumulative
Absolute
Contribution
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021 2017-2021
(in percent of GDP)
Total spending (in percent of GDP) 18.7 18.6 18.5 20.1 19.6
(in percent of total spending) (in percent, y/y)
Total spending (in percent of total spending) 100.0 100.0 100.0 100.0 100.0 .. 7.0 3.2 3.8 9.6 ..
(contribution to growth, in percent, y/y)
18 Energía 0.1 0.1 2.8 1.1 6.2 .. 0.0 2.8 -1.7 5.7 10.2
19 Aportaciones a Seguridad Social 15.5 16.2 17.6 18.2 18.3 .. 1.9 1.9 1.3 1.8 7.0
23 Provisiones Salariales y Económicas 8.5 6.0 2.5 2.8 1.8 .. -2.1 -3.4 0.4 -0.8 6.6
28 Participaciones a Entidades Federativas y Municipios 18.9 19.3 19.5 18.0 17.8 .. 1.8 0.8 -0.8 1.6 4.9
20 Bienestar 0.0 0.0 3.2 3.8 4.1 .. 0.0 3.3 0.7 0.6 4.7
24 Deuda Pública 9.9 10.3 11.1 11.6 10.0 .. 1.1 1.1 1.0 -0.7 3.9
33 Aportaciones Federales para Entidades Federativas y Municipios 16.2 15.9 16.4 16.4 15.3 .. 0.8 1.1 0.6 0.4 2.8
9 Comunicaciones y Transportes 2.4 3.1 1.5 1.2 1.1 .. 0.9 -1.6 -0.2 0.0 2.8
30 Adeudos de Ejercicios Fiscales Anteriores 0.5 1.4 0.2 0.3 0.1 .. 1.0 -1.1 0.1 -0.2 2.4
20 Desarrollo Social 2.4 2.3 0.0 0.0 0.0 .. 0.1 -2.3 0.0 0.0 2.3
4 Gobernación 1.7 1.8 1.0 0.2 0.2 .. 0.3 -0.9 -0.7 0.0 1.9
11 Educación Pública 7.3 7.1 7.3 7.3 7.2 .. 0.3 0.5 0.3 0.6 1.6
36 Seguridad y Protección Ciudadana 0.0 0.0 0.7 1.3 1.0 .. 0.0 0.7 0.7 -0.1 1.5
10 Economía 0.2 0.2 0.2 0.9 0.1 .. 0.0 0.0 0.7 -0.8 1.5
7 Defensa Nacional 1.8 1.8 2.3 2.5 2.7 .. 0.1 0.6 0.3 0.4 1.4
12 Salud 3.2 2.8 2.7 3.3 3.4 .. -0.2 0.0 0.7 0.4 1.3
34 Erogaciones para los Programas de Apoyo a Ahorradores y Deudores de la Banca 0.9 0.9 1.1 0.9 0.2 .. 0.1 0.3 -0.2 -0.7 1.2
21 Turismo 0.2 0.2 0.1 0.3 0.7 .. 0.1 -0.2 0.2 0.5 0.9
6 Hacienda y Crédito Público 1.1 1.0 1.1 1.6 1.3 .. 0.0 0.0 0.6 -0.2 0.8
8 Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación 1.5 1.6 1.4 1.0 1.0 .. 0.2 -0.2 -0.3 0.1 0.8
22 Instituto Nacional Electoral 0.4 0.5 0.3 0.3 0.5 .. 0.2 -0.2 0.0 0.3 0.7
14 Trabajo y Previsión Social 0.1 0.1 0.6 0.6 0.5 .. 0.0 0.5 0.1 -0.1 0.6
17 Procuraduría General de la República 0.4 0.4 0.3 0.0 0.0 .. 0.0 0.0 -0.3 0.0 0.4
40 Información Nacional Estadística y Geográfica 0.2 0.2 0.2 0.3 0.1 .. 0.0 0.1 0.1 -0.2 0.4
49 Fiscalía General de la República 0.0 0.0 0.0 0.3 0.3 .. 0.0 0.0 0.3 0.0 0.4
3 Poder Judicial 1.5 1.5 1.4 1.4 1.3 .. 0.1 -0.1 0.1 0.1 0.4
15 Desarrollo Agrario, Territorial y Urbano 0.4 0.5 0.4 0.3 0.4 .. 0.1 -0.1 -0.1 0.1 0.3
16 Medio Ambiente y Recursos Naturales 0.9 0.9 0.8 0.7 0.8 .. 0.1 -0.1 0.0 0.1 0.3
13 Marina 0.8 0.8 0.7 0.7 0.8 .. 0.0 0.0 0.0 0.2 0.3
5 Relaciones Exteriores 0.3 0.3 0.3 0.2 0.3 .. 0.1 -0.1 0.0 0.1 0.2
48 Cultura 0.3 0.3 0.3 0.3 0.3 .. 0.0 0.0 0.0 0.1 0.1
25 Previsiones y Aportaciones para los Sistemas de Educación Básica, Normal, Tecnológica y de Adultos 0.9 0.8 0.8 0.8 0.8 .. 0.0 0.0 0.1 0.0 0.1
38 Consejo Nacional de Ciencia y Tecnología 0.7 0.6 0.6 0.5 0.5 .. 0.0 0.0 0.0 0.0 0.1
1 Poder Legislativo 0.4 0.3 0.3 0.3 0.3 .. 0.0 0.0 0.0 0.0 0.1
2 Oficina de la Presidencia de la República 0.1 0.1 0.0 0.0 0.0 .. 0.0 -0.1 0.0 0.0 0.1
47 Entidades no Sectorizadas 0.2 0.2 0.2 0.2 0.2 .. 0.0 0.0 0.0 0.0 0.1
44 Instituto Nacional de Transparencia, Acceso a la Información y Protección de Datos Personales 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
42 Instituto Nacional para la Evaluación de la Educación 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
27 Función Pública 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
32 Tribunal Federal de Justicia Administrativa 0.1 0.1 0.1 0.1 0.0 .. 0.0 0.0 0.0 0.0 0.0
35 Comisión Nacional de los Derechos Humanos 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
45 Comisión Reguladora de Energía 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
31 Tribunales Agrarios 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
43 Instituto Federal de Telecomunicaciones 0.1 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
46 Comisión Nacional de Hidrocarburos 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
41 Comisión Federal de Competencia Económica 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
37 Consejería Jurídica del Ejecutivo Federal 0.0 0.0 0.0 0.0 0.0 .. 0.0 0.0 0.0 0.0 0.0
Grand Total 100.0 100.0 100.0 100.0 100.0
Source: WB Mexico BOOST.
42
Appendix 1.3: Spending by Economic Classification based on Authorities’
Categorization
Spending by Economic Classification, Data from SHCP
Composition of Federal Spending, by Economic Contribution to Growth in Federal Spending, by

Classification Economic Classification
Ministry of Finance Presentation Ministry of Finance Presentation
6000
20 Other
5000 Other 15 Financial investment
Financial investment 10 Capital investment

4000
Capital investment Transfers and subsidies
5
3000
Transfers and subsidies
0 Interest
2000 Interest
-5 Use of goods and
Use of goods and service service
1000 Compensation of
-10
employees Total
0 -15
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Source: SHCP and WB staff calculations.
43
Chapter 2: Revenues
Introduction
1. Increasing revenue collection can help address pressures on public service delivery and expanding
public investment while containing public debt. Resources are needed: (i) to support human capital
through quality education; (ii) to continue improving the health system, that needs to attend a changing
burden of disease and likely future shocks such as the one prompted by the recent pandemic; and (iii) to
maintain the existing social programs while expanding social assistance in critical areas. Demographic
changes (aging population) will continue to place pressures on long term care and certain pensions.
Moreover, higher growth rates over the medium term require a higher level of public infrastructure
investment (complemented with private resources). In this context, and despite the 2013 and 2021
reforms and recent adjustments, including strengthening tax administration and promoting
formalization, tax collection in Mexico remains below the LAC regional average and it is the lowest
among OECD countries. Tax expenditures (e.g., exemptions, zero rates, deductions) and other issues in
the tax structure continue to impose large fiscal costs. In the medium-term, a tax reform that renders
the resources to fulfil critical needs to build human capital and public service provision, particularly for
lower income families, and that supports a strong resurgence of the economy by financing critical
infrastructure spending is becoming urgent in the policy agenda.
2. Over the past decade, Mexico has taken important steps to improve tax policy and modernize the
tax administration. The 2013 Tax Reform helped raise additional non-oil tax revenue by around 3
percentage points (Figure 2.0, panel A), substantially lowering the revenue dependency from oil, with its
share in the total revenue going from more than 30 percent in 2010 to about 5 percent in 2020 (Figure
2.0, panel B). The current administration efforts to strengthen collection efficiency by combatting tax
evasion, optimizing tax collection without increasing or creating new taxes, and expanding the tax base
in e-commerce, have been highly positive and have paid off. In 2020, and despite the economic shock
triggered by the Covid-19 pandemic, tax revenue as a percentage of the GDP reached its highest point in
the last 10 years.
Figure 2.0. Past tax reforms helped to raise non-oil tax revenue
A. Revenue Structure (as share of GDP) B. Revenue Structure Mexico vs LAC and OECD
Note: Panel B contains data for 2019 and does not include Oil related revenues.
Source: Secretaría de Hacienda y Crédito Público (2020).
44
3. More recently, in December 2021, the government introduced a tax reform which included the
adoption of a new simplified regime (Régimen Simplificado de Confianza, or RESICO), effective as of
January 1, 2022. The reform seeks to reduce the procedures and time allocated to declare and pay taxes
in Mexico to increase tax compliance and facilitate formalization. The reform is also calibrated to ensure
revenue-neutrality in the short-term –i.e., maintaining similar effective rates for income tax for
individuals that meet the requirements. The reform also aims at lowering the barriers to formalization,
by simplifying the calculation, reducing the amount of taxes, and fighting the use of invoices—both false
and real—currently used to deduct taxes. The reform simplifies the tax regimes for MSMEs and self-
employed, and grants accelerated depreciation to support investment. The new simplified system for
income tax extends to a broader set of taxpayers (i.e., professionals) for individuals with a gross income
of up to MXN 3.5 million per year and legal entities with a gross income of up to MXN 35 million per
year. The new regime simplifies compliance with tax obligations, preventing the taxpayer from keeping
all the digital tax receipts or invoices, since the tax collection will be calculated directly on the total
income. The new regime allows for the auto-population of the monthly tax returns based on e-invoices,
simplifying tax filing for small and micro enterprises. While implementing measures to promote
formalization has become particularly urgent as the share of informal workers has increased post-
pandemic, it is rather premature to assess the impact of the reform in terms of revenue and
formalization resulting from decreasing compliance costs.
4. Mexico’s tax structure offers opportunities for a reform that can improve equity (enhancing
progressivity in the system), simplicity (reducing taxpayer compliance costs), and efficiency (including
fostering— not harming— the competitiveness of the economy). Currently, Mexico’s tax system does
not contribute enough to reduce income inequality, as reflected by the fact that the Gini coefficients for
market income and for disposable income are very close to each other, especially when compared to
other OECD countries (Castelletti, 2013; OECD, 2019). This is in part due to a relatively low tax-to-GDP
ratio (compared to other economies in Latin America and the OECD). Therefore, Mexico needs a tax
reform that enhances equity through more progressivity in the tax system, contributing to reduce the
negative distributional implications of market outcomes on lower income households, while creating
equal opportunities for all individuals and firms. The reform can be done in a way that makes the tax
system simpler, reducing complexity and tax-avoidance opportunities, easing compliance,
administration and enforcement costs. The reform can also be done while enhancing the country’s
competitiveness and improving efficiency, to help minimize distortions and support economic growth.
5. This chapter presents a menu of potential reform options that would help design an equity-
and-growth friendly reform. The chapter begins by looking at Mexico’s tax structure and comparing it
with other OECD and Latin American countries. It shows that the current tax system is not efficient, it
raises relatively low revenues and creates distortions that affect production and consumption decisions.
There is, therefore, scope for a revenue enhancement reform supporting the government’s commitment
to not raise tax rates, by incorporating best practices to improve efficiency and simplicity of the system.
This chapter looks separately at Personal and Corporate Income Taxes (PIT and CIT), Value Added Taxes
(VAT), selected Excise taxes, and the property tax. The options considered here follow a clear principle:
raise revenues in a growth-and-equity-friendly way by broadening tax bases, strengthening the overall
formalization. In other words, the chapter identifies reforms that align inclusiveness and growth goals to
contribute to significantly reducing efficiency-equity trade-offs. This chapter presents simulations on
how broadening the bases of income and consumption taxes would raise more revenue and reduce
distortions. A summary is presented in Error: Reference source not found at the end of the chapter.
45
Main characteristics of Mexico’s tax system
6. Tax revenues have increased but remain relatively low. Over the last decade Mexico has been
able to increase nonoil tax revenue (Error: Reference source not found). In 2013, a comprehensive tax-
reform eliminated tax benefits and preferential tax regimes, limited tax deductions, introduced dividend
and capital gains taxes, raised the top marginal income tax rate, and replaced the tax regime for small
businesses. The reform also eliminated a preferential VAT rate that had been applied in border regions.
They introduced new excise taxes on carbon, sugar-based beverages and high-calorie foods. The
negative excise tax on domestic fuel sales was phased out in the context of low international oil prices
and energy-sector reforms. As a result of this reform, government’s non-oil tax revenue rose from 9.6
percent of GDP in 2013 to 14.4 percent in 2020. Rising income tax collection contributed 1.8 percentage
points of GDP to the increase in non-oil tax revenue, fuel excise taxes contributed 1.8 percentage points,
VAT contributed 0.9 percentage points, and other excise taxes contributed the remaining 0.3 percentage
points. The total increase in non-oil tax revenue largely offset a decline in oil revenue equal to 4.4
percentage points of GDP.
Figure 2.1. Tax Revenues as a percentage of GDP
A. Evolution of Tax Revenues as % GDP B. Latin America and the Caribean (2019)
Source: SHCP and OECD Revenues Statistics.
7. Yet, tax collection in Mexico remains below the Latin America regional average and is the
lowest among OECD countries (Error: Reference source not found). Despite the increase in revenue
collection brough by the 2013 tax reform, Mexico had to deal with a parallel large decline in oil revenues
due to lower international prices and reduction in production. At 13.2 percent of GDP in 2019, non-oil
tax revenues remain well below the OECD and Latin America averages of 26 and 19 percent of GDP,
respectively. VAT revenues as share of GDP have flattened and remain much below that of other
countries, including most countries in Latin America.
8. Moreover, Mexico’s tax structure relies heavily on direct taxes compared to the OECD and,
particularly, LAC peers (Error: Reference source not found). Income tax revenue makes more than 40
percent of Mexico’s total tax revenue compared to 33 percent in the OECD. In fact, Mexico’s personal
income tax reduces gross income inequality more than any other personal income tax system in Latin
America. Indirect taxes in Mexico are underperforming due to large inefficiencies/tax expenditures in
VAT collections. Due to broad exemptions and zero-rating in the VAT regime, Mexico collects only 31.5
percent of the revenue that it could theoretically collect if VAT were to be applied at the standard rate
to all goods and services. By contrast, this VAT Revenue Ratio is 42.6 percent in Colombia, 55.1 percent
in Peru and 64.4 percent in Chile. Similarly, social contributions are well below regional peers and OECD
countries. Property taxes (recurring taxes on real property, on net wealth, on inheritances and
46
donations and on financial and capital transactions) in Mexico collect only 0.3 percent of GDP, compared
to 1.9 percent on average in the OECD.
Figure 2.2. Mexico’s tax structure relies heavily on income taxes (% of total tax revenues)
Mexico vs LAC countries
Mexico vs OECD countries
Mexico’s tax structure relies heavily on income taxes (% of GDP)

Mexico vs OECD countries
47
9. Mexico’s tax system does not contribute enough to reduce inequality. The Mexican tax system
is broadly progressive but given its relatively low collections, it makes only a limited contribution
towards the reduction of inequality. The analysis34 shows that fiscal policy in Mexico is broadly
redistributive, with the poorest 40 percent of the population being net beneficiaries of taxes and
spending (Figure 2.3). On their own, direct taxes are progressive (as measured by the Kakwani index 35),
with personal income taxes contributing to a moderate reduction in inequality of 0.02 Gini points ( Error:
Reference source not found). State payroll taxes are somewhat progressive, with much lower impacts
on inequality. On their own, indirect taxes were regressive in 2020 (with Kakwani coefficients that are
negative). Indirect taxes have insignificant impacts on inequality. When combined with the existing set
of direct transfers, the overall impacts of taxes and transfers on overall inequality is small compared to
that of other middle-income countries (Error: Reference source not found). The 2013 tax reform
improved the redistributive impact of the tax and transfer system (OECD, 2018). Among the reasons for
this are (i) high rates of informality; (ii) regulatory loopholes and (iii) need to strengthen tax
administration.
Figure 2.3. Mexico. The redistributive impact of taxes and transfers, 2020
(Cumulative interventions as a share of market income)
70%
60%
50%
40%
Benefit as a share of market income plus pensions
30%
20%
10%
0%
-10%
-20%
Market income plus pension deciles
PIT State payroll tax Benito Juárez Educación Básica
Benito Juárez Educación Superior Pension Adulto Mayor Jovenes Construyendo Futuro
Jovenes Escribiendo Futuro Transferencia discapacidad Transferencia Madres Trabajadoras
Transferencia Precios Garantias Transferencias Ganadero Palabra Transferencias Fertilizantes
Tandas para el Bienestar Sembrando Vida Produccion para bienestar
Electricity subsidies VAT Excises
Netcash position
34
The analysis uses the latest household survey to assess the distributional impact of taxes and transfers (see Appendix 1).
35
The Kakwani index is defined as the difference between the concentration index of taxes and the pre-tax Gini coefficient. If
taxes are more concentrated than income at the top of the distribution, then the Kakwani Index is positive and the system is
deemed progressive.
48
Source: World Bank Fiscal microsimulation model based on 2020 ENIGH.
Table 1.1. Mexico. Tax progressivity and Figure 2.4. Redistributive impact of taxes, transfers, and
marginal contribution to inequality, 2020 in-kind benefits
Marginal contribution 0.000
(Poverty -0.050
Kakwani
head count) (Gini Index) -0.100
Personal -0.150
income tax -0.013 0.019 0.284
-0.200
State
-0.250
payroll tax -0.004 0.001 0.125
-0.300
VAT -0.019 -0.003 -0.044
-0.350
Excise
taxes -0.006 -0.002 -0.015 -0.400
Source: World Bank microsimulation model based on ENIGH -0.450
(2020) and administrative data. -0.500
Note: Marginal contributions are estimated with respect to
J o ru s 0 1 6 )
liv (2 0 )
El P ia (2 0 1 3 )
Eswtate (2 0 1 7 )
rai (2 5 )
)
rd (2 )
Be rkey 2 0 1 )
20 )
Za uela 2 0 1 )
Maomcb(2 0 1 4 )
ol bl i 2 0 )
Ro Chil (2 0 0 )
ssi 2 0 )
la (2 1
n 2 )
Ve sot (2 6 )
m (2 7 )
Po ma 201 6 )
Ky Mon anm ia (2 1 7 )
SoRepu gro 2 0 1 7 )
Le land (2 0 1 )
Bodor 0 1 1)
ut b (2 0 )
Na atin(2 0 1 7 )
Uk xico (2 0 1 )
Me nya 2 0 1 )
14
Re ati 2 0 )
Tu ran ( 0 1 5
a ( 17
z ( 4
pC u a ( 1 5
Ru an ( 0 1 5
I (2 )
ma e ( 16
na ( 7
z ne r ( 7
ite nti (2 5 )
fri (2 0 5 )
l va u ( 1 6
7
K e us ( 0 1 7
Pa ibia (2 0 1 )
an Cro bia ( 0 1 3
ia 17
d S na 0 1
Un Arge pai n(2 0 16 )
ne ho 0 1
ne 0 2
m i 6
rgy te a 0 1
iti 2 )
market income for direct taxes and with respect to
h A lic 1
My men y (2 0
Sa er 0
u r i a (1 3
2
S ca 1
Ar gua
s
disposable income for indirect taxes. Ur
u
ni c
mi
Do
Source: CEQ indicator database and World Bank Fiscal
microsimulation model based on 2020 ENIGH.
Direct taxes
Personal income tax (PIT)
10. Mexico’s PIT is around 5 percent of GDP. While this number is relatively large compared to LAC
and weighs heavily in Mexico’s tax structure, it is (as percent of GDP) below the OECD average (see
Figure 2.2). This is mainly explained by large informality in the economy that reduces the base. Also, pre-
tax income inequality in Mexico translates into a large low-wage tax base with small contribution to
revenue collection due to the progressivity of PIT. While there is limited scope for increasing PIT revenue
through rate increases, the tax base can be expanded. Exclusions from the PIT are prevalent, such as
income on personal business activities and independent services. As a result, most of PIT collection is
Regressive
made from withholdings on wages and salaries, with very limited additional collection from the broader
base.
Figure 2.5. Direct Tax Progressivity and Marginal Contribution to Reducing Inequality
0.7 0.050
<------------------> Progressive
Marginal contribution to equity
0.045
0.6
0.040
(change in Gini points)
0.5 0.035
0.030
0.4
0.025
0.3
0.020
0.2 0.015
Kakwani Index
0.010
0.1
0.005
)
3
)
1
)
)
)
6
)
)
)
)
5
)
)
)
)
5
0
)
ep (2 )
7
0
)
0.0 0.000
3
)
0
lic 9
7
5
4
1
9
)
1
)
1
5
5
)
1
(2
1
1
1
7
1
b 0
1
1
1
1
0
0
es 1
1
0
1
0
0
0
0
1
0
u 0
0
1
0
0
(2
0
at 1
0
(2
0
(2
(2
(2
(2
0
(2
(2
0
(2
(2
(2
(2
(2
St (2 0
(2
2
(2
(2
(2
(2
a
i(
t
r
a
a
R a
ia
n
ia
ia
ka
s
ia
ic
as
o
a
s
c
el
n
an
u
ro
n
ia
ta
b
ed n
u
i
fr
ad
i
an
n
en
an g
ati
b
an
R
ai
o
u
m
it a
ar
d
za
w
ic ra
o
is
C
ez
n Ir
Sp
lv
a
rm
In
m
iL
jik
lo
el
ts
in ica
st
n
th
Za
ry
Sa
en
o
Es
B
Ta
o
Sr
Ta
o
A
o
R
u
o
C
m N
C
B
V
El
Iv
So
U
o
D
Kakwani Direct taxes Marginal contributions to equity (right axis)
49
Source: Commitment to Equity Database and World Bank Fiscal microsimulation model based on 2020 ENIGH.
Figure 2.6. The number of brackets seems excessive
Source: OECD Taxation Database
11. The number of PIT tax brackets seems excessive. Mexico has 11 PIT tax brackets compared the
OECD and LAC averages of close to four (Error: Reference source not found). All persons who obtain
income in cash, goods, credit, services, and prizes or donations that exceed 600,000 pesos are obliged to
pay personal income tax (PIT).36 Subordinate workers who receive income from wages make provisional
monthly payments on account of the annual income tax. Thus, the employer withholds the
corresponding tax in accordance with the monthly income tax rate, which increases with income (see
Appendix Error: Reference source not found). Workers earning less than the minimum wage are not
subject to withholding. Self-employed workers providing personal services are expected to make
provisional payments equivalent to 10 percent of income.
12. The income threshold for the top PIT rate in Mexico is high (Figure 2.7). The 2013 reform,
which increased the maximum marginal rate to 35 percent and introduced the tax on dividends,
together with the efforts to improve control, managed to increase income tax collection for individuals
from 2.5 to 3.4 percent of GDP between 2012 and 2019. Although the maximum marginal rate in Mexico
(35 percent applied to annual income above 3.5 million pesos) 37 is higher than the average for Latin
America (26.6 percent), it is well below the average of OECD countries (42.5 percent). Also, in Mexico,
the annual income earned to start paying the top PIT marginal rate is 28.7 times the average Mexican
salary, compared to 5.4 times in the average OECD country. Thus, despite higher tax rates fewer people
pay the top PIT rate in Mexico than in OECD countries. At the same time, the bottom statutory rate is
low and could be raised.
13. One factor constraining the collection of PIT in Mexico are the exemptions and deductions
that affect the integration of the tax base. The Ministry of Finance estimates that PIT tax expenditures
increased to 1.1 percent of GDP in 2020 (SHCP, 2020), and is now higher than Latin American and OECD
averages (Figure 2.8). Two thirds of them correspond to exemptions of pensions and benefits (Christmas
36
Income tax law, Article 90.
37
A taxpayer who declares 3.5 million would pay a fixed fee of 1.1 million. This means a rate of 31.4%, and a rate of 35% on the
margin. However, the maximum marginal rate was 55% in the early eighties, dropped through successive reforms to 28% in the
mid-2000s, rose to 30% in 2010 and 35% in the 2013 reform.
50
bonus, vacation bonus, overtime, among others) of employees. Beyond the revenue cost, tax incentives
increase the complexity of the tax system and reduce its effectiveness as an instrument to
promote equity if the uptake of tax incentives is higher among those with higher incomes. Similarly,
evidence show limited effectiveness of tax incentives for savings for retirement in promoting overall
savings and mostly point to the fact that taxpayers move existing savings to tax-preferred savings
instruments (Attanasio, Banks, and Wakefield 2004).
Figure 2.7. Top brackets kick in very late
A. Mexico vs selected OECD countries B. Mexico vs selected LAC countries
14. Generous tax allowances and exemptions alone represent 0.89 percent of GDP of loss PIT
revenue (Figure 2.8). They include medical expenses (0.04 percent of GDP), complementary
contributions for certain retirement accounts, payments of insurance premiums of pension plans and for
the acquisition of shares in societies below MXN152,000 annually (0.06 percent of GDP), education
expenditures (school transportation, and school tuition expenses) (0.013 percent of GDP), funeral
expenses, donations and real interest expenditure of mortgage loans. In addition, for subordinate
workers there are exemptions on income from incentives, bonuses, prizes, and profit sharing. Finally,
there is an exemption on pensions that are below 15 minimum wages. They could be eliminated, scaled
back, or converted into tax credits. Also, net capital gains in securities transactions and dividends are
taxed in a separate schedule. Therefore, although they are not considered fiscal expenditures, these
items usually represent a high proportion of the income of people at the top of the income distribution
and are subject to a flat rate of 10 percent since 2014, lowering their effective PIT rate.
Figure 2.8. PIT tax expenditures are high compared to LAC (2019)
51
Source: CIAT.
15. These deductions are regressive and increase income inequality. At least 0.5 percent of GDP of
these tax expenditures are inefficient or regressive. Over 70 percent of the benefits of existing
deductions in Mexico go to the top 10 percent of the pre-tax income distribution (Error: Reference
source not found). They are also regressive in relative terms, since they represent 0.5 percent of pre-tax
income distribution for the top 10 percent compared with nearly zero for those in the bottom 40
percent of the distribution (Error: Reference source not found). As such, they are both regressive
(Kakwani index is -0.311) and inequality increasing, amounting to an increase in the Gini of 0.1 points.
Figure 2.9. Absolute incidence of PIT Deductions Figure 2.10. Relative incidence of PIT
by decile, 2020 Deductions by decile, 2020
(share of total deduction benefit) (deduction as a share of market income)
70% 66% 0.5%
0.5%
60%
0.4%
50%
Share of total deductions
0.3%
40%
0.2%
30% 0.2%
0.2%
0.1%
0.1%
20% 0.1%
13% 0.1% 0.1% 0.1%
0.1%
10% 7% 0.0%
3% 5%
1% 2% 2%
0% 1% 0.0%
0%
Poorest 2 3 4 5 6 7 8 9 Richest Poorest 2 3 4 5 6 7 8 9 Richest
Market income plus pension deciles Market income + pension deciles
Source: World Bank fiscal microsimulation model based on ENIGH, 2020.

Simulations – Fiscal and Distributive impact
16. These issues point to significant gains in progressivity while also increasing potential
revenues. Using the fiscal microsimulation model, we assess the fiscal and distributional impacts of
different reform options to the PIT. These options include reducing the number of tax brackets,
increasing marginal tax rates, removing deductions, limiting exemptions, and the combination of
different policy options. The results of such simulations are presented in Table 2.2. Details of each
option are given in Appendix 2.1.
 Reducing the number of tax brackets in the PIT and changing bottom/marginal tax rates:
Option 1 assumes reducing tax brackets from 11 to 5 and setting the two-minimum wage
52
threshold (Figure 2.11). Option 2 assumes reducing the number of tax brackets to 4 and
increasing the marginal rates (Figure 2.12). Option 1 would improve the progressivity of the PIT
as it would reduce marginal rates at the bottom of the income distribution, while at the same
time increasing the marginal rate for higher incomes. This reform would also have a positive
impact on equity and poverty reduction compared to the baseline. However, it would lead to a
14.1 percent reduction in PIT revenue, equivalent to 0.48 percentage points of GDP. Option 2
would lead to a reduction in progressivity of the PIT and a 1.9 percentage point increase in
poverty. PIT collections would increase by 73.8 percent, equivalent to 2.5 percentage points of
GDP.
 Removing or reducing deductions: Option 3 assumes removing deductions for medical
expenses, mortgage interest and school tuition. This would increase PIT revenue by 1.6 percent
(equivalent to 0.05 percent of GDP). Given that most of these deductions largely benefit
individuals with higher incomes, the reform would increase slightly progressivity, reduce post-
tax inequality, while having no impact in poverty. Deductions would also become slightly less
regressive. Option 4 assumes reducing the global deductions. Current legislation sets a
maximum global limit to the sum of total deductions at 5 UMA 38 or 15 percent of gross income,
whichever is smallest. The proposed reform would reduce this limit to 3 UMAs or 10 percent of
gross income, whichever is smallest. This reform would have no significant impact on the
progressivity of PIT, nor would it have any impact on poverty or equity, but it would reduce the
regressivity of deductions. It would generate a 0.31 percent increase in PIT collections (0.1
percentage points of GDP). Option 5 assumes adding an individual PIT deductions limit
equivalent to one minimum wage. The legislation places individual limits (medical and funeral
expenses, donations, mortgage interest, school expenditures deductions) to people earning
more than MX$400,000 a month (about US$20,100). This would improve the progressivity of the
PIT, without a significant impact on poverty or inequality. PIT collections would increase by 0.37
percent (0.01pp. of GDP).
 Limiting exemptions on pensions: Option 6 assumes reducing the exemption on pensions to less
than 7 minimum wages. Current legislation places an exemption on pensions that have a value
that is below 15 times the minimum wage. PIT would become slightly more progressive,
increasing by 0.7 percent (or 0.02 percent of GDP), without an impact on poverty or inequality.
 Providing a general exemption: Most existing exemptions benefit individuals with higher
incomes, leading to vertical inequality. In addition, individuals working in large firms also benefit
more from exemptions than similar individuals working in small firms, simply because larger
firms provide benefits such as incentives, bonuses, prizes, and profit sharing, leading to
horizontal inequalities. Under Option 7, legislation could be modified to allow individuals a
choice between a fixed general exemption and the current exemption system. Under this
scheme, individuals would choose the maximum of the exemptions they are now eligible for and
a proposed general exemption of MX$88,588 a year on all taxable income. PIT collections would
fall by 45 percent (or 1.5 percent of GDP). A larger share of low and middle-income individuals
would be able to benefit from exemptions, this would lead to an improvement in progressivity
of the PIT system, a reduction in inequality, and an improvement in the progressivity of
exemptions. Option 8 assumes a general exemption applicable only to labor incomes (thus
excluding self-employed workers). This would lead to an improvement in progressivity of the PIT
system. The reform would cost 1.4 percent of GDP, with slightly smaller impacts on equity.
38
UMA stands for units of measurement and update (Unidad de Medida y Actualización), which serve as an economic reference
unit used to calculate payments, obligations or penalties for the government.
53
Figure 2.11. Option 1: Reducing the number of tax Figure 2.12. Option 2: Reducing the number of tax
brackets to 5, reducing the bottom rates brackets and increasing marginal tax rates
Marginal Rate Marginal Rate
40%
40%
35.0% 35%
35% 32%
30% 28% 33%

30%
25%
25% 23.5% 22%
21.4%
20%
20%
15%
15%
10.9%
Baseline Simulation Baseline Simulation
10% 10%
5% 5%
0.0%
0% 0%
- 50,000 100,000 150,000 200,000 250,000 300,000 - 50,000 100,000 150,000 200,000 250,000 300,000
PIT Thresholds PIT Thresholds
Source: World Bank Source: World Bank
54
Table 2.2. Alternative PIT Reform Scenarios
Policy Options 1 2 3 4 5 6 7 8 9 10 11
Remove
Reduce Reduce the Introduce
Reduce deductions Lower Exemption
Baseline brackets to 5 global individual General
brackets to for medical, pension on labor 3+6+8 3+6+8+2 3+6+7+2
2019 (2 min wage deduction deduction exemption
4 mortgage exemption income
threshold) limit limit
and tuition
Marginal
contribution to:
Personal Poverty
-0.010 -0.001 -0.044 -0.010 -0.010 -0.010 -0.010 -0.001 -0.002 -0.002 -0.023 -0.019
income tax reduction
Equality 0.019 0.022 0.017 0.019 0.019 0.019 0.019 0.014 0.014 0.015 0.014 0.015
Kakwani index 0.289 0.398 0.154 0.289 0.289 0.289 0.290 0.400 0.382 0.383 0.202 0.222
Decil 1 -0.231 -0.59 0.93 -0.23 -0.23 -0.23 -0.23 -0.39 -0.24 -0.24 0.88 0.22
Decil 2 0.01 -0.94 2.60 0.01 0.01 0.01 0.01 -0.29 -0.08 -0.07 2.23 1.52
Decil 3 0.59 -0.78 4.27 0.59 0.59 0.59 0.59 -0.06 0.17 0.18 3.10 2.40
Decil 4 1.40 -0.27 6.14 1.40 1.40 1.40 1.40 0.09 0.31 0.31 3.49 2.86
PIT as a share
Decil 5 1.96 0.15 7.26 1.97 1.96 1.96 1.96 0.27 0.44 0.44 3.92 3.38
of market
Decil 6 2.80 0.89 8.60 2.80 2.80 2.80 2.80 0.56 0.71 0.72 4.49 3.97
income
Decil 7 3.68 1.70 9.87 3.68 3.68 3.68 3.68 0.93 1.05 1.08 5.08 4.62
Decil 8 4.69 2.94 10.65 4.70 4.69 4.69 4.69 1.45 1.59 1.61 5.47 5.04
Decil 9 6.20 4.80 11.87 6.21 6.21 6.21 6.21 2.48 2.70 2.76 6.69 6.17
Decil 10 10.45 10.81 13.75 10.49 10.51 10.51 10.56 7.27 7.57 7.88 10.63 10.12
Δ% PIT Δ% PIT revenue -14.08% 73.80% 0.32% 0.37% 0.37% 0.71% -44.65% -41.03% -38.87% 14.14% 5.74%
revenue Δ% GDP (scaled)* -0.48% 2.50% 0.01% 0.01% 0.01% 0.02% -1.52% -1.39% -1.32% 0.48% 0.19%
Poverty Headcount ratio 22.85 22.28 24.76 22.85 22.85 22.85 22.85 22.51 22.66 22.66 24.02 23.57
Source: World Bank fiscal microsimulation model and estimates based on ENIGH 2018.
Note: Negative impacts on poverty imply an increase in poverty. Negative values of the incidence of PIT as a share of market income reflect the tax credit scheme, through
which PIT is actually a subsidy for low income individuals. Revenue impact is measured as a share of baseline PIT revenue observed in the household survey which amounts
to 53 percent of total PIT revenue collections according to administrative records, largely due to the fact that the household survey does not have information on top
incomes (see Appendix 1). Estimates on revenue collections in terms of GDP are scaled and assume that the same proportionality holds under the reform scenario.
Corporate Income Taxes (CIT)
17. Mexico’s CIT collection amounts to 3.6 percent of GDP compared to 2.8 and 3.2 percent of
GDP in OECD and LAC countries, respectively (Error: Reference source not found). This is mainly
explained by a high statutory tax rate in Mexico, which is, however, undermined by ample deductions
and special regimes. While in Mexico tax rates for CIT are by and large high compared to peer countries,
revenue performance is affected by holes in the tax system that provide ample opportunities to cheat
and evade taxation. Most of these loopholes are due to good intentions, such as encouraging
investment or simplifying tax administration, but in practice only a few of these special provisions meet
their objectives, and instead creating opportunities to evade.
18. Mexico’s CIT rate is high among both OECD and LAC countries (Figure 2.14). The CIT statutory
rate is 30 percent—above the U.S. federal rate of 21 percent and the average of OECD countries (24
percent). Nevertheless, due to accelerated depreciation on investments in buildings and machinery,
effective tax rates are closer to 20 percent, below the median of OECD countries (OECD, 2019). The
relatively high CIT rates can affect Mexico’s competitiveness, particularly given the recent trends in
reducing CIT rates among OECD countries. For example, the changes in the US corporate taxation will
not only make investment in other countries relatively less attractive, but also reduce the CIT tax base
elsewhere, including Mexico. This may have an impact on attractiveness of the Mexican economy due to
higher tax burden on corporations. The strong reliance on CIT revenue and rate differentials exposes
Mexico to potential revenue losses associated with profit-shifting and, over the medium-term,
relocation of production if the rate differential further widens. Consequently, there is limited room to
increase CIT rates.
Figure 2.13. Revenues from corporate income taxes are relatively high in Mexico
A. Mexico vs LAC countries B. Mexico vs OECD countries
Source: OECD Revenue Statistics Database
19. There is some space to broaden the tax base by eliminating some of the numerous
deductions, exemptions, special regimes, deferrals and administrative facilities. The application of
these differential treatments is very diverse in nature and together account for lost revenue of around
0.5 percent of GDP (Table 2.3). Tax expenditures of CIT related to deductions represent 0.11 percent of
GDP, three quarters of which come from deductions for the acquisition of internal combustion
automobiles and hybrid, electric or hydrogen-powered cars. International experience shows that these
sector- and size-contingent policies tend to hamper resource reallocation and firm growth. Productivity
of CIT is at the average of international peers and above several advanced economies. The revenue
agency (SAT) has been diligently pursuing collections against several corporations in recent months,
including in a number of high-profile settlements. The existence of this great variety of special
treatments, on the other hand, gives rise, especially in large companies, to aggressive tax planning that
reduces effective taxation.
Figure 2.14. Maximum Marginal Corporate Tax Rates

20. Recent measures to stimulate economic activity along the US border have contributed to
higher tax expenditures. In 2019 the government introduced tax incentives for businesses operating in
the northern border zone of Mexico. These included tax credits, which effectively reduce the corporate
income tax rate from 30 percent to 20 percent for legal entities, and a reduced value added tax (VAT)
from 16 percent to 8 percent. According to this measure, taxpayers in the border zone are entitled to a
credit equal to one third of the income tax due on their taxable income for the period. The tax credit
also applies to monthly estimated tax payments. The reduced income tax rate is applied to the taxable
income derived from activities performed in the northern region. The VAT reduction equals 50 percent
and is allowed for qualified sales of goods and services in the region. Although the authorities have
enacted a number of measures to prevent tax planning, recent experiences with tax rate differentials in
Mexico show that foregone revenue from tax arbitrage was substantial. Evidence for Mexico shows that
CIT revenues are affected by the design of the VAT (Ahmad et al, 2012). The presence of VAT
exemptions or lower rates do not only lead to evasion of VAT; they also lead to greater evasion of the
CIT, thereby exacerbating the loss of revenues caused by such policies. In sum, while the impact this
measure could have on promoting economic growth is uncertain, tax expenditures from this measure
are significant. According to SHCP, for the year 2020, tax expenditures from the incentives granted to
the taxpayers of the northern border region, in terms of VAT, CIT and VIT, amounted to 0.3 percent of
GDP.
Table 2.3. Tax expenditures from Corporate Income Taxes as a percentage of GDP (2020)
Tax expenditures % of GDP
Special or Sectorial Regimes 0.05
Deferrals 0.09
Administrative Facilities 0.02
Exemptions 0.05
Deductions 0.11
Employment Subsidy 0.21
Sub-total of tax expenditure from Corporate Income Taxes 0.5
57
Tax incentives in the northern border region 0.3
Source: SHCP (2020)
Property taxes
21. In Mexico, all levels of government are characterized by relatively low tax revenues, but
municipalities suffer the most. Mexico is one of the countries with the lowest level of subnational tax
collection as a proportion of total public revenue both in Latin America and the OECD (OECD, 2020). Sub-
national governments in Mexico only collect 0.7 percent of GDP in taxes compared to an OECD average
of 5 percent of GDP. Local or municipal taxes only represent 0.2 percent of GDP in Mexico compared to
3.9 percent of GDP in the OECD average. Among Latin American countries, Mexico is only above
Guatemala and Panama, and much lower than Argentina, Brazil, and Colombia.
Figure 2.15. Local Government Tax Revenue as a share of GDP (2019)
22. The most important source of tax revenues for the local governments are property taxes .
Collection from property taxes (“predial”) in Mexico is extremely low both in absolute and relative
terms. In 2019, Mexico collected 0.3 percent of GDP in revenues from property taxes, compared to
countries such as the United Kingdom, France and the United States that collected 4.1, 4 and 3 percent
of its GDP, respectively (Error: Reference source not found, Panel B). Similarly, Latin American countries
with similar levels of income to those of Mexico, such as Argentina, Brazil, Colombia and Chile, collected
2.6, 1.5, 1.8 and 1.1 percent of their GDP (Error: Reference source not found, Panel A). Low collection of
property taxes considerably reduces own resources of municipal governments.
23. A reform of property taxes can be a simple and efficient way to increase tax collection for local
governments. First, property taxes have a relatively stable tax base since the value of real estate does
not fluctuate as much as the income of companies or individuals. Second, it is a progressive tribute since
the amount of the tax increases according to the commercial value of the property. Third, since the
property tax is collected by the local government, the execution of public spending translates into goods
and services at the local level, which should decrease the resistance of taxpayers to pay the tax (Unda
and Moreno, 2015). Furthermore, property taxes are less harmful for growth than taxes on incomes:
recurrent taxes on immovable property are considered least distortionary relative to transaction taxes
(Arnold et al., 2011).
24. The potential to raise property tax revenues is huge in almost all municipalities in Mexico, but
there are important heterogeneities. One of the limitations for property collection in Mexico is related
to the federal transfer system, which does not encourage municipalities to increase tax collection. This
limitation affects all municipalities but has a greater impact on rural municipalities (Chavez Maza and
58
Lopez Toache, 2019). Thus, a reform of the federal transfer policy would encourage municipalities to
increase their own tax collection. However, the ability to raise revenues from property taxes varies
considerable across municipalities: over the past decade, Mexico City and municipalities in Quintana Roo
collected 0.44% and 0.48% of state GDP respectively, whereas municipalities in Campeche and Tabasco
collected just 0.02% and 0.04% of state GDP respectively. Coordinated efforts between levels of
government can lead to large increases in collection. Back of the envelop calculations suggest that
revenues from this tax in Mexico could increase by more than 1 percentage point of GDP should all
municipalities were to reach the level of property tax collection that Mexico City currently have.
Figure 2.16. Revenues from property taxes in Mexico, Latin American and OECD countries (2019)
25. One factor behind the differences in terms of property tax revenues, across municipalities in
different states and within a state, is the quality of cadastral system that does not enable a regular
actualization of the fiscal cadasters. Many municipalities have cadastral systems and fiscal cadasters
with low level of technological sophistication. This makes updates to the valuation of existing properties
in the jurisdiction and keeping up with new construction very challenging for many municipalities. Just
updating cadastral values can go a long way in improving tax collection, and modern valuation
approaches can facilitate the updating of the fiscal cadaster. For example, some states, like Baja
California, have a mandate for the municipalities to maintain an up-to-date geographic registry that
incorporates appropriate technologies and that is reviewed annually. This type of measure seeks to
avoid lags in the valuation mechanisms of local governments, ensuring that the quality of the
information used is homogeneous in all municipalities. To solve this problem, it may be necessary, on
the one hand, to establish a national cadastral registry program supported, for example, by the state,
and a scheme that stimulates municipal collection by rewarding, with additional states and federal
funds, the municipalities that make the most progress in the goals of the program.
26. Local capacity in many municipalities may undermine property tax collection. A first
requirement to improve collections is political will at the subnational level. It is also critical to provide
adequate incentives from the state and federal levels, while systematically taking steps towards
improving collection capacity at the local level. Most municipalities in Mexico, particularly those in rural
areas, do not have the capacity to effectively administer a property tax system (Unda and Moreno,
2015). They face staffing shortages and staff lack specific technical skills and/or are unfamiliar with ICT-
platform-based property tax administration. Municipalities do not have the means to administer a
59
regular training regime, which has an important role in strengthening process activities such as property
valuation, property tax administration, data analysis for decision making, IT intensive tools, and GIS
based property tax administration systems.
27. International experience offers some practical solutions, including a state (more centralized)
administration of the property tax without harming municipal autonomy for the tax. The experience
gathered from large countries and federations (Canada, Pakistan, Australia, India—early stages) is that
state level collection on behalf of municipalities may help improve collections, reduce collection costs,
and improve efficiency, effectiveness, and sophistication in terms of fiscal cadasters, all due to
economies of scale and capacity gains (see Chapter 3 for details).
Indirect taxes
28. Taxes on goods and services raised less than 6 percent of GDP in 2019 compared to 11.4
percent in LAC and 11 percent in OECD countries (Error: Reference source not found and Error:
Reference source not found). Taxes on goods and services are the second largest source of revenue,
representing more than 36 percent of total revenues in 2019, yet they remain significantly
underexploited. This is mainly due to the value-added tax, which raises a small amount of revenue
relative to other countries. Excise taxes are the next largest type.
Figure 2.17. Goods and services tax revenues are low and rely on value-added tax
Figure 2.18. Revenues from VAT as a share of GDP

Mexico vs LAC countries Mexico vs OECD countries
60
VAT
29. Low VAT revenues in Mexico are partly explained by the wide spectrum of goods that are
exempted or have reduced rates. The standard VAT rate is 16 percent, which is below the OECD
average (19.2 percent) (Error: Reference source not found)39. Furthermore, Mexico applies a reduced
rate of 0 percent to a large number of goods and services. According to the SHCP, tax expenditures from
VAT represent around 1.43 percent of GDP. The overall VAT collection would increase to 5.4 percent of
GDP should this foregone revenue collection would be added to the actual VAT collected (3.9 percent of
GDP). VAT exemptions mainly focus on benefits on the provision of education services and housing sales
(including rent and mortgage interest) which represent 0.24 percent of GDP. In addition, reduced rates
represent 1.17 percent of GDP. The main one representing 85 percent of the total is the zero rate on
food, amounting to 1.1 percent of GDP in 2019, followed by medicines (7 percent of the total).
Figure 2.19. Standard rates of VAT remain relatively low (2018)
30. Furthermore, VAT tax expenditures are inefficient in alleviating poverty, because higher-
income households capture most of the benefits. Higher income households have higher levels of
consumption and therefore pay more in value added taxes than lower income households in absolute
amounts. As a result, a higher share of total VAT tax expenditures goes to higher income households.
Moreover, some exempted items and zero-rated items are mostly consumed by the top of the income
39
As of 31 December 2020.
61
distribution, including some foods and services. For instance, households who choose to send their
children to private schools and universities are typically better-off, yet educational expenditures are
exempt. This is compounded in cases where there is a high share of informality. In Mexico, more than a
third of all spending took place in informal establishments 40 in 2020 (Error: Reference source not found),
in line with other low- and middle-income countries (Bachas et al, 2020). When informality is
considered, intended redistributive policies, such as reduced tax rates on basic goods, have a limited
impact on inequality as the benefits of zero rates and exemptions are further concentrated among
higher income households (Error: Reference source not found).
Figure 2.20. Mexico. Informal purchases as a % Figure 2.21. Mexico. Benefits from VAT zero rate
of total household expenditures, 2020 and exemptions, 2020 (share of total benefit from
reduced rates)
40%
30%
35% 34% 25%

25%
30% 29%
27% 20%
25%
25% 24% 14%
22% 15%
20% 12%
20% 19% 10%
17% 10% 9%
7% 8%
6%
15% 5%
13% 4%
5%
10%
0%
1 2 3 4 5 6 7 8 9 10
5%
Disposable income deciles
0%
1 2 3 4 5 6 7 8 9 10
Consumption deciles Benefits from zero rate Benefits from exemptions Zero rates + exemptions
Source: World Bank using ENIGH 2018. Source: World Bank microsimulation model.
31. Mexico’s Congress approved a tax legislative package in 2019 that also amended the VAT law
to improve taxation of the digital economy. The reform enables withholding and payment of (i) VAT on
digital services provided by non-resident service providers; and (ii) VAT and ISR (income taxes) on
services intermediated by digital platforms. This reform, which came into force on June 1, 2020,
introduces a simplified registration-based VAT collection mechanism for non-resident digital services
providers and requires digital intermediary platforms to exchange information and support VAT and
income tax filing obligations of the ultimate service providers in the country, inter alia through a
withholding mechanism. The reform is fully in line with the guidance provided by the OECD (2017) and
does not infringe on existing bilateral taxation treaties that aim to avoid double taxation. While tax
revenues will increase marginally by 2021 as the reform is rolled out, digital taxes are expected to grow
significantly over time in tandem with the growth of the digital economy. The “digital tax base” remains
small relative to the traditional tax base, but it is set to continue to grow rapidly as the digital economy
becomes much more prevalent in the market. Sales of digital platforms in Mexico grew on average over
12 percent annually in the five years prior to the crisis, faster than the rest of the economy. E-commerce
represented roughly 5 percent of GDP in 2018, more than doubling its size from a decade ago. Due to
the impact of the pandemic, including social distancing and the increase in use of e-platforms, rapid
growth of e-platforms is likely to continue over the medium term, perhaps at a more accelerated pace.
Taxing the digital economy would raise additional revenues and is likely to be progressive from the
distributional perspective.
40
The following places of purchase are classified as informal: open air markets, tianguis, street vendors, small grocery stores
(tiendas de abarrotes), convenience store, loncherías, fondas, torterías, and economic kitchens and dinners.
62
32. Research shows that the preferential border VAT rate generates little economic benefits
relative to the foregone tax revenues (Bacha et al., 2019; Ahmad et al., 2012). Historically, Mexico has
had a differentiated VAT rate between the border and the rest of the country. Although standard and
preferential rates suffered several modifications since the creation of the VAT in 1980, it was not until
the 2014 tax reform that rates were standardized at 16 percent throughout the country. This reform was
reversed in 2019, when the border area rate was lowered to 8 percent, keeping it at 16 percent in the
rest of the country. Further, in 2021, a border rate was extended to the municipalities of the southern
border. The goal was to stimulate economic activity and reduce cross-border shopping to the United
States where retail sales tax rates typically range from 5–7 percent. One of the purposes of this policy is
protecting the poor, as a lower VAT rate may translate into lower prices of goods, benefitting the
poorest in the income distribution. However, a difference-in-difference approach for taxed products
suggests that the estimated effect of the VAT on Mexican consumer prices is relatively low (around 20
percent). The effect is likely lower when one accounts for informal stores (Bachas et al., 2019). Since
poorer households tend to rely more on unregulated markets, the existence of the informal retail sector
can work as a market institution that shelters low-income consumers from adverse economic policy.
Furthermore, the border VAT represents a tax differential within Mexico, distorting economic activity
toward the preferential tax zone, and promoting tax evasion (Ahmad et al., 2012). Finally, this
preferential rate represents a significant loss in revenues in the form of tax expenditures (SHCP, 2020).
33. Mexico lags LAC and OECD countries in terms of revenue collection efficiency from VAT. In
Mexico the ratio of actual VAT collections to the theoretical revenues under a perfectly enforced tax
levied at the standard rate on all final consumption without any exemptions (what is known as the C-
efficiency) was 29 percent in 2019, one of the lowest among OECD countries and Latin America (that has
an average C-efficiency of 50 percent). An alternative measure, known as the VAT Revenue Ratio (VRR),
shows similar results.41 Mexico has been improving its VRR since 1996 when it was 0.21 to reach its
highest level in 2018 at 0.34, yet below the OECD average of 0.56.
34. The low level of VAT C-efficiency in Mexico can be explained by a narrow base (owing to tax
expenditures and informality) and by high non-compliance. In general, there are three main factors
that can explain a low level of tax collection in a country: (i) low rates; (ii) a narrow base; and/or (iii) non-
compliance (IMF, 2018). At 16 percent, the standard VAT rate in Mexico is not significantly lower than
the averages of OECD countries and Latin America. The base of the VAT in Mexico is reduced mainly for
two reasons: (a) tax expenditures on final products (such as exemptions of education and housing as
well as zero ratings of domestic transactions) represented 1.43 percent of GDP in 2019; and (b)
exclusions from the object of the tax (owing to informality, for example). Also, VAT compliance gap is
relatively high at 45.8 percent of potential revenue in 2016 (or 2.41 percent of GDP) as estimated by IMF
(2018). Non-compliance and tax expenditures are related: exemptions and zero rating of domestic
transactions, in addition to eroding the tax base, make control difficult and thus facilitate evasion. Any
reform to the VAT system will not generate a significant increase in tax revenues due to the prevalence
of informal trade in the country.
Figure 2.22. Mexico lags behind in terms of revenue collection efficiency from VAT (VAT Revenue
ratio)
A. Mexico vs LAC countries (2019) B. Mexico vs OECD countries (2019)
41
The VRR is a measure of the revenue raising performance of a VAT system. A ratio of 1 would reflect a VAT system that
applies a single VAT rate to a comprehensive base of all expenditure on goods and services consumed in an economy - with
perfect enforcement of the tax.
63
Simulations – Fiscal and distributive impacts

35. Using the fiscal microsimulation model, we assess the fiscal and distributional impacts of
alternative options for reform to the VAT. The model captures the embedded VAT in exempt items,
thus accounting for cascading. In addition, the model uses information on the place of purchase to
account for informality and calculates the economic incidence of embedded VAT since items sold in the
informal market contain inputs that were taxed earlier in the production process. Details of the
modeling approach are described in Appendix 1.
36. The simulations include alternative scenarios aimed at broadening the tax base (Table 2.4).
These scenarios show the impacts of changes in VAT alone, as well as when they are combined with new
social spending aimed at protecting the poor while at the same time generating some fiscal space. The
range of scenarios include more aggressive to less aggressive alternatives in terms of eliminating or
reducing tax expenditures. Each scenario aims to find a combination of policies that would be more
efficient in reducing poverty and inequality while at the same time generating additional revenue.
 Eliminating exemptions and zero-rated items: One scenario assumes that all zero-rated and
exempt items were moved to the standard rate of 16 percent. VAT collection would increase by
58.8 percent (or 2.31pp. of GDP). This measure would disproportionately affect the bottom of
the distribution. The poverty headcount would increase by 1.5pp. relative to the 2019 baseline,
and inequality would increase by 0.3 Gini points. If part of the increase in VAT tax collection
were spent on pro-poor programs, eliminating exemptions and zero-rated items could reduce
poverty while generating fiscal space. If households were fully compensated for the elimination
of exemptions and zero-rated items through the targeted scholarship program Benito Juárez,
this would reduce poverty by 0.3 pp while reducing inequality by 0.5 Gini points, at the cost of
an additional spending of 0.24 percent of GDP. This still would leave 2.07 percent of GDP in fiscal
savings after elimination of exemptions and zero-rated items. An alternative option would
assume eliminating exemptions and zero-rated items for everything other than food and public
transport. Food represents about half of the total consumption for the poorest decile, compared
to 38 percent for the top decile. In addition, public transport represents 6 percent of the total
consumption for the bottom decile, doubling the share of the richest decile. If all zero-rated and
exempt items were moved to the standard rate of 16 percent except for food items consumed
at home and public transport, this would generate a 29 percent increase in VAT collection (or
1.2pp. of GDP. This measure would lead to a 0.6pp. increase in poverty and an increase in
inequality.
64
 Eliminating exemptions and reduced rates except for a basic food basket and public
transportation: Several different scenarios were simulated with basic food baskets largely
consumed by the poor. The scenarios assume preferential treatment for certain foods and urban
public transport, while moving all other items to the standard rate. Moving zero-rated and
exempt items to the standard rate of 16 percent, would increase VAT collection between 28.0
and 41.4 percent depending on the scenario. The additional VAT collection would amount to an
increase of between 1.4 pp or 1.6 pp of GDP. However, it has significant distributional impacts.
Despite keeping a zero rate on the 35 most consumed items by the poor, moving all other food
items to a standard rate would still have an impact at the bottom of the distribution. Value
added tax would become regressive and inequality would increase. Poverty headcount rate
would increase relative to the 2019 baseline on each of the scenarios considered, respectively.
Given these distributional impacts, eliminating some exemptions and zero-rates for food and
public transport would require additional compensation for the poor to ensure poverty
reduction and net fiscal savings. Any reduction on exemptions and reduced rate of the most
consumed items among the poor would increase poverty unless they are adequately
compensated.
 Eliminating exemptions to educational services: If exemptions to educational services were
eliminated, VAT revenue would increase by 13 percent (or 0.5pp. of GDP). VAT would continue
to be progressive and there would be no significant impact on inequality. The US$5.5 poverty
headcount rate would increase by 0.3pp., as this change would mostly affect better-off
households who tend to be the ones using private schools and other private services.
 Eliminating exemptions to entertainment services: Currently there are exemptions on
entertainment services and social services such as clubs and associations. If these exemptions
were eliminated, VAT revenue would increase by one percent (or 0.04pp. of GDP). VAT would
continue to be progressive and there would be no significant impact on inequality. The US$5.5
poverty headcount rate would increase by 0.1 pp., as this change would mostly affect better-off
households who tend to use these services.
 Increasing the VAT rate and eliminating reduced border rate: An alternative set of reforms
would aim to equalize the standard rate across the country and potentially increase the
standard rate. Since 2019, a preferential rate of 8 percent was applied in municipalities at the
northern border, a reform that reversed the 2014 effort to equalize VAT rates by raising the
preferential northern border rates from 11 to 16 percent. Reversing this reform, equalizing the
standard rate at 16 percent across the country would improve both horizontal and vertical
equity, as households living in border towns have higher incomes than those living in the South.
Equalizing the standard rate across the country would increase the progressivity of VAT, while
having an insignificant impact on inequality and a very small impact on poverty (increase of
0.1pp.). This change would nevertheless increase VAT revenue by 5.6 percent (or 0.2pp. of GDP).
Increasing and equalizing the standard rate to 19 percent would increase tax collection by
0.87pp. of GDP but poverty would also increase by 0.39pp. Inequality, on the other hand, would
slightly decrease by 0.01 Gini points in all cases. A larger share of the gains in VAT collection
comes from the equalization of the border rate. Simulating the same increases of the standard
rate without eliminating the reduced border rate, would result in a much lower increase in tax
collection. If the poor were compensated with some of the additional VAT revenue, the reform
would make VAT more progressive, would reduce poverty relative to the baseline, while
achieving overall fiscal savings.
37.
65
Table 2.4. Alternative VAT Reform Scenarios
Marginal Poverty
Inequality Change in: Change in revenue:
Kakwani contribution to: headcount
index (US$5.5 %∆ of VAT %∆ of
Poverty Inequality (Gini) Poverty Inequality
PPP2011) revenue GDP
Baseline 2019 0.009 -0.021 -0.002 25.1 44.2
VAT policy scenarios

Eliminate exemptions and zero rates
1 Eliminating exemptions and zero-rated items -0.039 -0.037 -0.006 26.6 44.6 1.5 0.3 58.8 2.3
2 Eliminating exemptions and zero-rated items while compensating the poor -0.031 -0.037 -0.006 24.8 43.7 -0.3 -0.5 38.3 2.1
3 Eliminate all exemptions and zero-rates except for food items consumed at
-0.001 -0.027 -0.003 25.7 44.3 0.6 0.1 29.3 1.2
home and urban public transport
Eliminate exemptions and zero rates for food outside the basic basket
4 Except for a list of 35 non-processed food items consumed by the bottom
-0.015 -0.032 -0.004 26.0 44.4 0.9 0.2 41.4 1.6
30% and urban public transport
5 Option 4 and compensate the poor -0.010 -0.032 -0.004 24.9 43.8 -0.2 -0.4 28.0 1.5
6 Except for a list of 35 processed+non-processed food items consumed by
-0.006 -0.030 -0.003 25.8 44.3 0.8 0.1 37.5 1.5
the bottom 30% and urban public transport
7 Option 6 + maize products - 39 products consumed by the bottom 3 deciles -0.005 -0.030 -0.003 25.8 44.3 0.7 0.1 37.3 1.5
8 Option 4 + 6 + 7 – total of 54 products -0.003 -0.029 -0.003 25.8 44.3 0.7 0.1 35.7 1.4
Eliminate exemptions and zero rates for products not consumed by the poor
9 Eliminate exemptions to educational services 0.004 -0.024 -0.002 25.4 44.3 0.3 0.0 13.1 0.5
10 Eliminate exemptions to entertainment services 0.011 -0.021 -0.001 25.2 44.3 0.1 0.0 1.0 0.0
Increase border and standard rates
11 Eliminate border rate, equalize rate at 16% across the country 0.013 -0.022 -0.001 25.2 44.2 0.1 0.0 5.6 0.2
12 Eliminate border rate and increase standard rate to 17% 0.013 -0.024 -0.002 25.3 44.2 0.2 0.0 11.3 0.4
15 Option 13 and compensate the poor 0.015 -0.026 -0.002 25.2 44.0 0.1 -0.2 18.2 0.8
Note: Kakwani coefficients, and marginal contributions to poverty and inequality are measured with respect to disposable income in baseline and reform scenarios. Poverty
and Inequality are measured for consumable income, the post-VAT welfare aggregate. Revenue impact is measured as a share of baseline VAT revenue observed in the
household survey which amounts to 27 percent of total VAT revenue collections according to administrative records, largely due to the fact that the household survey does
not have information on top incomes (see Appendix 1). Estimates on revenue collections in terms of GDP are scaled and assume that the same proportionality holds under
the reform scenario.
38.
Excise taxes
39. While VAT taxes could yield the biggest gains, there is also room to improve excise tax
collections in Mexico. Mexico’s revenue from excise taxes is low relative to other LAC countries.
Collections are also 1 percentage point or more below the average for OECD countries. Most excise
revenue come from fuel taxes, while “bads” excise taxes make up less than half of collections.
Figure 2.23. Revenue from Excise taxes as Table 2.5. Incidence of Excise Taxes and Contributions to
a share of GDP Poverty and Inequality, 2020
Marginal
contribution Marginal
Kakwani
(Poverty contribution
head count) (Gini Index)
Excise taxes -0.006 -0.001 -0.015
Alcohol 0.000 0.000 0.115
Tobacco 0.000 0.000 -0.029
Fuel -0.004 0.000 0.022
Energy drinks 0.000 0.000 -0.005
Sugar beverages -0.001 0.000 -0.292
Industrial fuel 0.000 0.000 -0.238
Junk food 0.000 0.000 -0.153
Lottery 0.000 0.000 0.187
Telecommunications -0.001 0.000 -0.105
Source: OECD Revenue Statistics. Source: World Bank Fiscal Microsimulation Model based on ENIGH
2020.
40. Excise taxes are slightly progressive overall, mostly on account of automobile fuel excise
taxes. In contrast to other countries, excise taxes are mostly progressive and have little impact on
inequality in Mexico (Error: Reference source not found). This is largely driven by progressivity in fuel
taxes, which accounted for 56 percent of all excise tax revenues in 2021, according to administrative
data from the Ministry of Finance. Whereas the bottom 40 percent of the distribution spends between
0.6 and 0.7 percent of their disposable income on fuel excises, the richest 20 percent spends about 1
percent of their disposable income on fuel taxes. Moreover, 37 percent of the revenue collected from
fuel excise taxes comes from the top decile of the income distribution and only 1 percent from the
bottom decile. This progressivity exists even when both the direct and indirect effects of taxes on fuel
are taken into account.42 While taxes on alcohol, fuel, and lottery are progressive, taxes on tobacco,
energy drinks, sugar beverages, industrial fuel, junk food, and telecommunications are regressive in the
short run. Nevertheless, none of these taxes have a significant impact on inequality because they
represent less than one percent of disposable income, even for top earners. Moreover, as shown below,
once the lifetime benefits of reducing consumption on unhealthy products is considered, many of these
taxes become progressive in the long run.
41. As in most countries in Latin America and the OECD, Mexico is increasingly using excise taxes
to influence consumer behavior to achieve health outcomes as well as raise revenue. Empirical
research based on OECD countries suggests that there is no growth penalty to increases in excise taxes
(Akgun, Cournède and Fournier, 2017). This may result from the relatively inelastic tax base, which
makes the tax efficient. The distributional impact of increasing excise tax rates will depend on
consumption patterns but these effects are mitigated by the long term distributional impacts of health
costs, as described below.
Figure 2.24. Revenues from excise taxes in Mexico are higher than LAC average but lower than OECD
42
Given the fact that fuel is a critical intermediate input, we use a cost-push model and use an input-output matrix to estimate
the impact of fuel taxes on the prices of all goods. See methodological details in Appendix 1.
67
Revenues from excise taxes as a share of GDP (2019)
Simulations
42. While excise taxes are generally progressive, taxes on some “bads” make a larger portion of
disposable incomes of the poor. Excise taxes on tobacco have a larger impact on lower income
households, representing 0.14 percent of the disposable income of households in the bottom decile
compared to 0.06 households in the top decile (Error: Reference source not found). Similarly, excise
taxes on sugar-sweetened beverages (SSB) and junk food represent 0.04 and 0.02 percent of the
disposable income of households at the top decile but this share is significantly larger for the households
at the bottom decile (0.46 and 0.14 percent, respectively).
Figure 2.25. Relative Incidence of Excise Taxes on “Bads”

(share of disposable income by disposable income deciles)
0.60%
Percent of disposable income
0.50%
0.40%
0.30%
0.20%
0.10%
0.00%
Alcohol Tobacco Energy drinks Sugary drinks Junk food Lottery
Source: World Bank estimates based on ENIGH 2020 and microsimulation model.
43. Nevertheless, this point-in-time analysis does not account for potential lifetime benefits of
reducing consumption of unhealthy products. Despite the regressive nature of some excise taxes on
“bads”, these taxes are powerful tools to reduce the consumption of unhealthy products and thus can
bring important long-term benefits, such as a reduction in diseases and premature deaths. In addition,
68
they are associated with economic gains arising from lower medical expenses, higher worker
productivity, and longer working lives. Those benefits can offset the short-term negative effects.
Tobacco taxes could be further increased

44. Tobacco consumption represents a major public health concern in Mexico. Smoking prevalence
among individuals aged 12-65 was 17.6 percent in 2017, remaining almost unchanged since 2011
(ENCODAT, 2017). An extensive body of research has established that tobacco use, and in particular
cigarette smoking, causes numerous serious illnesses. In 2019, tobacco-related diseases caused 57,518
premature deaths in Mexico, representing about 8 percent of all deaths in the country (Global Burden of
Disease, 2020; INEGI, 2020). Chronic obstructive pulmonary disease and ischemic heart disease
accounted for 47 percent of those lives lost. Beyond the adverse health effects, tobacco use also entails
large economic losses, both in terms of medical expenses and lost productivity. 43 Smoking-attributable
healthcare expenditures were estimated at MX$81 billion in 2015, or 0.44 percent of GDP (Centro de
Estudios de las Finanzas Públicas, 2018).
45. Increasing tobacco prices through higher taxes is one of the most cost-effective interventions
to reduce tobacco consumption and its consequent burden of disease (WHO, 2019).44 The retail price
of cigarettes is a crucial determinant of cigarette consumption. By increasing taxes on tobacco products,
retail prices will increase, making them less affordable and thus discourage their consumption. The WHO
recommends a minimum 75 percent tax share of the retail price of tobacco (WHO, 2015). 45 In LAC, the
average taxation is 49 percent for a pack of 20 cigarettes. Although tobacco taxation in Mexico (67
percent) is above the regional average, it still lags behind developed countries and several regional
peers, such as Brazil (83 percent), Chile (82 percent), and Colombia (78 percent), and remains below the
WHO’s recommended level of 75 percent. Tax increases on tobacco products not only reduce tobacco
use and improve health, but they also generate more government revenues. Moreover, taxation has
proved to be a cost-effective means of reducing smoking (WHO, 2017).
46. Tobacco taxes in Mexico are progressive, partly explained by a large concentration of smokers
in the top deciles of the income distribution. Conditional on smoking, low-income households typically
allocate a significantly larger share of their budget to tobacco spending. According to the ENIGH (2018),
tobacco consumption represented on average 10.4 percent of the disposable income of households at
the first decile of the income distribution, whereas this share was 1.9 percent for households at the top
decile. Yet, households with higher income in Mexico consume more cigarettes, that is, the richest decile
consumes up to four times more than the poorest (Huesca et al, 2019). Since cigarette consumption
varies according to income level, an increase in the specific component of the tax will fall mainly on
households with higher incomes. Regardless of the increase in the specific component, higher income
groups continue to pay more than half of total tax income: about 52 percent of the collection of the
special tobacco tax comes from deciles 8, 9 and 10 (with the highest income) while deciles 1, 2 and 3
contribute 11 percent.
47. Although Mexico has recently updated its tobacco tax policy, tobacco taxes could be further
increased. For almost a decade, tobacco tax rates remained constant at relatively low levels. Tobacco
products were subject to a general VAT rate of 16 percent and an ad valorem rate of 160 percent
43
In 2019, 17,537 tobacco-related death events occurred before the age of 65, resulting in a national aggregate of 146,448
years of productive life lost among the working age population.
44
For greater effectiveness, increases in tobacco taxes should be combined with other tobacco control measures, such as bans
on tobacco industry marketing activities, prominent pictorial health warning labels, smoke-free policies, and population-wide
tobacco cessation programs to help people stop smoking.
45
The total tax burden on cigarettes is calculated as the sum of all taxes on cigarettes – including general sales taxes, such as a
value-added tax – expressed as a percentage of the retail price.
69
combined with a specific tax of MX$0.35 per cigarette. With the amendments to the excise law on
tobacco (Ley del Impuesto Especial sobre Producción y Servicios, IEPS), Mexico started to adjust its rates
by inflation (annual adjustments by inflation to the specific IEPS). This led to an increase on January 2020
in the specific tobacco tax to MX$0.498 per cigarette to account for inflation over the period 2011-19
(SHCP, 2019). The WHO recommends a greater reliance on specific excise taxes, as an increase in these
taxes would reduce the gap in prices between premium and low-priced brands and thus limit
opportunities for users to switch down in response to tax increases.
48. In the short-term, an increase in tobacco taxes would fall disproportionately on lower income
families. Holding tobacco consumption constant, the results from the microsimulation model show that
if Mexico raised specific tobacco taxes further, poverty and inequality would tend to increase ( Error:
Reference source not found). However, these changes are of small magnitude. Annual excise revenues
from cigarettes could increase between 3 and 34 percent compared to excise collection in 2018,
depending on the scenario. This additional revenue could be used for tobacco control programs, as well
as other health and social initiatives.
Table 2.6. Short-term effects of increasing tobacco taxes
Source: World Bank estimates based on ENIGH (2018) and microsimulation model.
Note: Results from the microsimulation model capture 11% of the total tobacco tax collection reported in the
national accounts. Part of this differential is explained by the underreporting of households’ consumption of sin
goods and the limited data coverage of the upper tail of the income distribution.
49. However, the short-term analysis does not consider behavioral responses to price changes nor
the potential lifetime benefits of reducing consumption of unhealthy products, which could impact
households’ welfare in the long-term. While demand for tobacco products is not as elastic as demand
for other consumer products, research has shown that increases in the price of tobacco products are
followed by falls in their consumption (Jimenez-Ruiz et al., 2008; Macias Sanchez et al., 2020).
Therefore, the change in tobacco spending resulting from an increase in tobacco taxes will depend on (i)
the size of the price increase, and (ii) the price elasticity of demand. When facing a similar price shock,
lower-income households tend to cut consumption at relatively higher rates than their wealthier peers. 46
income
disposable
46
Decile-specific elasticities are estimated based on a log-log OLS model, following the equation:
of
Per cent
. Where Q i , dis the quantity of tobacco purchased by household i in decile d,

0 .6 0 %
0 .5 0 %
0 .4 0 %
0 .3 0 %
0 .2 0 %
0 .1 0 %
0 .0 0 %
A l c ohol T oba c c o E ne rg y dri nks S ug a ry dri nks J unk f ood L ott e ry
P oore s t 2 3 4 5 6 7 8 9 R i c he s t
Pi , d is the price of tobacco, and X i , d is a vector of sociodemographic covariates, including region, urban vs. rural location,
household size, sex of the household head, educational attainment of the household head, and an indicator for children in the
household, among others. In the absence of price data, prices are proxied by unitary values of tobacco, inferred from the
household survey data on quantity and expenditures. Our elasticity estimates are monotonically decreasing with income and
range between -0.97 (for the bottom decile) and -0.84 (for the top decile). However, the distributional pattern in Mexico is
flatter and elasticities are larger than those estimated by previous studies (Jimenez-Ruiz et al., 2008; Macias Sanchez et al.,
2020).
70
In the long-term, lower tobacco consumption would be accompanied by a reduction in smoking-
attributable diseases and premature deaths, translating into lower medical expenses, higher worker
productivity, and longer working lives. Those benefits can offset the short-term negative effects of
higher prices on household budgets, potentially resulting in long-term net welfare gains (see Fuchs and
Meneses (2017) for the extended cost-benefit analysis to estimate the long-term distributional impact of
the proposed changes). The long-term net effect of increasing tobacco taxation is calculated as the sum
of three effects: the change in household tobacco spending, the change in smoking-related medical
expenses as smoking becomes less affordable, and the change in household earnings arising from a
reduction in the years of working life lost (YLL). 47 The net effect is added and interpreted as a percentage
of household income.
50. In the long-term, the net effects of increasing tobacco taxes are both progressive and welfare-
improving. Tobacco taxes might be progressive once the greater price sensitivity of the poor and the
internalities associated with tobacco use are taken into account. 48 The first-order estimations on Error:
Reference source not found, 49 which fail to incorporate consumer price responses, show that in the
short-term income changes are negative and regressive across all tax scenarios. However, by
incorporating decile-specific elasticities and accounting for the reductions in medical expenses and YLL,
the net effects of increasing tobacco excise taxes are generally positive and larger for lower-income
households (Error: Reference source not found).50 Note that these estimates incorporate the reduction
in tobacco consumption as a response to the price increase, as well as the reduction in medical expenses
and years of productive life lost. For example, in the long-term, an increase of the specific component to
MX$0.54 is estimated to increase income by 0.4 percent for the average low-income household,
whereas this increase is estimated to be 0.1 percent for households at the top decile of the income
distribution. This net effect is mostly driven by a reduction in medical expenditures.
Figure 2.26. Distributional effect of an increase in Figure 2.27. Net distributional effect of an increase in
tobacco taxes without behavioral changes tobacco taxes, including long-term behavioral responses
47
As data on medical expenses and YLL are reported at the country-level, we first distribute them across income deciles
according to the share of households reporting positive tobacco expenditures. Changes on medical expenses and YLL due to
increasing tobacco taxation are estimated based on the size of the price shock and estimated price elasticities (Fuchs and
Meneses 2017).
48
Not only poor households are more responsive to tobacco price changes, but they are also exposed to larger price shocks for
a given tax increase. This is related to the fact that they tend to consume cheaper cigarette brands. For example, based on
brand preferences from ENCODAT (2017) and price data reported in the National CPI (2018), an increase of the specific IEPS to
MX$0.54 would raise cigarette retail prices by 12 percent for consumers of the cheapest brands, whereas prices would increase
by 9% for consumers of premium brands.
49
This figure shows the welfare effect of increasing tobacco taxes by income decile. It holds tobacco consumption constant and
it evaluates the effect of a price increase.
50
Qualitatively similar results are obtained when restricting the analysis to the fraction of medical expenses that is paid out-of-
pocket (42 percent in 2018, according to the WDI).
71
Source: World Bank estimates based on ENIGH 2016 and 2018, GBD (2017), INPC (2018), ENCODAT (2016-17), WDI, Centro de
Estudios de las Finanzas Publicas (2018), and the income aggregate methodology by CONEVAL. Deciles are constructed based on
household total income per adult equivalent.
Environmental taxes
51. Mexico imposes several environmentally related taxes (defined as taxes levied on bases
deemed to be of environmental relevance) and was one of the first major countries to implement a
carbon tax. Since 1980 the country applies a specific excise tax on transport fuels (IEPS or Impuesto
especial sobre producción y servicios) that applies to gasoline, diesel and their biofuel equivalents. 51 In
2005, Mexico introduced a special consumption tax (ISAN) that is levied on the sale of new automobiles.
The basis of the ISAN is the sale price, including common or luxury optional equipment, charged to the
consumer by the manufacturer, assembler, distributor or authorized dealer in the vehicle sector.
Although the IEPS and ISAN taxes were not introduced with environmental objectives (but rather
revenue collection objectives), they have a strong environmental impact. Finally, with the 2013 tax
reform, Mexico became one of the first large economies to introduce a carbon tax that applies to a
much broader base (gasoline, diesel, fuel oil, kerosene, LPG, and coal for use by households, industry,
and electricity generation). This carbon tax was designed to send a price signal towards a lower-carbon
economy and obtain revenues from externalities. Supported by these instruments, Mexico was able to
achieve a reduction in its carbon intensity (measured as carbon footprint per unit of GDP) by the end of
2017 (Muñoz-Piña et al., 2020).
52. Despite improvements reducing the carbon intensity of the economy, environmental
effectiveness of Mexico’s tax system remains weak and collect few revenues due to narrow
environmental tax bases and low effective rates. Environmentally related tax revenue has been
increasing but remains relatively low in Mexico (Error: Reference source not found). According to the
Ministry of Finance (Secretaría de Hacienda y Crédito Público) in 2019 Mexico raised a little over 1
percent of GDP from the federal and local IEPS and the carbon tax, below the average for OECD
countries (over 2 percent of GDP). The registration and circulation tax raises on average 0.6 percent of
GDP in European countries, over ten times more than in Mexico (0.05 percent of GDP). The carbon tax
raises less than 0.1 percent of GDP in revenues (SHCP, 2020). While the introduction of the carbon tax
was a clear step forward, the carbon tax rates currently applied are relatively low ( Error: Reference
source not found) and they do not reflect the carbon content of the fuels (Error: Reference source not
found). For instance, despite its high estimated social cost, the tax rates on coal are set at a much lower
level than for other fuels. Natural gas, which accounts for around 45 percent of Mexican energy use, is
51
In the 1990s Mexico initiated a policy to keep gasoline and diesel domestic prices at stable levels, smoothing-out temporary
upward surges in international oil prices shocks through immediate reductions in these excise taxes. By doing this, the excise
taxes became negative and transformed, de facto, into a heavy subsidy. Starting in 2010, Mexico initiated its efforts to phase-
out fuel subsidy.
72
subject to a zero tax rate. As a result, the share of carbon emissions priced at or above EUR 30 per tCO2
remains very low outside of road transport. If the purpose of the carbon tax is to send a meaningful
signal on the social cost of carbon emissions, and thus represent a credible component of the policy
strategy for the mitigation of greenhouse gas emissions in Mexico (Mexico’s NDCs to the Paris
Agreement has a goal to achieve a 22 percent reduction in GHG emissions), corresponding tax rates
should be increased gradually.
53. In the long run, an overhaul of Mexico’s environmental taxes could facilitate a reduction of
formal-sector taxes on labor and capital. This could be done in such a way that overall tax revenues are
unchanged, delivering environmental improvements (such as lower emissions) while having a positive
effect on the economy (such as higher formal employment). Evidence for Mexico (Iacovone et al, 2019)
suggests that raising fuel prices can—in certain circumstances—foster firm-level efficiency gains by
encouraging energy efficiency and incentivizing investment in modern equipment. These gains in
efficiency and productivity can, in turn, make firms more competitive internationally.
Figure 2.28. Tax revenue from Environmental Taxes (% of GDP, 2019)
Source: SHCP, Eurostat and OECD.
Table 2.7. Implicit carbon prices per tCO2 by fuel in Mexico

Fuel Tax Implicit price per tone
Value Unit MXN per tin CO2 USD per ton CO2
Propane 7.48 MXN cents per liter 50.3 2.5

Butane 9.68 MXN cents per liter 53.2 2.7
Gasoline 13.12 MXN cents per liter 57.2 2.9

Diesel 15.92 MXN cents per liter 58.7 2.9
Fuel Oil 16.99 MXN cents per liter 57.9 2.9

Petroleum Coke 19.72 MXN per ton 7.3 0.4
Coal Coke 46.23 MXN per ton 16.9 0.9

Coal 34.81 MXN per ton 13.8 0.7
Other fossil fuels 50.32 MXN per ton CO2 50.3 2.5
Natural Gas 0 MXN cents per liter
Figure 2.29. Average effective energy tax rates by country
73
Source: OECD
54. Introducing a uniform economy-wise carbon tax would help reduce inefficiencies, meet
climate mitigation commitments, and increase revenues. Preliminary estimations using the Carbon
Pricing Assessment Tool (CPAT) show that gradually increasing the carbon tax rate to $50 per ton CO2 by
2026, starting with $25 per ton in 2021, could contribute to achieving about half of Mexico’s climate
mitigation commitment by 2030 and could increase annual revenues by approximately 1.1 percent GDP
by 2030, with the largest revenues coming from natural gas (35 percent of total revenue increase) and
gasoline (22 percent) (Error: Reference source not found). The model suggests that prices for liquid fuels
would rise moderately, e.g., US$0.06 cents per liter of gasoline (15 percent increase), US$0.07 per liter
of diesel (19 percent increase). Coal prices in turn would rise substantially (75 percent increase), leading
to a US$0.23/kwh rise in electricity prices (28 percent increase), but this could be limited by fostering
changes in the energy production matrix -i.e., using part of the revenues to subsidize renewables. For
consumer products, price rises would mostly affect energy-intensive sectors like cement, with lower
impacts on e.g., food and services which households consume more. A US$25 per ton tax on carbon is
estimated to be roughly distributionally neutral (Kakwani of -0.002) as it would represent less than 0.4
percent of disposable income across deciles, with insignificant impacts on poverty and inequality, even if
indirect effects (through higher prices of other products) are taken into account (Error: Reference source
not found). A further increase to US$50 per ton would also have an insignificant impact on inequality,
and only a 0.3 percentage point increase in poverty (Error: Reference source not found).
Figure 2.30. Annual revenues with an economy-wide carbon tax
(US$25 per ton carbon tax in 2021, rising to US$50 by 2026)
A. Fiscal revenues above baseline (US$) B. Fiscal revenues by source
74
Figure 2.31. Incidence of US$25 per ton carbon Figure 2.32. Incidence of US$50 per ton carbon tax
tax (Carbon tax as a share of disposable income)
(Carbon tax as a share of disposable income)
0.9% 0.9%
0.79%
0.8% 0.8%
0.73% 0.73%
0.7% 0.7% 0.64%
0.60%
0.6% 0.6% 0.54%
0.52%
0.5% 0.5% 0.44%
0.42%
0.39% 0.39% 0.38%
0.4% 0.34% 0.4%
0.32%
0.28% 0.29% 0.28%
0.3% 0.24% 0.3%
0.20%
0.2% 0.15% 0.2%
0.1% 0.1%
0.0% 0.0%
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
Disposable income deciles Disposable income deciles
Direct effect Indirect effect Total effect Direct effect Indirect effect Total effect
Source: Microsimulation Model based on ENIGH 2018.
Align the vehicles tax with the emissions

55. Mexico has a lot to benefit from changing its environmental strategy on road transportation
by aligning vehicle taxes with the emissions they generate. Mexico has more than 50 million passenger
vehicles that cause negative externalities such as massive air pollution and congestion, particularly in the
most densely populated areas. Just in Mexico City, road transport causes 46 percent of CO2, 17 percent
of PM2.5, 85 percent of CO, 82 percent of NOx, and 27 percent of COV (all relative to total emissions in
Mexico City). To reduce this externality, Mexico’s current environmental strategy on road transport
consists of command-and-control, setting fuel economy standards, requiring emission tests, and issuing
restrictions like Hoy No Circula, and providing support for the public transport system. But this strategy
can be further reinforced. For instance, the vehicle purchasing tax (ISAN) depends on vehicle price and
could be more clearly related to vehicle emissions. Restrictions like Hoy No Circula tend to be regressive,
as rich households can purchase a 2nd vehicle and still drive those days.
56. Adding a vehicle tax based on emissions or changing the relative price of clean vehicles could
keep the cost of average vehicles and revenues while reducing emissions. This could be achieved by a
combination of (i) a new vehicle purchasing feebate dependent on CO2 emission rate linear tax (function
75
of emission rate in gCO2/km); and (ii) a lump-sum subsidy for efficient cars. This would make the
externality cost more salient and reflect car’s fuel efficiency in their prices. The measure would be
progressive, as tends to reduce the cost of small (emission-efficient) vehicles often bought by middle
class, while making large, heavy SUVs expensive.
Replacing current tax exemptions with Output-Based Rebates
57. Phasing out some Mexico’s many energy-related tax expenditures could save revenue.
Although fossil fuels used in sectors such as Industry, Agriculture and Fisheries are taxed, these sectors
have access to a refund for the excise tax on diesel they use. Tax expenditures from this are important,
representing more than 0.5 percent of GDP (SHCP, 2020), and phasing them out would further incentive
lower fossil consumption and increase revenues. As an alternative, a second-best approach is to replace
them with Output-Based Rebates (OBR). Under OBR, protected sectors do pay the carbon tax, but
revenues are rebated back to them according to production in physical units of the output (i.e., tons of
aluminum, tons of steel). Under the current system with tax rebates, all firms in the protected sector
have interest to prevent carbon taxation. With OBR, the less polluting firms would have even an interest
to support increases in carbon taxes as it would increase their competitiveness against more polluting
firms. With OBR, countries can combine strong mitigation incentives with industrial policy strategies
that use comparative advantages in carbon-intensive industries.
The new Simplified Regime for Small and Medium Enterprises52
58. Mexico recently introduced a new simplified regime for small and medium taxpayers
(Regimen Simplificado de Confianza, RESICO) to increase formalization. 53 The reform aims to reduce
the time and number of procedures required to pay taxes in Mexico with the objective of increasing tax
compliance and facilitate formalization. The reform was aimed at taxpayers with business and
professional income, seeking to minimize the administrative burden of tax compliance, principally using
electronic billing to automate procedures. In addition, the reform is aimed at individuals with
entrepreneurial activities paying personal income taxes (PIT) and at small and medium firms, paying
taxes under the corporate income tax (CIT) regime. The introduction of the new regime replaces a
previous small taxpayer regime for individuals, the Fiscal Incorporation Regime (Régimen de
Incorporación Fiscal, RIF) which granted deductions to individual taxpayers with new entrepreneurial
activities, reducing their liabilities on personal income taxes, social security contributions, valued added
taxes and excise taxes.
59. The reform aimed to reduce the administrative burden of taxation of small taxpayers,
extending the previous special regime aimed at individuals to also include firms. Annex [xx] presents
the main differences between the new and old regimes. In contrast to the RIF, RESICO is aimed at both
individuals with entrepreneurial activity, as well as legal entities. For individual taxpayers the RIF aimed
at those with entrepreneurial activity but excluded professional self-employed individuals. In contrast,
RESICO includes those professionals. The RIF taxed individuals based on their revenue net of allowable
expenditures, while the tax base for RESICO are gross revenues, which should be easier to keep track of
and declare compared to having to declare all expenses. The change in the tax base from net to gross
has also meant there is an important change in the threshold up to which individuals and firms can
qualify into the simplified regime. Previously, individuals with annual net incomes below MX2mn were
eligible for the RIF, while firms or legal entities were not eligible. Now, individuals with an annual gross
52
Based on Inchauste and Baquero (2022): The Distributional Impact of the Introduction of the Simplified Regime for Small and
Medium Enterprises.
53
Loayza (2018) argues for a well-conceived formalization strategy aimed at reducing the costs and increasing the benefits of
formalization that would consist of four pillars: (i) increasing labor productivity; (ii) making labor markets more flexible; (iii)
making the regulatory framework more efficient; and (iv) reforming social protection. Simplification of the tax regime may
foster formalization, but it should be complemented with other reforms.
76
income of up to MXN3.5mn (around US$170,000) and legal entities with an income of up to MXN35mn
on annual gross income (around US$1.7 million) are eligible.
60. Tax rates for individuals have changed, but rates for firms remain the same. Tax rates under
the RIF were the same marginal tax rates as in the general personal income tax regime, but a deduction
was applied based on the age of the firm, starting from a 100 percent deduction in the first year, 90
percent in the second year, and further 10 percentage points declines each year. In addition, RIF
taxpayers with entrepreneurial activities were given deductions on their VAT and excise tax payments.
Legal entities were not eligible to participate in the RIF and were subject to the standard 30 percent
corporate income tax. With the reform, RESICO tax rates for individuals are much lower than in the
standard PIT regime. These rates have been calibrated to be revenue neutral since the taxable base
changed from net to gross revenue. However, taxpayers with low profit margins who previously
contributed on their net incomes under the RIF could face higher tax liabilities under RESICO. The tax
rate for legal entities is still 30 percent, however they are now eligible to contribute under RESICO,
which grants them an accelerated depreciation schedule to support investment. Finally, the change
from accrual to cash basis for legal entities will likely reduce the liquidity constraints that many firms
faced under the previous regime.
61. There are several reasons why the reform could change the incidence of personal income
taxes in Mexico. First, the reform includes a broader set of taxpayers compared to the RIF, including
higher income individuals, as the eligibility threshold is higher than the RIF, even if it is now applied to
gross income rather than net income. RESICO also includes self-employed individuals delivering
professional services, who were previously excluded from the RIF. Second, taxpayers with new
entrepreneurial activities with low profit margins that previously benefited from relatively large
deductions under the RIF could face higher liabilities under RESICO. Third, the applicable tax rates
applied to gross income are much lower in the new regime compared to the RIF. RESICO tax rates were
calibrated to be revenue neutral, however at the individual level there may be winners and losers.
Finally, existing taxpayers previously filing under RIF were grandfathered, giving them a choice through
end-January 2022 to stay under the RIF regime (the default is that they would be switched to the RESICO
regime). However, new entrepreneurs filing in 2022 for the first time did not have a choice to file under
the RIF regime and would automatically fall under RESICO if eligible. As a result, there will be
differentiated treatment of personal income taxpayers with similar characteristics leading to horizontal
inequality, at least in the short run, as three regimes coexist in 2022: the standard PIT regime, the RIF,
and the RESICO regimes. The net distributional impact of the reform in the short-term is uncertain but
can be empirically simulated by applying the changes in legislation to a representative sample of
taxpayers.
62. In the long term, the distributional impacts are also uncertain: the reform can reduce
informality and lead to improved welfare outcomes but could also increase incentives for tax evasion
and potentially lead to lower levels of aggregate productivity and wellbeing. Although special regimes
encourage new contributors to register, they also reduce the tax rate on those small firms already
registered and potentially incentivize large formal firms that are close to the regulatory threshold to
under-report their revenue to the tax authorities or limit their growth to avoid the additional tax
burden. For instance, Azuara et al. (2019) find the bunching of firms around the eligibility threshold of
various tax regimes in Peru. To the extent that formal firms have incentives to remain small, forgoing
opportunities to innovate and grow, this leads to smaller and less productive firms on average in the
economy, which is reflected in a decrease in welfare or total productivity (Guner et al, 2008; Garicano et
al, 2016; Almunia and Lopez-Rodriguez, 2018). In addition, to the extent that workers' social security
contributions in micro firms are reduced, as in the case of the RIF, the indirect pressure on federal
77
budgets could be significant (Azuara et al. 2019). Finally, simplified regimes also potentially misallocate
resources across sectors and distort the incentive arrangements within firms. This is because simplified
regimes force productive firms to choose between limiting their income growth so that is just below the
regulatory threshold or bearing a higher tax, both of which potentially reduce employment. Moreover,
to the extent that some of the tax burden in the corporate income tax regime falls on workers,
equilibrium wages could be lower in firms subject to the general regime, encouraging too many agents
with low managerial ability to become small entrepreneurs rather than working as employees for more
productive firms (Garicano et al 2016). By focusing on firms with specific organizational structures (e.g.,
family-owned businesses, unincorporated firms), often in the least productive activities, these
distortions could seriously affect total factor productivity (Busso et al, 2012).
63. Ultimately, the distributional impact of the reform will depend on the behavior of taxpayers.
To the extent that taxpayers were given the choice between RIF and RESICO, at least in the short run, it
is likely that taxpayers will choose the regime that minimizes their tax liabilities. This could potentially
reduce the burden on small taxpayers, but also reduce revenues in the short run.
64. Preliminary estimates suggest that the recently approved small taxpayer reform is expected to
lead to a small decline in revenue and a small negative distributional impact in the short term. The
reform is expected to lead many individuals switching to the new simplified regime. Taxpayers in the
new regime are expected to have higher profit margins than those deciding to stay in the RIF regime.
Moreover, most contributors expected to switch to the RESICO regime belong to the top quintile of the
pre-tax distribution. To the extent that taxpayers had time to assess which regime minimized their
liabilities, the reform could lead to a decline in collections and a small negative distributional impact.
The negative distributional effect is likely the result of a higher eligibility threshold and professional self-
employed workers qualifying under the RESICO regime. The distributional effects are small because
wage workers, who contribute 90 percent of all personal income taxes, do not have access to the new
regime. In the medium term, preliminary estimates suggest that 5.5 percent of previously informal self-
employed workers would become formal with the recent reform, potentially leading to an increase in
tax collections in the medium term. This could amount to a 2.1 percent decline in informal employment.
Firms that switch to formal status show that these firms are larger and potentially more productive than
those who switch to informality. To the extent that the firms switching into formality are more
productive, this could lead to a small increase in revenue collections, counteracting the reduction in tax
collections estimated in the short run.
Final remarks
65. This chapter looks at efficiency-equity trade-offs on a tax-by-tax basis and at the tax system
altogether. This is relevant because tax policy affects inequality in many ways and because the
magnitude of the efficiency-equity trade-offs depends on the interactions between many factors within
the tax system. The note looks at Mexico’s tax structure, as a whole and on a tax-by-tax basis, comparing
it with other OECD and Latin American countries. The options considered are based on clear principles:
raise revenues in a growth-and-equity-friendly way by broadening tax bases, strengthening the overall
formalization. In sum, the goal is to identify reforms that align inclusiveness and growth goals to
contribute to significantly reducing efficiency-equity trade-offs.
66. The analysis shows that Mexico’s tax structure offers many opportunities for a reform that
improves equity, simplicity, and efficiency. First, Mexico’s tax system does not contribute enough to
reduce income inequality, as the difference in the Gini coefficient for market income and disposable
income is small, particularly when compared to other OECD countries. This is in part due to a low tax-to-
78
GDP ratio relative to other LAC and OECD countries which prevents further fiscal redistribution.
Therefore, Mexico would benefit with a tax reform that enhances equity through more progressivity in
the tax system, contributing to reduce the negative distributional implications of market outcomes on
lower income households, while creating equal opportunities for all individuals and firms. Second, the
chapter shows there is room to make the tax system simpler, reducing complexity and tax-avoidance
opportunities, while easing compliance, administration, and enforcement costs. Finally, this chapter
shows that Mexico can enhance the country’s competitiveness by improving efficiency, helping minimize
distortions and supporting economic growth. Mexico could consider a comprehensive tax reform that
focuses on eliminating the exceptions to the generalized taxation of income sources and the tax
incentives that erode the bases and distort the economic activity. These exemptions and incentives
largely benefit the top of the distribution and complicate taxation. Introducing these reforms would
allow Mexico to increase tax revenues without increasing tax rates.
67. This chapter does not offer a definitive proposal for reform of the tax system, rather it
presents a menu of potential options that would help design an equity-and-growth friendly reform.
This menu is based on a large number of simulations on how broadening the bases of income and
consumption taxes would make the tax system fairer and more efficient while raising more revenues.
Some options to consider that could expand the tax base with limited impact on poverty or inequality
are given by:
 Direct taxes:
(iv) PIT: removing deductions for medical expenses, mortgage interest and school tuition,
which would increase PIT revenue by 0.05pp of GDP. Most of these deductions largely
benefit individuals with higher incomes, thus increasing progressivity, reducing post-tax
inequality, and no impact in poverty.
(v) PIT: reducing the exemption on pensions to less than 7 minimum wages (from 15 times).
PIT would become slightly more progressive, increasing revenue by 0.02pp of GDP,
without an impact on poverty or inequality.
(vi) CIT: eliminating tax expenditures from the incentives granted to the taxpayers of the
northern border region (0.3pp of GDP).
(vii) Property taxes: State level collection on behalf of municipalities may help improve
collections, reduce collection costs, and improve efficiency, effectiveness, and
sophistication in terms of fiscal cadasters, all due to economies of scale and capacity
gains.
 Indirect taxes:
(vi) VAT: Eliminating exemptions and reduced rates except for a basic food basket and
public transportation. Depending on the scenario, the additional VAT collection would
amount to an increase of between 1.4 pp or 1.6 pp of GDP. Both inequality and poverty
would increase relative to the baseline. Given these distributional impacts, eliminating
some exemptions and zero-rates for food and public transport would require additional
compensation for the poor to ensure poverty reduction and net fiscal savings.
(vii) VAT: If exemptions to educational services were eliminated, VAT revenue would
increase by 0.5pp. of GDP. There would be no significant impact on inequality. This
change would mostly affect better-off households who tend to be the ones using private
schools and other private services.
(viii) VAT: If exemptions to entertainment services and social services were eliminated, VAT
revenue would increase by 0.04pp. of GDP. There would be no significant impact on
79
inequality. This change would mostly affect better-off households who tend to use these
services.
(ix) VAT: Equalizing the standard rate at 16 percent (eliminating the reduced border rate)
across the country would improve both horizontal and vertical equity. This would
increase the progressivity of VAT, while having an insignificant impact on inequality and
a very small impact on poverty. This change would increase VAT revenue by 0.2pp. of
GDP.
(x) Tobacco taxes: Increase the specific component to MX$0.54, following the WHO
recommendation of a 75 percent tax share of the retail price of tobacco, resulting in
additional revenues of 0.01pp. of GDP. Short-term negative distributive impacts to be
offset by positive long-term distributive impact from health costs.
Table 2.8. Summary of Options for Reform

Options for reform Impact Impact on Impact
Impact on Impact on
on Poverty on
progressivity revenue
equity Reduction efficiency
(US$5.5 (% of tax (% of
(Kakwani) (Gini)
poverty) revenue) GDP)
PIT
1 Reduce brackets to 5, combine [T1+2 + - + + -14.1 -0.48
min wage] at 0%, [2 min wage + T3],
[T4+T5+T6] and [T8+T9+T10+T11] at
the top rates
2 Reduce the number of tax brackets to 4 - + - + 73.8 2.50
while increasing the rates
3 Removing deductions for medical + i - + 1.6 0.05

expenses, mortgage interest and school
tuition
4 Reduce global deduction limit to 3 i i i i 0.4 0.01

UMAS or 0.10 of gross income,
whichever is lower
5 Introduce an individual deduction limit i i i i 0.4 0.01
6 Limit exemptions on pensions to those + i i + 0.7 0.02

that are below 7 minimum wages
7 Replace current exemptions by a fixed + + + -44.6 -1.52

individual general exemption on all
80
income
8 Replace current exemptions by a fixed + + + -41.0 -1.39
individual general exemption on labor
income
9 Combine options 3 + 6 + 8 + + + + -38.9 -1.32
10 Combine options 3 + 6 + 8 + 2 - + - + 14.1 0.48
11 Combine options 3 + 6 + 7 + 2 - + - + 5.7 0.19
VAT
1 Eliminating exemptions and zero-rated - - - + 58.8 2.31
items
2 Option 1 while compensating the poor - + + + 38.3 2.07
3 Eliminate all exemptions and zero-rates - - - + 29.3 1.15

except for food items consumed at
home and urban public transport
Impact on Impact on
on Poverty on
(US$5.5 (% of tax (% of
(Kakwani) (Gini)
4 Eliminate all exemptions and zero-rates - - - + 41.4 1.60

except for a list of 35 non-processed
food items consumed by the bottom
30% and urban public transport
5 Option 4 while compensating the poor - + + + 28.0 1.47
6 Eliminate all exemptions and zero-rates - - - 37.5 1.47

except for a list of 35 processed+non-
processed food items consumed by the
bottom 30% and urban public transport
7 Option 6 plus maize products for a total - - - 37.3 1.47

of 39 products consumed by the
bottom 3 deciles
8 Option 4 + 6 + 7 – total of 54 products - - - 35.7 1.40
9 Eliminate exemptions to educational - - - 13.1 0.51

services
10 Eliminate exemptions to entertainment + - - 1.0 0.04

services
11 Eliminate border rate, equalize rate at + + - + 5.6 0.22
16% across the country
81
12 Eliminate border rate and increase + + - + 11.3 0.44
standard rate to 17%

15 Option 14 and compensate the poor + + + + 18.2 0.83
Excise taxes
1 Update the specific component for + i - + 2.7 0.01
inflation over the period 2011-19 to
MX$0.49;
2 Increase the specific component to i i - + 4.0 0.01

MX$0.54, following the WHO
recommendation of a 75 percent tax
share of the retail price of tobacco;

Impact on Impact on
on Poverty on
(US$5.5 (% of tax (% of
(Kakwani) (Gini)
3 Increase the specific component to - i - + 9.4 0.02

MX$0.74, equivalent to increasing the
specific IEPS from Scenario 1 by 50
percent;
4 Increase the specific component to - i - + 16.3 0.03

MX$1;
5 Increase the specific component to - - - + 33.7 0.06

MX$1.49, in line with a policy
discussion in Congress (El Financiero,
2020).
Source: World Bank microsimulation model.

Note: A positive sign “+” signals and improvement in progressivity, poverty and inequality reduction. A negative sign “-”
signals worsening progressivity, poverty and inequality reduction. A “i” signals insignificant changes.
82
References
Akgun, O., D. Bartolini and B. Cournède (2017), “The capacity of governments to raise taxes”, OECD
Economics Department Working Papers, No. 1407, OECD Publishing, Paris,
https://dx.doi.org/10.1787/6bee2df9-en.
Ahmad, E., M. Best, and C. P¨ oschl (2012): “Tax reform in the presence of informality in developing
countries: incentives to cheat in Mexico,” Tech. rep.
Arlinghaus, J. and K. van Dender (2017), “The environmental tax and subsidy reform in Mexico”, OECD
Taxation Working Papers, No. 31, OECD Publishing, Paris, https://dx.doi.org/10.1787/a9204f40-
en.
Arnold, J. et al. (2011), “Tax policy for economic recovery and growth”, The Economic Journal, Vol.
121/550, pp. F59-F80, http://dx.doi.org/10.1111/j.1468-0297.2010.02415.x.
Attanasio, O., J. Banks, and M. Wakefield. (2004). Effectiveness of tax incentives to boost (retirement)
saving: theoretical motivation and empirical evidence, No W04/33, IFS Working Papers, Institute
for Fiscal Studies, https://EconPapers.repec.org/RePEc:ifs:ifsewp:04/33.
Bachas, Pierre; Gadenne, Lucie; Jensen, Anders. 2020. Informality, Consumption Taxes and
Redistribution. Policy Research Working Paper;No. 9267. World Bank, Washington, DC. © World
Bank. https://openknowledge.worldbank.org/handle/10986/33851 License: CC BY 3.0 IGO.
Batis, C., M. Arantxa Colchero, F.M. Hernandez, and J.A. Rivera (2019). “Reduction in purchases of
energy-dense nutrient-poor foods in Mexico associated with the introduction of a tax in 2014”,
Preventive Medicine 118, 16-22.
Casar, J. I. (2017). Deuda pública y política fiscal en México. Revista de Economía Mexicana Anuario
UNAM. Facultad de Economía UNAM. México
Casar, José I. (2020). Hacia una Reforma Fiscal para el crecimiento y la igualdad / José I. Casar. Primera
edición. Ciudad de México : Universidad Nacional Autónoma de México, Programa Universitario
de Estudios del Desarrollo, 2020.
Castelletti, B. (2013). How redistributive is fiscal policy in Latin America? The case of Chile and Mexico
OECD Development Centre, Working Paper Nº 318, julio. París
83
Centro de Estudios de las Finanzas Públicas, (2018). “El tabaquismo y su efecto en las finanzas públicas
2007-17”.
Chavez Maza, L. A. & V. Lopez Toache (2019). Determinantes del recaudo del impuesto a la propiedad
inmobiliaria: un enfoque estadístico para México. Rev. iberoam. estud. munic. 2019, n.19
http://dx.doi.org/10.4067/S0719-17902019000100089.
Colchero, M. Arantxa, J.C. Salgado, M. Unar-Munguia, M. Hernandez-Avila, and J.A. Rivera-Dommarco,
(2015). “Price elasticity of the demand for sugar-sweetened beverages and soft drinks in
Mexico”, Economics and Human Biology 19, 129-37.
Colchero, M. Arantxa, J. Rivera-Dommarco, B.M. Popkin, and S. Wen Ng, (2017a). “In Mexico, evidence
of sustained consumer response two years after implementing a sugar-sweetened beverage
tax”, Health Affairs 36 (3), 564-71.
Colchero, M. Arantxa, C. Guerrero-Lopez, and M. Molina, (2017b). “After Mexico implemented a tax,
purchases of sugar-sweetened beverages decreased and water increased: Difference by place of
residence, household composition, and income level”, Journal of Nutrition 147 (8), 1552-57.
Cordera, R. y Provencio, E. (coords.), (2018) Informe del desarrollo en México. Propuestas estratégicas
para el desarrollo 2019-2024, México: UNAM.
El Financiero (2020). “Con impuestos al cigarro esperan reducir 1.5 millones de fumadores en 10 años”.
Septiembre 30, 2020.
Encuesta Nacional de Consumo de Drogas, Alcohol, y Tabaco (ENCODAT). (2017). “Encuesta Nacional de
Consumo de Drogas, Alcohol y Tabaco 2016-17: Reporte de tabaco”.
Espinosa, Martínez y Martell (2018). ¿por qué algunos municipios en méxico son mejores recaudadores
de impuesto predial que otros? Gestión y Política Publica, 27(2), 375.
Fuchs Tarlovsky, A. and M.F. Gonzalez Icaza (2020). “The welfare and distributional effects of taxing SSB
to reduce the risks of obesity in Ukraine”, mimeo.
Fuchs Tarlovsky, A. and F. Meneses. (2017). “Regressive or progressive? The effect of tobacco taxes in
Ukraine”, World Bank Policy Research Working Paper 8227.
Harris, T., D. Phillips, R. Warwick, M. Goldman, J. Jellema, K. Goraus, and G. Inchauste (2018).
“Redistribution via VAT and cash transfers: an assessment in four low and middle-income
countries.” IFS Working Paper W18/11. Available at: https://www.ifs.org.uk/publications/12867
Jiménez Ruiz, J.A., B. Sáenz de Miera, L.M. Reynales-Shigematsu, H.R. Waters, and M. Hernandez-Avila,
(2008). “The impact of taxation on tobacco consumption in Mexico”, Tobacco Control 17, 105-
10.
Lustig, Nora (2018). CEQ Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty.
CEQ Institute at Tulane University and Brookings Institution Press. Available at:
https://commitmentoequity.org/publications-ceq-handbook
Levi, Santiago. (2019). Una prosperidad compartida: propuestas para crecer y transformar la seguridad
social en México. BID. Washington
84
Macias Sánchez, A., H.J. Villarreal Páez, J.S. Méndez, and A. Garcia Gómez. (2020). “Análisis costo-
beneficio del consumo de tabaco en México”, Centro de Investigación Económica y
Presupuestaria.
Muñoz-Piña, C., M. Rivera Planter, & M. Montes de Oca (2020), “Implicit and Explicit Carbon Pricing in
Mexico: From inefficiency-inducing subsidies to climate-positive taxes”, Mimeo.
OECD (2018), Getting it Right: Strategic Priorities for Mexico, OECD Publishing, Paris,
https://dx.doi.org/10.1787/9789264292062-en.
OECD (2019), OECD Economic Surveys: Mexico 2019, OECD Publishing, Paris.
https://doi.org/10.1787/a536d00e-en
OECD, et al. (2020). Revenue Statistics in Latin America and the Caribbean 2020. OECD Publishing, Paris.
https://doi.org/10.1787/68739b9b-en-es
OECD et al. (2021), Revenue Statistics in Latin America and the Caribbean 2021, OECD Publishing, Paris,
https://doi.org/10.1787/96ce5287-en-es.
Secretaria de Hacienda y Crédito Público (SHCP). (2019). Acuerdo136/2019. Diario Oficial de la
Federación. Gobierno de México.
Scott, John, Enrique de la Rosa and Rodrigo Aranda. (2017). "Inequality and Fiscal Redistribution in
Mexico," Commitment to Equity (CEQ) Working Paper Series 65, Tulane University, Department
of Economics. https://ideas.repec.org/p/tul/ceqwps/65.html#download
Tello, C. (2014). La economía política de las finanzas públicas: México 1917-2014. Facultad de Economía,
UNAM. México
Unda Gutiérrez, M. & C. Moreno Jaimes (2015). Property Tax Collection in Mexico: An Analysis of Its
Economic Determinants in the Period 1969-2010, Revista Mexicana de Ciencias Políticas y
Sociales, Volume 60, Issue 225, 2015, Pages 45-77, ISSN 0185-1918,
https://doi.org/10.1016/S0185-1918(15)30019-2.
World Health Organization (WHO). (2015). “WHO report on the global tobacco epidemic, 2015: Raising
taxes on tobacco”.
World Health Organization. (2019). “WHO report on the global tobacco epidemic, 2019: Offer help to
quit tobacco use”.
World Bank. (2019). Mexico – Systematic Country Diagnostic (English). Washington, DC: World Bank.
World Bank. (2021). La Participacion Laboral de la Mujer en Mexico (Spanish). Washington, D.C. : World
Bank Group. http://documents.worldbank.org/curated/en/753451607401938953/La-
Participacion-Laboral-de-la-Mujer-en-Mexico
85
Appendix 2.1: Methodological Appendix
We begin by first assessing the incidence of fiscal policies under a baseline. Using the 2018 ENIGH
survey, we first model the fiscal interventions in place in 2018 using the Commitment to Equity
Approach (Lustig, 2018), which has been used to describe fiscal redistribution in Mexico over the last
three decades (Scott et al 2017). The usefulness of this approach is that it allows for detailed analysis of
the distributional impact of each of tax and spending component, as well as the impact of the
combination of policies. The approach relies on state-of-the art tax and benefit incidence analysis.
However, it does not include behavioral responses, lifecycle or general equilibrium effects, and it
excludes some important taxes (such as corporate and international trade taxes) and spending
categories that surely affect the distribution of income, but where methodologies to ascertain which
individuals are affected are still to be fully developed (including infrastructure investments).
With 2018 as a baseline, we build a microsimulation model to project the incidence of alternative
policy reforms going forward. To achieve this, instead of directly identifying taxpayers and program
beneficiaries observed in the household survey, as done in standard incidence analysis, each
intervention is modeled according to tax and spending rules, and the resulting simulations compared
against actual revenue and spending outcomes. This then allows for simulations of the policy changes
undertaken in 2019 and alternative reform scenarios going forward. The microsimulation model
provides partial equilibrium simulations in line with international practice. 54 Given the interest in
modeling alternative reform scenarios, we abstract from the 2020 crisis, as this could distort the results,
and use 2019 policies as the baseline.
The microsimulation model captures details of the tax and transfer system. When it comes to the
direct tax and transfer system, the model captures: (i) the full range of institutional features of the tax
and benefit system, including the appropriate income aggregate relevant for computing the taxable base
and income-tested benefits; (2) a precise characterization of thresholds, floors, ceilings, and relevant tax
rates; and (3) specific eligibility rules used in computing benefit entitlements. Detailed modeling of
direct taxes allows for analysis of exemptions, deductions, and allowances as well as the potential
impact of removing these. Similarly, detailed modeling of the beneficiary households allows for
simulations of alternative reforms. When it comes to value added taxes, considerable effort is made to
model exemptions and reduced rates, as well as the embedded VAT in exempt items. Excise taxes on
fuel take account of direct effects on households as well as indirect effects through the increase in the
price of all other goods.
Treatment of informality
Special effort is made to capture non-compliance for each type of tax. For personal income and payroll
taxes, we use information from the household survey on whether the individual participates in the
contributory health insurance system as a proxy for informality. For indirect taxes, we use information
on the place of purchase of each item in the consumption module as a proxy for informality. Based on
the information from the household survey, we classify the place of purchase as formal or informal in
line with the recent literature (Bachas et al 2020). Then, we assume both direct and indirect effect for
purchase in formal places but only indirect effect for informal purchase; that is, for informal purchase
direct effect is zero and only cascading effects exist.55
54
For EU countries and the UK see Euromod models at https://www.microsimulation.ac.uk/euromod/, for Canada see:
https://www.statcan.gc.ca/eng/microsimulation/index , for Australia see:
https://www.governanceinstitute.edu.au/centres/national-centre-for-social-and-economic-modelling-natsem#models
55
This scenario is considered as an upper bound of VAT. Alternately, other scenarios could be assumed with zero effect for
informal purchase (lower bound of VAT) and both direct and indirect effects for formal purchase.
86
Direct taxes
For direct taxes, the microsimulation model estimates the taxable base in the household survey,
taking account of minimum wage, informal workers, and exempt incomes. Observed monthly incomes
for wage workers are assumed to be reported in net terms and are therefore grossed up in line with the
PIT tax schedule, tax credit schedule and expected monthly social contributions (See schedules in
Appendix 2). Self-employed income is assumed to be reported in gross terms. Once estimated, the
monthly labor income taxable base is added with the cumulative gross income from non-labor income
sources including income from leases, interest, home-owned businesses, private transfers, and
retirement income. The model then annualizes the taxable base and applies personal deductions which
can be identified in the consumption module of the household survey. 56 When compared to tax
administration records, the model generates a tax structure that is close to the observed structure
(Error: Reference source not found), the largest difference being at the top of the income distribution, as
high-income earners typically do not respond to household surveys. As shown in Error: Reference source
not found, the household survey does not have information for top income earners – the sample is
truncated- and it also observes far fewer contributors at the top, leading to a lower density of
density: log_y
households with relatively high incomes.57
Figure A.1. Distribution of PIT contributors: Figure A.2. Kernel density of market income:
.4
.3
.2
.1
0
Survey vs Tax Administration data, 2018 Survey vs Tax Administration data, 2018
40%
35%
Share of contributors
30% SHCP Mexico Simtool
25%
20%
15%
10%
5%
0%
1 2 3 4 5 6 7 8 9 10 11 -5 0 5 10 15 20
Brackets Log market income
Survey Admin. data
Source: World Bank estimates based on ENIGH 2018 and SHCP tax administration data.
Note: Right side graph shows a first threshold at which administrative data starts to have a much higher density, as
well as the threshold at which survey no longer has any observations.
Simulation results (PIT)
Option 1: Reduce the number of tax brackets to 5, reducing the bottom rates
68. Reducing the number of tax brackets from 11 to 5, while increasing rates would reduce
progressivity of the PIT and increase poverty. Assuming the existing tax credit scheme works perfectly,
if the first simulation reduces the 11 current tax brackets {T1, T2,….T11} to 5 brackets by combining the
first bracket plus incomes that are less than twice the minimum wage [T1+2 min wage] and setting the
marginal rate of 0 percent, and then combining the next four brackets as follows [2 min wage + T3],
[T4+T5+T6], [T7] and [T8+T9+T10+T11] and setting these at the current top marginal rate in each set
(Error: Reference source not found). This structure would improve the progressivity of the PIT
(increasing the Kakwani index from 0.289 to 0.398), as it would reduce marginal rates at the bottom of
56
The methodology considers payments that must be made by digital means to make the deduction effective.
57
Analysis based on 8,000 bins from tax administration records, where the highest monthly market income is close to MX$
35,000,000 compared to the highest market income observed in the household survey of MX$ 1,300,000.
87
the income distribution, while at the same time increasing the marginal rate for higher incomes. This
reform would also have a positive impact on equity and poverty reduction compared to the baseline
(Error: Reference source not found). However, it would lead to a 14.1 percent reduction in PIT revenue,
equivalent to 0.48 percentage points of GDP.
Figure A.3. Option 1: Reducing the number of tax Figure A.4. Option 2: Reducing the number of tax
brackets to 5, reducing the bottom rates brackets and increasing marginal tax rates
Marginal Rate Marginal Rate
40%
40%
35.0% 35%
35% 32%
30% 28% 33%

30%
25%
25% 23.5% 22%
21.4%
20%
20%
15%
15%
10.9%
Baseline Simulation Baseline Simulation
10% 10%
5% 5%
0.0%
0% 0%
- 50,000 100,000 150,000 200,000 250,000 300,000 - 50,000 100,000 150,000 200,000 250,000 300,000
PIT Thresholds PIT Thresholds
Source: World Bank Source: World Bank

Option 2: Reduce the number of tax brackets to 4, beginning with a 22 percent marginal rate
69. Reducing the number of tax brackets from 11 to 4, while increasing rates would reduce
progressivity of the PIT and increase poverty. Assuming the existing tax credit scheme works perfectly,
individuals at the bottom of the distribution would continue to face negative income taxes, while those
with incomes above the tax credit threshold would now face a higher marginal rate, ranging from 22 to
33 percent. Under such a scenario, the incidence of PIT as a share of market income would increase for
all deciles, leading to a 73.8 percent increase in PIT collections, equivalent to 2.5 percent of GDP.
However, this reform would lead to a reduction in progressivity of the PIT (reducing the Kakwani index
from 0.289 to 0.154) and a 1.9 percentage point increase in poverty (Error: Reference source not found).
However, some of this could be compensated by complementary reforms to the tax credit system and
to the system of exemptions and deductions, as described below.
Option 3: Remove deductions
70. Removing deductions for medical expenses, mortgage interest and school tuition would raise
revenue with no significant impacts on poverty and inequality. A reform removing deductions for
medical expenses, mortgage interest deduction and school expenditures (used by private educational
costs) would increase PIT revenue by 1.6 percent (equivalent to 0.05 percent of GDP). Most of this
additional revenue would come from mortgage interest deductions (0.028 percent of GDP), compared
to medical deductions (0.002 percent of GDP) and school deductions (0.003 percent of GDP). Given that
most of these deductions largely benefit individuals with higher incomes, the reform would increase
slightly progressivity, reduce post-tax inequality, while having no impact in poverty (Error: Reference
source not found). 58 Deductions would also become slightly less regressive.
58
All of the simulation scenarios presented assume that the tax credit system works as designed, with low-income individuals
effectively receiving a credit. The direction and magnitude of the changes reported below are not very different when this
assumption is lifted.
88
Option 4: Reduce the global deductions limit to 3 UMAS or 10 percent of gross income
71. Reducing the global deductions limit would have no impact on poverty or inequality, while
generating a small increase in revenue. Current legislation sets a maximum global limit to the sum of
total deductions at 5 UMA 59 or 15 percent of gross income, whichever is smallest. The proposed reform
would reduce this limit to 3 UMAs or 10 percent of gross income, whichever is smallest. Assuming the
tax credit system works as designed, this reform would have no significant impact on the progressivity of
PIT, nor would it have any impact on poverty or equity, but it would reduce the regressivity of
deductions (Error: Reference source not found). The reform would generate a 0.31 percent increase in
PIT collections, equivalent to 0.1 percentage points of GDP.
Option 5: Reduce individual PIT deductions limits
72. Adding an individual PIT deductions limit equivalent to one minimum wage would have no
impact on poverty or inequality, while generating a small increase in revenue. In addition to the global
limit on deductions, the legislation places individual limits to people earning more than MX$400,000 a
month (about US$20,100). Individual limits on medical and funeral expense deductions are equivalent to
one minimum wage. In addition, there are individual limits on donations equivalent to 7 percent of gross
income, while mortgage interest deductions have a limit on the size of the loan, such that it does not
exceed one unit of investment, equivalent to about MX$52,000 (US$2,620). If individual deduction limits
were all set to one minimum wage, including on mortgage interest rate and donations and if additional
limits were introduced for school expenditure deductions equivalent to one minimum wage deductions,
all while keeping global deductions constant, this would improve the progressivity of the PIT, as
deductions would become less regressive, without a significant impact on poverty or inequality ( Error:
Reference source not found). Moreover, the reform would generate an 0.37 percent increase in PIT
revenue, equivalent to 0.01 percentage points of GDP.
Option 6: Limit exemptions on pensions from 15 to 7 minimum wages
73. Similarly, limiting PIT exemptions on pensions would generate additional revenue with
insignificant impacts on poverty and inequality. Current legislation places an exemption on pensions
that have a value that is below 15 times the minimum wage. If such an exemption were to be reduced to
a value below 7 minimum wages, this would broaden the tax base without an impact on poverty or
inequality (Error: Reference source not found). PIT would become slightly more progressive while
generating a 0.7 percent increase in PIT revenue, equivalent to 0.02 percent of GDP.
Option 7: Provide a general exemption to labor incomes to eliminate horizontal inequality
74. Replacing the current set of exemptions by a fixed individual exemption would reduce both
vertical and horizontal inequality. As shown earlier, most existing exemptions benefit individuals with
higher incomes, leading to vertical inequality. In addition, individuals working in large firms also benefit
more from exemptions than similar individuals working in small firms, simply because larger firms
provide benefits such as incentives, bonuses, prizes, and profit sharing, leading to horizontal
inequalities. For instance, a worker earning 1 to 2 minimum wages working in a formal firm with 1 to 5
workers receives 6 types of benefits, while a worker earning the same amount working in a firm with
more than 500 workers receives over 10 types of benefits (Error: Reference source not found).
75. A general exemption would improve progressivity and improve equity, and lead to a small
reduction in revenue. Legislation could be modified to allow individuals a choice between a fixed
general exemption and the current exemption system. Under this scheme, individuals would choose the
59
UMA stands for units of measurement and update (Unidad de Medida y Actualización), which serve as an economic reference
unit used to calculate payments, obligations or penalties for the government.
89
maximum of the exemptions they are now eligible for and a proposed general exemption of MX$88,588
a year on all taxable income. Under this scenario, assuming the tax credit system works perfectly, there
would be a reduction in the incidence of PIT for lower income individuals, as a larger share of low and
middle-income individuals would be able to benefit from exemptions, which would lead to an
improvement in progressivity of the PIT system (the Kakwani would increase from 0.29 to 0.4) and a
reduction in inequality, while at the same time substantially improving the progressivity of exemptions
(Kakwani of exemption benefits improves from -0.37 to 0.12) (Error: Reference source not found). The
reform would have a cost, leading to a 45 percent reduction in PIT collections, equivalent to 1.5 percent
of GDP.
Figure A.5. Number of benefits received by Figure A.6. Impact on Poverty and Inequality of a general PIT
size of firm and level of income exemption
14 0.500
0.400
12 0.400
0.289
10 0.300
8 0.200
Average number of benefits received
0.115
6 0.100 0.019
0.014 0.000 0.001
0.000 -0.010
4 -0.001 -0.003 -0.001
-0.100
2
-0.200
0
<1 1-2 2-4 4-6 6-8 8-10 > 10 -0.300 Baseline 2019 General exemption
Number of minimum wages -0.400 -0.374
-0.500
Kakwani
Kakwani
Marginal contribution to greater equality

Marginal contribution to poverty reduction

Firm with 1-5 workers 6-10 11-15 16-20 21-30 31-50 51-100 101-250 251-500 501+ workers
Option 8: Provide a general exemption to labor incomes to eliminate horizontal inequality

76. A general exemption applicable only to labor incomes would improve progressivity and
improve equity, with a slightly smaller impact on revenue. If the general exemption were to be applied
only to labor income (thus excluding self-employed workers), would lead to an improvement in
progressivity of the PIT system (the Kakwani would increase from 0.29 to 0.38). The reform would cost
1.4 percent of GDP, with slightly smaller impacts on equity. If combined with some of the reforms noted
below, a general exemption could help to mitigate some of the negative distributional impacts as
detailed below.
Combination of policies
77. A combined package of policies could generate additional revenue sources while protecting
the poor. Below we present three combined scenarios which combine the alternative reforms presented
above in Error: Reference source not found. The first scenario, shown in column 9, combines the
elimination of deductions on mortgage interest, health expenditures and educational expenses (column
3), limits exemptions on pensions (column 6), and introduces a general exemption for labor incomes
only (column 8). The second scenario, shown in column 10, is the same, but also reduces the number of
tax brackets to 4 (column 2). The final scenario introduces a general exemption for all types of income,
instead of limiting the exemption to labor income alone (column 7). As discussed earlier, the elimination
of exemptions and deductions lead to small increases in revenue, with little or no change in poverty. The
reduction in the number of brackets raises revenue, while increasing poverty by 1.9 percentage points
and increasing inequality. However, when a general exemption is introduced, the impact on poverty
declines, while still generating an increase in revenue. A reduction in the number of tax brackets along
90
with a general exemption applied to all types of income, leads to a poverty increase of 0.7 percentage
points, while a general exemption applied to labor income alone leads to a poverty increase of 1.2
percentage points.
Indirect taxes
For indirect taxation, the microsimulation model estimates both direct and indirect effects of VAT. To
calculate the direct effect, we apply the VAT rate to all consumption items in the household expenditure
survey following the statutory standard rate or preferential treatment (zero-rate or exemptions) 60.
Additionally, the border rate is applied to relevant jurisdictions in 2019. For indirect effect, we use the
Cost-Push model to estimate the cascading from VAT-taxed inputs to exempted final goods 61. The
augmented Input-Output matrix, that is mapped to every item in the household survey, is the main
input data. We assume that all input sectors (except those exempt or zero-rated) carry a price shock
equal to T/(1+T), where T is the VAT statutory rate. Input sectors that are exempted/zero-rated do not
generate a “VAT shock”. Since we assume that sectors with positive rate or zero-rate are fixed (no
cascading effect), only for exempted sectors, we calculate the increase of price (or tax burden) from
VAT-liable inputs.
Excise taxes are paid for the production, sale, and importation of certain products and services, such
as fuels and non-essential products like tobacco and alcohol. There are three types of excise taxes in
Mexico: ad valorem, specific (a fixed amount per unit), and mixed. The microsimulation model estimates
the direct effect of all excise taxes and the indirect effect of gasoline taxes. For the direct effect, ad
valorem rates and/or specific taxes are applied to the consumption of the good or service (the Appendix
presents the list of excise taxes used for the microsimulation). To estimate the indirect effect on
gasolines, the model uses the Input-Output matrix and a cost-push model.62
Simulation results (VAT)
60
Preferential treatments include the following goods and services: non-processed vegetables and animals, medicines,
processed food (except: drinks different than milk, syrup used for concentrated drinks, caviar/smoked salmon,
flavorings/additives, chewing gum, pet food), ice and water, tractors and other equipment for agricultural uses, fertilizers,
pesticides, herbicides and fungicides, hydroponic greenhouses, gold and jewelry, books, newspapers and magazines, lottery
tickets, services provided to members of associations/clubs/labor unions (e.g. religious, cultural, sports, professional), and
public shows (except: cinema, theater, circus and those provided in restaurants/bars).
61
In a VAT system, final goods that are VAT-liable or subject to zero-rate have in common the fact that sellers/producers are
eligible for VAT credit. In contrast, for exempt goods, sellers cannot claim VAT credit, so it is expected that they will trespass
part of their Input VAT to consumers in terms of final higher prices; this effect is called “cascading” or “indirect effects”.
62
In the cost-push model, the prices in the fuel sector are fixed while the rest of the sectors have the following price shock : dp
(0.58) + t premium /P premium (0.11) + t Diesel / Pdiesel (0.31) where t is the statutory rate and P is the price of
0.500
0.400
0.400
0.289
=
0.300
0.200
0.115
0.100 0.019
0.014 0.000 0.001
0.000 - 0.010
- 0.001 - 0.003 - 0.001
- 0.100
- 0.200
- 0.300 B a s eline 2019 G enera l exem pti on

Kakwani
Kakwani
- 0.400 - 0.374
- 0.500
fuels. The average price change is a weighted average based on the consumption of the different types of fuels (magna,
premium and diesel).
91
Eliminate exemptions and zero-rated items
78. Eliminating exemptions and zero-rated items would significantly increase tax collection but it
would mostly affect the poor. In the most aggressive scenario, if all zero-rated and exempt items were
moved to the standard rate of 16 percent, this would generate a 58.8 percent increase in VAT collection,
equivalent to an increase of 2.31 percent of GDP. However, this measure would disproportionately
affect the bottom of the distribution. The tax would become regressive, the Kakwani index would go
from 0.009 to -0.039, poverty headcount would increase by 1.5 percentage points relative to the 2019
baseline, and inequality would increase by 0.3 Gini points.
79. However, if part of the increase in VAT tax collection were spent on pro-poor programs,
eliminating exemptions and zero-rated items could reduce poverty while generating fiscal space. The
targeted scholarship program Benito Juárez provides economic support to children from vulnerable
populations enrolled in public schools. The program grants a monthly support of $800 pesos per family
throughout the school year and covers about 3.7 million families. In 2019, the transfer represented
about 6.1 percent of the disposable income of the poorest decile. If households were fully compensated
for the elimination of exemptions and zero-rated items through this program, this would reduce poverty
by 0.3 pp relative to the 2019 baseline and still leave net fiscal savings. Increasing the transfer from $800
to $2,300 would cost an additional spending of 0.24 percent of GDP but still leave 2.07 percent of GDP
fiscal savings after elimination of exemptions and zero-rated items. Moreover, eliminating these
exemptions while increasing transfers would reduce inequality by 0.5 Gini points.
Eliminate exemptions and zero-rated items except for food and public transport
80. Any changes in the taxation of food and public transport would disproportionally affect the
poor as they represent close to 60 percent of the total consumption of the bottom decile. Food
represents about half of the total consumption for the poorest decile, compared to 38 percent for the
top decile. In addition, public transport represents 6 percent of the total consumption for the bottom
decile, doubling the share of the richest decile. Therefore, in a less aggressive reform scenario, if
exemptions and zero-rated were to apply only to food and transport items, these could still have a larger
impact on the lower part of the distribution.
81. Eliminating exemptions and zero-rated items for everything other than food and public
transport would also increase tax collection with slightly better impacts on the poor. If all zero-rated
and exempt items were moved to the standard rate of 16 percent except for food items consumed at
home and public transport, this would generate a 29 percent increase in VAT collection, equivalent to
1.2 percent of GDP. This measure would lead to a 0.6 percentage point increase in poverty and an
increase in inequality, as the VAT would become slightly regressive.
Eliminate exemptions and reduced rates except for a basic food basket and public transport
82. Several different scenarios were simulated with basic food baskets largely consumed by the
poor. In the first scenario, all exemptions and zero-rates are removed except for a list of 35 non-
processed food items and urban public transport that were mostly consumed by the first 3 deciles 63. In
the second scenario, recognizing that households at the bottom of the distribution also consume
processed food items (Error: Reference source not found), a different list of 35 products (containing
both processed and non-processed foods) mostly consumed by the poorest deciles remained with rate
zero and urban public transport exempt 64. In the third scenario, the list of 35 processed and non-
processed foods is augmented to account for maize products, for a total of 39 products consumed by
63
Non-processed food scenario considers zero-rate for: avocado, beef steak, rice, pork steak, pumpkin, zucchini, onion,
jalapeno pepper, chile serrano, beef, pork rib and chop, beans, eggs, tomato, powdered milk, cow's milk, lettuce, lemon, corn,
apples, ground beef, nopal, bananas, potato, fish, chicken, green tomato and carrots.
92
the bottom 3 deciles. The fourth scenario combines the previous four scenarios for a total of 54
products that remain zero-rated.
Figure A.7. Share of selected food baskets as Figure A.8. VAT tax collections under alternative
a percentage of total consumption by preferential treatments
disposable income deciles
12% % VAT revenue % of GDP
45 41.4 1.65
10% Non-processed food basket Processed and non-processed food basket 40 37.5 37.3
1.60 35.7 1.60
35
27.8 1.55
8% 30
25 1.50 1.50
Percent of total consumption
1.47 1.47
20 1.45
6%
15
1.40
Percent of VAT revenue

1.40
10
Percent of GDP
4% 5 1.35
0 1.30
2%
(4) 54 products
(1) Non-processed
(1) w/compensation
(2) Non- & processed
(3) 39 products (w/maize)

0%
1 2 3 4 5 6 7 8 9 10
Disposable income deciles
Source: World Bank Microsimulation Model and ENIGH 2018

83. Preferential treatment for certain foods and urban public transport, moving all other items to
the standard rate would increase tax collection at the cost of more poverty and inequality. Moving
zero-rated and exempt items to the standard rate of 16 percent, would increase VAT collection between
28.0 and 41.4 percent depending on the scenario (Error: Reference source not found). The additional
VAT collection would amount to an increase of between 1.4 pp or 1.6 pp of GDP, depending on the
scenario. However, it has significant distributional impacts. In the first scenario, despite keeping a zero
rate on the 35 most consumed items by the poor, moving all other food items to a standard rate would
still have an impact at the bottom of the distribution. Value added tax would become regressive
(Kakwani index of -0.015, -0.006, -0.005 and -0.003, respectively) and inequality would increase (by 0.16,
0.10 and 0.09 Gini points, respectively) (Error: Reference source not found). The poverty headcount rate
would increase by 0.9, 0.8 and 0.7 percentage points relative to the 2019 baseline on each of the
scenarios considered, respectively (Error: Reference source not found).
Figure A.9. Progressivity and Inequality under Figure A.10. Poverty under alternative preferential
alternative preferential VAT treatments VAT treatments
(US$5.5 a day per capita)
64
Non-and processed food scenario considers zero-rate for the following items: Vegetable Oil, bottled water, rice in grain,
sugar, pork steak, beef steak, coffee, onion, pork crisp, chile serrano, chorizo, pork rib and chop, beans, cookies, eggs, natural
juices and nectars, tomato, cow's milk, corn, apples, ground beef, bread, pastries, potato, pasta for soup, fish, chicken, cheese,
green tomato and corn tortilla.
93
Increase in inequality Kakwani (right axis) 1.2
0.2 0.16 0.000 1.0 0.9
0.10 0.10 0.09
0.1 -0.002 0.8 0.8 0.7 0.7
0.0 -0.004
0.6
-0.006
-0.1 0.4
-0.008
-0.2 0.2
-0.010
-0.3 -0.012 0.0
Increase in poverty (in ppt)

Increase in inequality (in Gini points)
-0.4
Kakwani coefficient
-0.39 -0.014 -0.2 -0.2
-0.5 -0.016
-0.4
(4) 54 products
(4) 54 products
(1) Non-processed
(1) w/compensation
(1) Non-processed
(1) w/compensation


Source: Microsimulation Model based on ENIGH 2018.
Note: “Non-processed Food” refers to a scenario where all exemptions and reduced rates are eliminated except for a list of 35
non-processed food items and public transport; in the “non-processed and processed Food” scenario another list of 35 items
and public transport remains exempt, and in “All Foods” scenario all food items (except those consumed outside home) and
public transports are exempt.
84. Given these distributional impacts, eliminating some exemptions and zero-rates for food and
public transport would require additional compensation for the poor to ensure poverty reduction and
net fiscal savings. Any reduction on exemptions and reduced rate of the most consumed items among
the poor would increase poverty unless they are adequately compensated. Taking the most extreme
scenario -non-processed foods-, a compensation through the scholarship program Benito Juárez of
$1,800 pesos per month would offset the negative impact on poverty as the headcount rate would be
0.2 ppts lower than the baseline, and slightly reduce inequality (0.4 Gini points respect to baseline). This
measure would cost an additional 0.16 pp of GDP and yet it would still leave net fiscal savings of 1.47 pp
of GDP.
Eliminate exemptions to educational services
85. Eliminating exemptions on educational services would have a small impact on poverty and
inequality, while raising additional revenue. If exemptions to educational services were eliminated, VAT
revenue would increase by 13 percent, equivalent to 0.5 percentage points of GDP. VAT would continue
to be progressive and there would be no significant impact on inequality. The US$5.5 poverty headcount
rate would increase slightly, by 0.3 ppts, as this change would mostly affect better-off households who
tend to be the ones using private schools and other private services. A small increase in transfers or
scholarships could compensate households at the bottom of the distribution while attaining fiscal
savings.
Eliminate exemptions to entertainment services
86. Eliminating exemptions on entertainment services would have insignificant impact on poverty
and inequality, and a small increase in revenue. Currently there are exemptions on entertainment
services and social services such as clubs and associations. If these exemptions were eliminated, VAT
revenue would increase by one percent, equivalent to 0.04 percentage points of GDP. VAT would
continue to be progressive and there would be no significant impact on inequality. The US$5.5 poverty
94
headcount rate would increase slightly, by 0.1 pp., as this change would mostly affect better-off
households who tend to use these services.
Increase the VAT rate and eliminate reduced border rate and compensate the poor
87. An alternative set of reforms would aim to equalize the standard rate across the country and
potentially increase the standard rate. Since 2019, a preferential rate of 8 percent was applied in
municipalities at the northern border, a reform that reversed the 2014 effort to equalize VAT rates by
raising the preferential northern border rates from 11 to 16 percent. Reversing this reform, equalizing
the standard rate at 16 percent across the country would improve both horizontal and vertical equity, as
households living in border towns have higher incomes than those living in the South. In addition, three
different scenarios were simulated, where the standard rate was raised to 17, 18 and 19 percent,
respectively, while eliminating the reduced border rate.
88. Increasing the border rate to 16 percent would increase the progressivity of the VAT, with
insignificant impacts on poverty and inequality, but would generate additional revenue. Equalizing the
standard rate across the country would increase the progressivity of VAT (Kakwani index increases to
0.013), while having an insignificant impact on inequality and a very small impact on poverty (increase of
0.1 percentage points). This change would nevertheless increase VAT revenue by 5.6 percent, equivalent
to 0.2 percentage points of GDP.
89. Increasing and equalizing the standard rate to 19 percent would increase tax collection by
close to 1 percentage point of GDP but poverty would also increase by almost half of a percentage
point. VAT collection would increase 11.3, 16.8 and 22.3 percent as the standard rate increased from 16
percent to 17, 18 and 19 percent, respectively. The additional tax collection corresponds to an increase
of 0.44, 0.66 and 0.87 pp. of GDP. Nevertheless, poverty would also go up by 0.17, 0.26 or 0.39 pp.
depending on the scenario. Inequality, on the other hand, would slightly decrease by 0.01 Gini points in
all cases.
90. A larger share of the gains in VAT collection comes from the equalization of the border rate.
Simulating the same increases of the standard rate (17, 18 and 9 percent) without eliminating the
reduced border rate (i.e maintaining border rates at 8 percent), would result in a much lower increase in
tax collection (0.2, 0.39 and 0.58 pp. of GDP, respectively). Therefore, a large share of the increase in tax
collection (almost half in the 17-rate scenario) would come from applying the same standard rate
throughout the country.
91. If the poor were compensated with some of the additional VAT revenue, the reform would
make VAT more progressive, would reduce poverty relative to the baseline, while achieving overall
fiscal savings. Combining the elimination of the border rate with an increase in the standard rate would
make VAT more progressive (Kakwani increases from 0.009 to 0.015) while a well-targeted transfer
could more than compensate the impact on the poor. For instance, an increase of the standard rate up
to 19 percent nationwide, combined with an increase in the Benito Juárez scholarship to $1,100 pesos
per month would reduce the US$5.5 a day poverty headcount rate by 0.8 ppts. The additional social
spending would amount to an additional 0.05 pp of GDP, but the overall fiscal saving would still be 0.83
pp. of GDP. Error: Reference source not found summarizes the scenarios presented above.
Appendix 2.2: Tax schedules

Personal Income Tax Rates
Table A.1. Monthly PIT rate 2018 Table A.2. Employment Subsidy 2018
95
Lower Upper Tax credit /
Marginal Lower Upper
threshold threshold employme
rate (%) threshold threshold
(MX$) (MX$) nt subsidy
(MX$) (MX$)
0 579 1.90% (MX$)
579 4910 6.40% 0.01 1768.96 407.02
4910 8629 10.90% 1768.96 2653.38 406.83
8629 10031 16.00% 2653.38 3472.84 406.62
10031 12010 17.90% 3472.84 3537.87 392.77
12010 24222 21.40% 3537.87 4446.15 382.46
4446.15 4717.18 354.23
24222 38178 23.50%
4717.18 5335.42 324.87
38178 72888 30.00%
5335.42 6224.67 294.63
72888 97183 32.00%
6224.67 7113.9 253.54
97183 291550 34.00%
7113.9 7382.33 217.61
291550 35.00%
7382.33 0
Payroll taxes
State Rate
Table A.3. Payroll tax Aguascalientes 2.0%
2018
Baja California 1.8%
Baja California sur 2.5%
Campeche 2.0%
Chiapas 2.0%
Chihuahua 3.0%
Coahuila 2.0%
Colima 2.0%
Ciudad de México 3.0%
Durango 2.0%
Estado de México 3.0%
Guanajuato 2.0%
Guerrero 2.0%
Hidalgo 1.5%
Jalisco 2.0%
Michoacán 2.0%
Morelos 2.0%
Nayarit 2.0%
Nuevo León 3.0%
Oaxaca 3.0%
Puebla 3.0%
Quintana Roo 3.0%
Querétaro 2.0%
San Luis Potosí 2.5%
Sinaloa 2.4%
Sonora 1.0%
96
Tabasco 2.5%
Tamaulipas 3.0%
Tlaxcala 3.0%
Veracruz 3.0%
Yucatán 2.5%
Zacatecas 2.5%
Level of schooling Max annual deduction

Table A.4 Máximum Preschool
anual 14200
deducción for schooling
Primary 12900
expenses
Middle school 19900
Vocational/technical 17100
High school 24500
Source: SAT
97
Value added taxes
Table A.5 Mexico: VAT Zero-Rated Goods and Services (Not Including Exports)
Goods Services
- Unprocessed vegetables and animals (except pets) - Those provided to farmers and
- Patented medicines and products destined to feeding 65 ranchers, if they are intended for
(except: drinks different than milk, syrup used for agricultural activities
concentrated drinks, caviar/smoked salmon, - Grinding or crushing corn or wheat
flavorings/additives, chewing gum, pet food) - Milk pasteurization.
- Ice and water - Services provided to hydroponic
- Ixtle, palm and frill greenhouses
- Tractors and other equipment for agricultural uses (more - Removal of cotton in branch
details in the Law) - The slaughter of cattle and poultry
- Fertilizers, pesticides, herbicides, and fungicides for - Reinsurance
agricultural and livestock uses - Water supply for domestic use
- Hydroponic greenhouses
- Gold, jewelry, metalwork, artistic or ornamental pieces and
ingots, whose minimum content of such material is 80%
(except final sales to consumers)
- Books, newspapers, and magazines edited by contributor
Source: World Bank staff based on Article 2 of the Value Added Tax Law.
Table A.6 Mexico: VAT Exemptions
Goods Services
- Soil - Commissions on mortgage credits for housing, except
- Housing construction those generated after the authorization of the credit or
- Books, newspapers, and magazines. As those paid to third parties.
well as the right of use and their - Commissions on retirement funds administration
exploiting (performed by the author) - Free services (except when the beneficiary is related to the
- Used movable assets (except those sold member/partner of the entity)
by firms) - Educational services (public and private)
- Lottery tickets - Public ground transportation for passengers (provided in
- Currencies, gold or silver. urban, suburban and metropolitan areas)
- Social parties, documents pending - International shipping (provided by foreign residents)
collection and credit titles - Insurance against agricultural risks, home credit insurance,
- Gold bullion with a minimum content of financial guarantee insurance that covers payment for non-
99% of said material (sold to the general compliance with issuers of securities, titles of credit or
public) documents that are subject to public offering or brokerage
- Goods sold among foreign residents in stock markets;
- Interests from financial intermediation:
65
This means that processed food is a zero-rated good. Instead, what is charged is the food “prepared to be eaten in the place
they are sold”. Place: includes both restaurants and supermarkets. Source: Source: Art. 4.3.1. of Resolución Miscelanea Fiscal
2019. Examples of prepared food taxed: https://elpais.com/economia/2015/07/01/actualidad/1435717729_658486.html
98
o For legal persons:
 Those resulting from documents pending
collection
 Those paid by decentralized entities of the
Federal Public Administration
 Those received by insurance companies
o For natural persons:
 Interests are exempt if the credit is for acquiring
investment goods/housing acquisition or
repairment;
 If they proceed from workers’ savings account
 If the credit is for financing exempt or zero-rated
goods or services
 Those resulting from financial assets of public
institutions or the federal government
 Those resulting from credit titles considered as
placed in the market by the investors
- Services provided to members of associations/clubs/labor
unions (e.g. religious, cultural, sports, professional)
- Public Shows (except: cinema tickets, theater, circus, those
provided in restaurants/bars)
- Medical services (public and private)
- Housing rent
- Land-leasing for agricultural purposes
- Benefits resulting from authorizing third parties to publish
pieces written by the author in newspapers and magazines.
(excluding advertisement, logos, industrial designs, artwork
or when the benefits result from business activities other
than selling)
- Imports:
o If they are temporary
o Luggage
o Exempt or zero-rated goods or services
o Donated goods to the government
o Publicly recognized artwork
o Gold (minimum content of 80%)
o Vehicles
- Intermediate goods for which VAT has already been paid
Source: World Bank staff based on Articles: 9, 15, 20 and 25 of the Value Added Tax Law.
99
Excise Taxes
Table A.7 Excise taxes in 2018
Item Description Ad valorem Specific

< 14 G.L 26.5%
Beer and spirits >14 G.L & <20 G.L 30%
> 20 G.L 53%
0.35 pesos per
Tobacco cigarettes 160% cigarette
cigars 160%
Magna 4.59 pesos per liter
40.52 cents per liter
Premium 3.88 pesos per liter
Automobile fuels
Diesel 5.04 pesos per liter
Energy drinks 25%
Sugar-sweetened
beverages 1.17 pesos por liter
Fosil fuels liquid gas 8.98 8.98 cents per liter
High-calorie food 8%
Gambling 30%
Telecommunications 3%
Source: Excise Tax on Production and Services Law (LIEPS) 2018
100
Appendix 2.3: RIF vs. RESICO
Table 1. Differences between the Standard PIT, RIF and RESICO Regimes
Standard RIF RESICO - Individuals Standard RESICO – Legal entities
Personal Corporate Tax
Income Tax Regime - Legal
Regime entities
 Individuals who exclusively carry out entrepreneurial Individuals who exclusively carry Legal entities Legal entities resident in
Individuals activities, sell goods or render services that do not out entrepreneurial and resident in Mexico incorporated with
obtaining require a professional degree and with annual professional activities, or who grant Mexico individual partners or
Eligibility income from turnover below the threshold the temporary use or enjoyment of shareholders, whose total
any source.  Entrepreneurial activities under a joint ownership, goods (including professionals), income in the immediately
subject to certain requirements. provided that their annual turnover preceding fiscal year does not
 Income from salaries, rent or interest. is below the threshold. exceed the threshold.
Threshold to None Net annual income of up to MXN 2 million Gross annual income of up to MXN None Gross annual income of up to
qualify 3.5 million MXN 35 million.
Net income (total income less authorized deductions and statutory Gross income on cash basis Net income on Net income (total income less
Tax base employees’ profit sharing) on a cash basis accrual basis purchases of merchandise and
raw materials) on cash basis
Net annual income (MXN) Marginal Rate Gross annual income (MXN) Rate Rate Rate
0 – 6,942 1.9% Up to 25,000 1.0%
6,942 – 58,922 6.4% 25,000 - 50,000 1.1%
58,922 – 103,550 10.9% 50,000 - 83,333.33 1.5% 30% 30%
103,550 – 120,373 16.0% 83,333.33 - 208,333.33 2.0%
Tax rates 120,373 – 144,119 17.9% 208,333.33 – 3,500,000 2.5%
144,119 – 290,668 21.4%
290,668 – 458,132 23.5%
458,132 – 874,650 30.0%
874,650 – 1,166,200 32.0%
1,166,200 – 2,000,000 34.0%
100% discount during the 1st year, 90% during 2nd year, Accelerated depreciation for
Exemptions and further 10% reduction in discount every year after new investment if below MX 3
and for up to 10 years. million.
Deductions Exemptions: Pensions below 15 minimum wages; wage workers bonuses, incentives, prizes, and profit sharing.
Deductions: medical, educational, and funeral expenses. Insurance premiums, contributions to retirement
accounts, donations, real interest expenditure on mortgage loans.
Withholding Wages and Legal entities that carry out operations with
101
salaries. individuals in RIF must retain 1.25% of the
amount of the payments made to them
Table 2: Tax obligations and exclusions
RIF RESICO - Individuals RESICO – Legal entities

Tax  Tax returns must be filed on a bimonthly basis. If this  Tax returns must be filed on a monthly  Monthly tax payments must be made no later
obligations requirement is not met or in case the tax return is filed after basis and an annual tax return must also than the 17th day of the following month.
the deadline two consecutive times, or three times within a be filed in the month of April of the  Annual tax calculations follow the same basis
term of 6 years as of the first time that it failed such a following year. of the general corporate tax regime,
requirement, the taxpayer may not continue paying taxes  If the monthly tax return is not filed three considering cumulative income, authorized
under this regime. consecutive times within a calendar year deductions, profit and tax result.
 Save receipts that meet tax requirements or the annual tax return is not filed, the  Keep accounting in accordance with the Fiscal
 Provide tax receipts to clients. taxpayer may not continue paying taxes Code.
 Submit bimonthly income and expense reports. under this regime.  Issue, deliver, obtain and keep the CFDI’s of
 Register or update their Federal Taxpayers the total income, expenses and investments.
Registry.  Formulate statement of financial position and
 Generate their advanced electronic raise inventories.
signature and an active tax mailbox.  Present the unpaid balances of the loans as of
 Issue digital tax receipts online (CFDI) for December 31 of the previous year.
all of the income actually collected.  Keep a record of the operations carried out
 Obtain and keep digital tax receipts online with securities.
that cover the expenses and investments.  In the case of payment of dividends, provide
 Issue and deliver to its clients digital tax the tax receipts that indicate the amount of
receipts over the Internet. income tax withheld.
 Report on loans and capital increases
Exclusions  Partners, shareholders or members of legal entities, or  Partners or shareholders of legal entities;  Legal entities that any of their partners or
related parties of taxpayers who have previously been subject  foreign residents with one or more shareholders have control or participate in
to this regime. This exception is not applicable to the establishments in the country; and another company or when they are related
following taxpayers, which may opt to apply the  those who have income from preferential parties, and
incorporation regime: regimes.  those who carry out activities through trusts
o partners or shareholders of non-profit organizations, in or joint ventures
the case that they do not receive distributable profits;
o individuals that are partners, shareholders or members
of entities dedicated to the administration of saving
funds, and savings and loans cooperatives; and
o partners, shareholders or members of sports associations
that pay taxes as legal entities, in the case that the
102
partners, shareholders or members do not receive
income from the legal entities of which they form part.
 Taxpayers who carry out activities related to real estate,
real estate capitals, real estate business or financial
activities, except in certain cases;
 Commissioners, agents or brokers;
 Taxpayers deriving income from public performances or
franchisers; and
 Taxpayers carrying out activities through trusts or joint
ventures.
 Taxpayers who sell goods or render services through digital
platforms, only for the income received from these
activities.
103
Appendix 2.4: Distributional Impact of the Introduction of the Simplified Regime for
Small and Medium Enterprises66
To estimate the distributional impact of the reform, we simulate taxes under all regimes and assume
workers minimize their tax liabilities. We use the household survey data (2020 ENIGH) 67 and apply the
Personal Income Tax legislation to simulate tax liabilities at the individual level, including all exemptions,
allowances, and deductions. We first estimate a baseline scenario which computes the tax liabilities for
each person in the household in 2020, assigning them to the RIF or the standard PIT regime, depending
on their eligibility. We then estimate tax liabilities under the three regimes available in 2022 under the
reform scenario, including the standard, RIF, and RESICO regime, depending on their eligibility. We
assume taxpayers minimize their liabilities and assign them the regime that minimizes their tax
payments. We then compare the number of taxpayers likely to stay in the same regime and those
switching to the RESICO regime. Finally, we estimate the distributional impact of personal income taxes
before and after the reform by looking at measures of progressivity, their marginal contributions to
poverty and inequality (defined as the change in poverty and inequality before and after personal
income taxes).
As much as 28 percent of personal income taxpayers are expected to switch to RESICO, leading to a
decline in tax collections. Keeping the total number of taxpayers constant, we find that 19 percent of
previous RIF taxpayers and 46 percent of standard rate taxpayers would switch to the RESICO regime
(Table A2.4.1). While more than half of new RESICO taxpayers would be moving from the standard rate
regime, 23 percent employers previously contributing under the RIF would also make the change. The
overall impact on observed tax collections is estimated to be an 18 percent decline, largely from self-
employed individuals previously contributing under the standard regime who now qualify for RESICO.
Table A2.4.1. Taxpayers and Tax Collections under Baseline and RESICO Reforms
66
Source: Inchauste and Baquero (2022): The Distributional Impact of the Introduction of the Simplified Regime for Small and
Medium Enterprises and Inchauste (2022): Formalization of SMEs and the Introduction of a Simplified Regime in Mexico.
67
The household survey captures 2.5 million PIT taxpayers in 2020, or 86 percent of the total reported in tax administration
data. However, total tax collections estimated with the household survey are only 50 percent of what is reported by
administrative records. The survey is especially bad at the higher end of the income distribution, leading to a lower share of tax
collections being observed. The result is that the pre-tax level of inequality is underestimated under the baseline. However, to
the extent that both the RIF and the RESICO regimes target small and medium taxpayers, the estimated impacts of the reform
likely point in the right direction. Finally, note that this approach ignores distributional impacts that arise from the change of
regime for legal entities and firms, and the potential impacts on wage workers, either through changes in employment or
wages.
104
Total Personal Income Taxpayers Total Collections (MX$ million)
Baseline RESICO Baseline RESICO
2020 reform Change 2020 reform Change
Total 2,565,178 2,565,178 0 $ 30,051 $ 24,650 -18%
RIF 1,718,182 1,394,324 -323,858 $ 1,343 $ 690 -49%
Self-employed 741,037 641,744 -99,293 $ 590 $ 281 -52%
Employers 977,145 752,580 -224,565 $ 753 $ 410 -46%
Standard PIT 846,996 453,241 -393,755 $ 28,708 $ 22,102 -23%
Self-employed 352,839 78,723 -274,116 $ 5,436 $ 669 -88%
Employers 494,157 374,518 -119,639 $ 23,272 $ 21,432 -8%
RESICO 717,613 717,613 $ 1,858 100%
Self-employed 373,409 373,409 $ 897
Employers 344,204 344,204 $ 961
Source: World Bank Fiscal Microsimulation Model using ENIGH 2020.
Taxpayers transitioning into RESICO have higher profit margins than those choosing to stay in the RIF.
As expected under the simulation scenario, taxpayers with entrepreneurial activities and low profit
margins will prefer to stay under the RIF regime (which are based on net incomes or profits), while
taxpayers with high profit margins will prefer the much lower rate on their gross revenue in the RESICO
regime. This is especially true for self-employed workers with entrepreneurial activities but is also the
case among employers (who generally tend to have higher costs).
Most contributors expected to switch to the RESICO regime belong to the top quintile of the pre-tax
distribution. While many taxpayers in the middle of the pre-tax distribution will prefer to stay in the RIF
program, taxpayers who are in the top quintile of the pre-tax distribution are more likely to switch to
the RESICO regime, particularly those with entrepreneurial activities that previously did not qualify
under the RIF. As a result, about 38 percent of RESICO contributors are expected to belong to the top
quintile of the observed pre-tax distribution.
As a result, the RESICO reform is expected to have a small negative distributional impact. The
incidence of PIT is expected to decline slightly for higher income individuals, while a higher share of total
PIT collections would come from lower income deciles (Figure A2.4.1). This could signal that a potentially
higher share of PIT collections would rely on withheld taxes on wages and salaries. As a result, the
reform would slightly reduce the progressivity of the PIT regime in Mexico as shown by the Kakwani
coefficient (Table A2.4.2). Moreover, the PIT’s redistributive impact would slightly decline, as the
marginal contribution to inequality reduction falls by 0.1 percentage point. This effect is likely the result
of (i) the higher eligibility threshold, (ii) the fact that professional self-employed workers who previously
did not qualify for the RIF and contributed under the standard regime will now qualify under the RESICO
regime, and (iii) the ability of self-employed workers with entrepreneurial activities to choose the tax
regime they belong to (at least initially), which is a privilege not afforded to wage workers.
Table A2.4.2. Distributional impact of the reform
Marginal Marginal
Contributio Contribution to
Kakwani
n to Poverty Inequality
reduction reduction
105
PIT baseline 0.284 -0.013 0.019
PIT under
RESICO 0.283 -0.012 0.018
Figure A2.4.1. Incidence of Personal Income Taxes before and after the reform
Taxpayers beginning in the 4th decile will see a A higher share of total PIT revenue will rely on
small decline in their tax liabilities as a share of lower income deciles
their pre-tax income
0.10 0.70
0.08 0.60
0.50
0.06
Share of market income + pensions
0.40
0.04
Share of total taxes paid

0.30
0.02
0.20
0.00
Poorest 2 3 4 5 6 7 8 9 Richest 0.10
-0.02 0.00
Market income + pensions deciles
-0.10
PIT baseline PIT under RESICO PIT baseline PIT under RESICO

The recently approved small taxpayer reform is expected to lead to a small decline in revenue and a
small negative distributional impact in the short term. The reform is expected to lead many individuals
switching to the new simplified regime. Over 700,000 taxpayers (28 percent of all personal income
taxpayers) are expected to switch to the RESICO regime. Taxpayers in the new regime are expected to
have higher profit margins than those deciding to stay in the RIF regime. Moreover, most contributors
expected to switch to the RESICO regime belong to the top quintile of the pre-tax distribution. To the
extent that taxpayers had time to assess which regime minimized their liabilities, the reform could lead
to a decline in collections and a small negative distributional impact. The negative distributional effect is
likely the result of a higher eligibility threshold and professional self-employed workers qualifying under
the RESICO regime. The distributional effects are small because wage workers, who contribute 90
percent of all personal income taxes, do not have access to the new regime. Further analysis on the
potential impacts on informality and on the long-term distributional impacts of the reform will be
needed.
Inchauste (2022) finds that 5.5 percent of previously informal self-employed workers would
become formal with the recent reform, potentially leading to an increase in tax collections in
the medium term. Among self-employed workers with entrepreneurial activities and self-employed
service providers for whom the data allow a calculation of tax liabilities. This behavioral response is
significant, but it must be put in context as it would affect less than one percent of all employed
workers. However, abstracting from other changes in the economy, it could amount to a 2.1 percent
decline in informal employment. Firms that switch to formal status show that these firms are larger and
106
potentially more productive than those who switch to informality. To the extent that the firms switching
into formality are more productive, this could lead to a small increase in revenue collections,
counteracting the reduction in tax collections estimated in the short run. The results also point to other
constraints to formalization, including the difficulty in complying with other regulatory requirements and
other incentives to formalization. Policies across the government will need to be coordinated to avoid
counteracting forces to this effort.
107
Chapter 3: Intergovernmental Transfers
Introduction
1. Mexico’s fiscal federalism framework play a critical role for the finances of subnational
governments (SNGs) and the provision of public services. As in most federal systems, major tax bases –
i.e., personal and corporate income taxes and consumption taxes— are more efficiently administered at
the federal level, whereas SNGs are often better equipped to provide public goods and services that
meet the needs of local populations. The resulting combination of centralized tax collection and
decentralized public service delivery —known as “asymmetric decentralization”— creates vertical fiscal
gaps in federal systems worldwide.
2. The extent of asymmetric decentralization varies from country to country. Large fiscal gaps
tend to be present in countries with highly centralized revenue systems and decentralized service
provision. In many emerging economies, expenditure responsibilities of SNGs far exceed the revenue
potential of their assigned tax bases. Such disparities between expenditure needs and own-source
revenue capacity are bridged by transfers from the federal government to SNGs and, in some instances,
SNGs’ borrowing. Decentralization ratios for OECD countries are depicted in Figure 3.1.
Figure 3.1: Decentralization ratios in OECD countries, 2019 and 2020
80
70
60
CAN
50 GBR
AUS
40
DEU
CHE
30
ICL FIN SWE DNK
BEL
20 CZE
LVA MEX
FRA PRT
POL
ITA
10 NORSLK ESP AUT
ISR
HUN USA NLD
LUX
CRI GRCIRL SLV EST
LIT
0
0 10 20 30 40 50 60 70 80
108
70
60
Tax decentralization > CAN
50
spending decentralization
CHE
AUS
USA
40
DEU
SWE
30 ICL FIN ESP
DNK
BEL
20 CZE MEX
FRA LVA NOR
POL
ITA
SLV AUT
10 PRT GBR
ISR
HUN NLD
CRI GRCIRLLUX SLK EST
LIT
0
0 10 20 30 40 50 60 70
Source: OECD Fiscal Decentralization database

3. Intergovernmental transfers play a critical role in Mexico’s fiscal federalism framework due to
both the size of the vertical fiscal gap and the size of the gap between expenditure needs and fiscal
capacities of individual SNGs (horizontal fiscal gap). Regional differences in socioeconomic conditions
create horizontal fiscal gaps, as SNGs vary in terms of their expenditure needs and ability to generate
own-source revenue. Countries that face large horizontal fiscal gaps often adjust intergovernmental
transfers to reduce disparities in the provision of public goods and services across SNGs.
Intergovernmental transfers can also be used to advance national social and economic objectives or to
boost spending on public goods and services that generate positive spillovers across subnational
jurisdictions.
4. Mexico has one of the largest vertical fiscal gaps among comparable countries worldwide.
Mexico’s SNGs are now responsible for a large share of total spending, while they collect a small portion
of total tax revenues, resulting in one of the largest vertical fiscal gaps among OECD countries (Figure
3.1). Large regional disparities also create acute expenditure needs in Mexico’s less-developed regions,
where SNGs’ own-source revenue capacity is particularly limited: intergovernmental transfers finance
more than 95 percent of subnational spending. To advance national development objectives, the federal
government earmarks a large share of federal transfers for key social sectors, such as education, health,
public security, and basic social infrastructure. Finally, capital transfers finance a large share of public
infrastructure investment given the limited access to credit markets among certain SNGs, especially for
local governments.
5. Intergovernmental transfers are designed to attenuate vertical and horizontal fiscal gaps,
broadly equalize public service provision across regions, and advance national development goals, but
the heavy dependence of many Mexican SNGs on federal transfers creates serious challenges. While
the optimal vertical fiscal gap for each federal system is difficult to determine, the analysis presented in
this chapter suggests that large fiscal gaps and low levels of subnational own-source revenue weaken
incentives for expenditure efficiency and undermine incentives for SNGs to strengthen own-source
revenue collection, creating a cycle of dependence.
109
6. Recent evidence (World Bank 2022) suggests that Mexican states show no evidence of
convergence, with southern states falling behind. The poorest states in Mexico, like Chiapas, Guerrero
and Oaxaca are growing more slowly than the national average, leading to regional divergence.
Interventions in key areas are needed to tackle regional inequality and promote growth in lagging
regions: (i) reforming intergovernmental fiscal relations to reduce horizontal imbalances; (ii) raising the
municipalities’ capacity to mobilize own revenues through property taxes, which would enhance local
capacity to provide public goods; (iii) improving governance of land, by phasing out restrictions on
agricultural land ownership and transfer, while strengthening rural income support and access to
finance; (iv) addressing poor infrastructure; (v) enhancing regulatory quality; and (vi) fostering human
capital accumulation and strengthening incentives and support to raise the contribution of universities to
regional knowledge and innovation.
7. The following note examines Mexico’s intergovernmental transfer system, assesses its fiscal-
equalization effects, and offers recommendations to improve its efficiency and effectiveness. The note
is organized as follows. Section 2 describes the intergovernmental transfer system. Section 3 depicts
Mexico’s subnational borrowing. Section 4 concludes the analysis and presents some policy options to
improve the effectiveness of the intergovernmental transfer system, SNGs’ revenue collection with
emphasis on property taxes, and subnational borrowing.
Mexico’s Intergovernmental Transfer System at a Glance
Institutional framework
8. The current fiscal federalism framework in Mexico has been around for over 40 years. It is
structured around the National Fiscal Coordination System (Sistema Nacional de Coordinación Fiscal, or
SNCF). Its objective is to establish and coordinate the fiscal federal framework, outline specific
responsibilities in terms of decision making between the different government levels and channel
budgetary resources. This is done by: (i) setting the transfer of competencies and responsibilities in
terms of levying taxes between the federal government, states, and municipalities; (ii) implementing the
decentralization of federal public resources to subnational entities; and (iii) establishing the formal
channels of coordination for public finance matters between the three different levels of government.
9. The transfer of fiscal competencies between the federal government and subnational
governments (SNGs) results in transfers from the federal government to SNGs. Under the Fiscal
Agreement in the Constitution of Mexico, SNGs transfer part of their tax sovereignty to the federal
government. As a result, SNGs receive both non-earmarked transfers and earmarked transfers
(participaciones, aportaciones68 and convenios de descentralización de recursos federales 69).
Participaciones are non-earmarked transfers that complements own-resources from SNGs. Aportaciones
are earmarked transfers set by the Fiscal Coordination Law to achieve a specific objective. Finally,
Convenios could be earmarked or non-earmarked, with an objective and time horizon that are agreed
between the SNGs and the federal government.
10. The legal framework in Mexico encompassing the federal fiscal coordination system as well as
activities related to revenue collection, spending, subnational borrowing, and monitoring and
reporting is comprehensive. This includes laws and regulations issued by different layers of the
government: federal, state, and local. At the federal level, it includes articles 115, 116, 117, and 122 of
the Constitution. In addition, there are seven general/federal laws, which are complemented by
legislation and regulation issued by the states and local governments. This includes: (i) Fiscal
68
These are for education, health, security, public investment, and financial restructuring.
69
These are for the objective of: “reasignar la ejecución de funciones, programas o proyectos federales y, en su caso, recursos
humanos y materiales”
110
Coordination Law (1979); (ii) The Federal Budget Law (2006); (iii) General Public Accounting Law (2008);
(iv) Fiscal Responsibility Law for Subnational Governments (2016); (v) Federal Oversight and
Accountability Law (2016); (vi) Federal Revenue Law (annual); and (vii) Federal Budget Law (annual). 70
11. At the state and local level, the legal framework includes the state constitutions and
legislation and regulations issued by the states and municipalities. Each state has a constitution that
sets the overarching legal framework for both the state and comprising municipalities of such state.
Each subnational legislature has the authority to enact specific subnational public finance legislation –
aligned with the principles stated in the national Constitution, with rules and regulations that are
particular to each subnational government. In practice, these differences are not large. Similarly, local
governments can also issue public finance regulations. However, the scope is very limited given that
most revenue collection competencies, budget execution and oversight are already governed by public
finance regulation issued at the federal and state level. Even for the case of taxes that pertain
exclusively to local governments, such as property taxes (“predial”), state level legislation regulates all
municipalities in each state. Each of the 2,458 municipalities of Mexico issue their own annual revenue
and budget laws approved by their local governments, in which they adjust, in many instances, the tax
bases and tax rates and other sources of own revenues.
12. Under the fiscal responsibility law, SNGs are mandated to report the overall budget balance as
well as the budget balance of available resources to monitor 71 and evaluate the budgetary
performance of SNGs.72 Through these two balances, the goal is to monitor the trajectory of the
earmarked spending and the non-earmarked spending of state and local governments and establish the
underlying spending pressures and imbalances. The overall budget balance is defined as the difference
between total revenues included in the Revenue Law, and total spending under the Budget Law of the
EEFF, excluding debt amortizations. In other words, it looks at all revenue collection and executed
spending over the fiscal year, without a distinction on whether these are earmarked or non-earmarked.
The FRL defines budget balance of available resources as the difference between non-earmarked
revenues, including those in the Revenue Law, plus net financing, and non-earmarked spending under
the budget law, excluding debt amortization. In other words, it is the difference between non-
earmarked revenues73 and non-earmarked spending of the SNGs, without accounting for net financing
that has been secured using components of non-earmarked revenues.74
The transfer system
13. Intergovernmental transfers can be classified into four types according to their policy
objectives: (i) revenue-sharing transfers used to close the vertical fiscal gap; (ii) equalization or
70
These are: i. La Ley de Coordinación Fiscal (1979); ii. La Ley Federal de Presupuesto y Responsabilidad Hacendaria (2006); iii.
La Ley General de Contabilidad Gubernamental (2008); iv. La Ley de Disciplina Financiera de las Entidades Federativas y los
Municipios (2016); v. La Ley de Fiscalización y Rendición de Cuentas de la Federación (2016); vi. La Ley de Ingresos de la
Federación (anual), y vii. El Presupuesto de Egresos de la Federación (anual).
71
FRL imposes a reporting mandate, with EEFFs required to publish these indicators periodically. Prior to 2016, it was not
mandatory for the states to report the budget balance of available resources. Information to construct such indicators was,
however, available.
72
Balance Presupuestario de Recursos Disponibles
73
Non-earmarked revenues are resources subnational governments have full authority to determine where to spend those
resources. Their origin and estimated amount are established each year in the revenue laws, enacted by the local legislatures.
They are comprised of i) own resources from each EEFF; (ii) participaciones from the federal government; (iii) resources from
financing which payments derive from non-earmarked revenues. Own-resources represent a small share of total revenues for
SNGs. They are comprised of (i) Impuestos locales o estatales; (ii) Cuotas y Aportaciones de Seguridad Social; (3) Contribuciones
de Mejoras; (4) Derechos; (5) Productos, y (6) Aprovechamientos.
74
Non-earmarked spending is all spending contained in the approved annual Budget Law by state legislatures, which does not
include spending that is paid using resources coming from earmarked transfers by a norm or regulation at the federal level.
111
compensatory transfers used to reduce the horizontal fiscal gap; (iii) conditional or earmarked transfers
used to finance expenditures that advance national development objectives or generate positive
spillovers across jurisdictions; and (iv) capital transfers used to overcome subnational credit constraints
that would otherwise inhibit public investment. Some countries have also established regional
development funds designed to promote economic growth in underdeveloped areas. Some funds
provide equalization transfers that are earmarked for specific expenditure categories (a hybrid of type
(ii) and (iii)).
14. Mexico’s fiscal federalism framework includes all four types of intergovernmental transfers
(see Figure 3.2):
 Revenue-sharing transfers: The largest source of revenue-sharing transfers
(participaciones) from the federal government to state governments is the General Fund for
Revenue-Sharing Transfers (Fondo General de Participaciones, or FGP). The FGP is financed by 20
percent of all federal tax revenue and federal income from extractive industries. States, in turn,
are required to transfer at least 20 percent of FGP resources to municipalities, and each state
establishes its own allocation criteria. In addition, the Revenue Collection and Auditing Fund
(Fondo de Fiscalización y Recaudación) receives 1.25 percent of projected federal sharable
revenue (recaudación federal participable); the Fund for the Promotion of Local Governments
(Fondo de Fomento Municipal) receives 1 percent; and the Fund for Municipalities Located along
the Border or in Coastal Areas (Fondo para Municipios Colindantes con la Frontera o los
Litorales) receives 0.136 percent, which it allocates according to the value of international trade
processed at seaports and border crossings in each municipality. Additional revenue-sharing
mechanisms also apply to specific federal taxes and other revenue streams, including excise
taxes, taxes on gasoline, taxes on the oil industry, customs duties, and income taxes paid by civil
servants.
 Equalization transfers: In addition to covering the vertical fiscal gap, the revenue-
sharing transfers described above are also intended to address disparities in expenditures across
jurisdictions. The federal revenue shared through the FGP is primarily collected in Mexico’s
most-developed regions, and less-developed regions tend to receive more than they contribute.
However, the FGP is not designed to achieve full fiscal equalization, and many of its sub-
transfers have other policy goals.
 Conditional transfers: Conditional transfers (aportaciones) finance sector-specific
spending by SNGs. These transfers are distributed through eight funds under the Ramo 33
budget line, the largest of which finance education, health, and public security services provided
by state governments. Conditional transfers also include discretionary transfers under Ramo 23,
which subsidize state programs that contribute to national objectives. Finally, federal resources
are channeled to state governments through individual agreements (convenios) negotiated on a
case-by-case basis.
 Capital transfers: Capital transfers finance investments in basic social infrastructure by
state and local governments. These transfers are conditional and distributed according to
equalization criteria. The subsidies provided under Ramo 23 and the resources transferred
through individual agreements with state governments can also be used to finance capital
spending.
Figure 3.2: Taxonomy of Mexico’s Intergovernmental Transfer System
112
Revenue sharing Equalizing Conditional Capital
Aportaciones (Education Basic Social

Fondo General de Fondo General de (FONE), Health (FASSA), infrastructure for state
Participaciones Participaciones Public Security (FASP), (FISE) and municiplaities
others under Ramo 33. (FAIS)
Other revenue-sharing
schemes (Fomento
Municipal, oil revenues,
tax collection of Subsidies under Ramo 23
individual taxes, etc)
Agreements
Source: Fiscal Coordination Law, SHCP
15. The total value of Mexico’s intergovernmental transfers rose from 7.2 percent of GDP in 2005
to 7.7 percent in 2021—equal to about half of all federal revenue or 90 percent of SNG revenue . The
continuous growth of revenue-sharing transfers reflects a steady increase in federal tax revenue over
the period, while the expansion of conditional transfers has been driven by the decentralization of
service delivery. Conditional transfers are not automatically linked to federal revenue and can, in
principle, be adjusted by policymakers. In practice, however, a large share of earmarked transfers is
effectively mandatory, as their level and allocation are highly inertial. Nevertheless, conditional transfers
have declined since the federal government launched its current fiscal consolidation effort in 2015. The
federal authorities exercise a greater degree of discretion over the subsidies provided under Ramo 23
and the decentralization agreements, both of which contracted after 2014 (Figure 3.3).
16. The multiple objectives of the FGP and other revenue-sharing transfers distinguish Mexico’s
fiscal federalism framework from more conventional intergovernmental transfer systems. In a
conventional system, vertical and horizontal fiscal gaps are addressed by two different instruments—
pure revenue-sharing transfers distributed according to devolutionary or derivation criteria and pure
equalization transfers designed to reduce disparities in expenditure capacity after accounting for the
fiscal impact of revenue-sharing transfers—and neither mechanism is used to create incentives for SNGs
or advance policy goals beyond its core purpose. In Mexico, however, revenue-sharing transfers are
used to address both the vertical and horizontal fiscal gaps, to stimulate regional economic growth, and
to incentivize own-source revenue collection by SNGs. 75 In addition, the numerous small transfers
financed by federal sharable revenue are designed to achieve a wide range of objectives, which further
dilute the fundamental revenue-sharing and fiscal-equalization goals of Mexico’s transfer system.
Figure 3.3: Intergovernmental Transfers in Mexico, 2005-2021
75
This is also a feature of fiscal transfers in other federal republics in Latin America, including Argentina and Brazil.
113
10
Health Insurance
8.8 8.9 8.9
9 8.6
8.1 8.2 8.3 8.2 8.3 Conditional - Agreements
7.9 8.0 7.9 7.8 (Convenios)
8 7.7
7.2 7.3 7.2
Conditional - Subsidies
7 Ramo 23
6 Conditional (Aportaciones)
- Other Ramo 33
% of GDP
5
Conditional (Aportaciones)
- Social Infra
4
Conditional (Aportaciones)
3 - Health
2 Conditional (Aportaciones)
- Education
1 Revenue sharing (Partic-
ipaciones)
0
Total
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: SHCP, INEGI
17. The excessive complexity and inconsistent purposes of Mexico’s revenue-sharing transfers
prevent them from contributing to a single, unified policy objective; as a result, they do not effectively
address vertical or horizontal gaps, accelerate regional economic growth, or incentivize subnational
tax collection. Revenue-sharing transfers were introduced in 1978, when the central government
replaced many indirect state and local taxes with a national value-added tax. The initial purpose of the
transfers was to compensate SNGs for the loss of own-source revenue, and the original formula
distributed resources based on the amount of federal revenue collected in each state. 76 Revisions to this
formula in the early 1990s shifted its focus from devolution toward relative population size and own-
source tax collection, effectively introducing a fiscal-equalization component into the allocation criteria.
However, the new formula had only a weak equalizing effect, as it did not directly incorporate fiscal
capacity or expenditure needs. In 2007, the distribution formula was revised again, and resource
allocation is now based on the growth rate of states’ GDP (a devolutionary component), relative
population size (an equalization component), and the growth rate of the state’s own-source revenues
(an incentive to improve tax effort). In addition, a “hold harmless” clause was included to ensure that no
SNG may receive fewer resources than it received before 2007. Due to this clause, the revised formula
retains elements of both the original formula’s devolutionary component and the population-size
criterion from the 1990 reform. The complex structure and inconsistent objectives of Mexico’s revenue-
sharing transfers have produced a system that is neither devolutionary nor equalizing and that creates
only weak incentives for SNGs to support local economic growth or strengthen own-source revenue
collection.
18. The current distribution formula allocates the increase in FGP resources since 2007 at close to
an equal per capita basis, which heightens its equalization effect but undermines incentives to boost
economic growth or increase tax effort. Although the formula is designed to reward increases in state
economic activity (proxied by federal tax revenue from each state) and encourage own-source revenue
collection at the state level, improvements in economic activity or tax effort above the national averages
generate only a small increase in revenue-sharing transfers, as population size determines the bulk of
the allocation.77 The distribution of the post-2007 increase in FGP resources on a simple per capita basis
76
A compensatory fund provided additional resources to states that received less than the average per capita revenue-sharing
transfer.
77
Herrera, 2012.
114
further dilutes the fund’s incentive structure. When the rising trajectory of FGP transfers renders the
hold-harmless clause obsolete, zeroing out the weight of the pre-2007 distribution, the FGP will come
close to being an equal per capita transfer for most states.
19. Conditional transfers were established in 1998 to finance the decentralization of basic
education services, healthcare, and social infrastructure. Unlike traditional conditional grants, which
are typically used to encourage SNGs to allocate more resources to certain public goods or services than
they otherwise would, Mexico’s conditional transfers were created to finance the decentralization of
staff and facilities from the federal government to SNGs. Consequently, the allocation criteria for
transfers under Ramo 33 reflect supply-side considerations rather than demand. Focusing on inputs
rather than outcomes weakens the efficiency of subnational service provision. Moreover, because the
criteria for conditional transfers were defined when services were decentralized, they have a strong
inertial component, and cost considerations are often absent from their allocation formulas.
20. Mexico’s most important capital transfer is distributed through the Transfer Fund for Social
Infrastructure (Fondo de Aportaciones para la Infraestructura Social, FAIS). The FAIS represents 2.5
percent of federal shareable revenue and is by far the largest financing source for capital investment at
the municipal level.78 The FAIS could also be described as an equalization transfer, since its resources are
allocated to states and municipalities based on poverty indicators. However, the FAIS formula combines
state and municipal indicators, which attenuates its equalizing effects, as a poorer municipality in a
relatively wealthy state would receive fewer resources than a similar municipality in a poorer state.
21. Ramo 23 is a more flexible budget line consisting of discretionary transfers, the size and
beneficiaries of which are determined by the federal government and approved by Congress.
Resources from Ramo 23 are distributed to subnational governments through specific funds, including
regional-integration projects, the metropolitan fund, the culture fund, and a border-areas fund, among
others. Decentralization agreements between federal ministries and state governments finance the
implementation of federal programs at the state level.
Main stylized facts of the current system of transfers
Large vertical fiscal imbalances
22. The current intergovernmental finance system exhibits one of the largest vertical fiscal
imbalances (VFI) in the world. Mexico has one of the largest vertical fiscal gaps among comparable
countries worldwide. Due to the centralization of revenue collection at the federal level –which has its
origins in the 1978 Pacto Federal, combined with the decentralization of expenditures at the state level,
Mexico’s SNGs are now responsible for a large share of total spending, yet they collect very little tax
revenue on their own, resulting in one of the largest vertical fiscal gaps among OECD countries.
Equalization Impact of Intergovernmental Transfers
23. Mexico exhibits substantial regional disparities in socioeconomic development. Differences
are especially large between the industrialized north and center-north and the less developed south
(Figure 3.4 and Figure 3.5). In 2020, the country’s wealthiest federative entity, Mexico City, had a GDP
per capita of MXN376,205, over six times that of the poorest state, Chiapas, at MXN60,671. Per capita
GDP in Nuevo León, one of Mexico’s richest states, was close to the level of Poland, whereas Chiapas
had a per capita GDP similar to that of Honduras or Bhutan. Poverty and other socioeconomic indicators
follow a similar pattern. Extreme poverty in 2020 was high in Chiapas (44.1 percent), Guerrero (34.8
percent), Oaxaca (28.7 percent), Tlaxcala (27.7), Puebla (26.2), and Veracruz (24.4 percent).
78
FAIS transfers are disbursed through two funds: (i) the Fund for State Social Infrastructure Transfers (Fondo de Aportaciones
para la Infraestructura Estatal, FISE), and (ii) the Fund for Municipal Social Infrastructure Transfers (Fondo de Aportaciones para
la Infraestructura Social Municipal, FISM). FISM accounts for about 88 percent of FAIS transfers.
115
Figure 3.4: GDP per capita by state Figure 3.5: Poverty by state
50
494,341
2018 2020
40
376,205
30
301,562
20
266,321
251,007
213,618
213,096
212,672
210,491
210,322
210,102
191,267
190,943
190,581
10
180,664
173,663
167,154
157,025
153,951
146,162
143,092
131,085
122,655
119,547
119,500
117,278
117,339
114,471
107,863
0
91,830
85,567
85,225
60,671
B Agu uer co
Nu Sinal a
Tlax ebla
evo oa
GMuéxi a
Gu aca
de Sonor t
Mo éxico
Oaxcala
aca algo
Tab atán
s
Qu mp sco
ari
im
Yucelos
go
Chirero
apa
as
Coa aja Ca ascali étaro
na e
Q Jalis
Nauya
Dulipas
au o
n
inta ech
Ciaol
ran
tec
Pu
a
Tam juat
Roo
o Hid
Leó
r
uah
la d rnia s
oMsí
er
hui lifo ente
orn
ragC r
ancao
e Za Su
P ot
ohziha
alif
a
Z
C
mp
nousis
aC
lave
Oca
xicna L
Baj
dad
MeSa
la L
d e
Ciu
de
cán
os
acio
h oa
nid
Ign
Mic
sU
de
ado
uz
s co
Hid bla
a
Gu xaca
dad vo a
arit
Est
Mo xico
Pueala
algo
G uran n
Oa as
Agu Calif sco

Tam Colim
Naylos
Ciu Nue Sonor

alo
Qu uanaju go
acr
Tlax ro
Yuccas
D atá
Baj Tabaaro
he
Qu uahua
h s
ap
na o
c
Jíali
de León
erre
Chi aulipa
re
o Mé
Sin
Ver
inta at
Roo
Coa aja Ca scalie rnia
pec
e
hui lifo ntes

Ch i
t
eré
tos
e acat
rag r
Caxmico
e Za Su
o
oz a
ss Po
Llav Z
la d rnia
Mé
mp
Lnoui
a
B a
Oca
xiacna
de
MeS
e la
cán
io d
os
h oa
nid
nac
Mic
sU
e Ig
ado
zd u
Est
acr
Ver
Source: INEGI, CONEVAL
24. Large disparities in per capita spending levels among state and local governments can be
observed across Mexico. Such disparities in per capita spending are reflected in access to basic services.
The share of the population with lack of access to healthcare, education lags, or without access to basic
services, water and sanitation is higher among southern states (Figures 3.6-3.9). Socioeconomic
disparities are also reinforced by limited own-source revenue mobilization by SNGs. Because wealthier
states tend to have much larger tax bases and more efficient tax administrations, their own-source
revenue capacity often far outstrips that of their poorer counterparts (see below).
25. Intergovernmental transfers in the current system mostly play a devolutionary role: they do
very little to reduce the horizontal fiscal gap and attenuate existing regional socioeconomic disparities
across SNGs. Differences in expenditures per capita between better off and worse off states are large,
which translates in unequal access to basic public services among the population, depending on where
they live in the country. The current system of transfers (including Participaciones or revenue sharing,
Aportaciones or conditional grants, and Convenios or Agreements) falls short to properly equalize across
subnational governments. Individually and in the aggregate, transfers appear to be regressive: relatively
richer states with higher own revenues per capita generally tend to receive higher per capita transfers.
This is not only highly inequitable, but it is also economically inefficient, incentivizing fiscal (as opposed
to economic) migration and retarding economic growth by underinvesting in human capital. 79 Horizontal
disparities also are also present in terms of capital infrastructure across the states. The current system
of capital grants is not only fragmented, leading to management and planning inefficiencies, but also
due to its size, it has fallen short of closing the existing infrastructure gaps across states, and thus
fostering regional economic convergence.
Figure 3.6: Total SNG Spending per Capita (2020, Figure 3.7: Share of the population without
in current MXN) access to healthcare services, 2020 (%)
79
Preliminary analysis, based on panel regressions for 2012-2019, suggests that “participaciones” and “aportaciones” are
regressive, whereas “convenios” are fairly neutral.
116
30,000
25,000
20,000
15,000
10,000
5,000
-
an bla
mp a
an ca
Gu aloa
Tab ora
Mi lifor o
Ba C Méx rit
Nu Yuca la
Mo xico
ua ica go
San ev tán
Gu Pue as
s M Hid ila
C co
ja C Jali s
ate o
eré z
Du rrero
e
d d Nay as
Co taro
orn hua
Ca olim
Ve cán
Qu O ipas
Méato
is P ón
a sc
o
Qu racru
ch nia
Sinoo
xca
Zac rang
au osí
Tlates
ech
int axa
a
iap
Sour
al
o
u
rel
as
n
Lu Le
ja C hih ic
ah
aR
oa
aju
alif a
ia S
liens
Tam ot
l
e
Ch
u
scano
e
Ag ex
Ba
da
Ciu
ido
Un
os
ad
Est
Non-earmarked spending Earmarked spending
Figure 3.8: Share of the population without Figure 3.9: Share of the population with
access to basic services, 2020 (%) education lags, 2020 (%)
Source: INEGI, SHCP, World Bank staff calculations
26. Among these transfers, the revenue-sharing transfers are devolutionary rather than equalizing
and they tend to magnify disparities in state revenue. The amount of federal revenue-sharing transfers
received by each state broadly correlates with its level of economic activity. SNGs with higher levels of
per capita GDP receive more in revenue-sharing transfers than do their poorer counterparts. The six
states that contribute the most to national GDP—Mexico City, State of Mexico, Nuevo León, Jalisco,
Veracruz, and Guanajuato—are also the states that receive the largest shares of revenue-sharing
transfers (Figure 3.10). While conditional transfers and other transfers (i.e., Ramo 23 subsidies and
individual agreements) are all regressive, revenue-sharing transfers are substantially more regressive
than conditional transfers. As all of its components are regressive, the overall intergovernmental
transfer system is also regressive.
Figure 3.10: Shares of GDP and Revenue-Sharing Transfers by State, 2019-2020
117
GDP 2019 Participaciones 2019 GDP 2020 Participaciones 2020
18 18
16 16
14 14
12 12
10 10
8 8
6 6
4 4
2 2
0 0
la
Tla ima
cán Tama ebla
Tla ima
Ve lisco
V lisco
ra
Qu Hid loa
Yu aca
u a
Tab aloa
Ch aca
Co rit
Co rit
la
la
evo ico
evo ico
an o
Oa lgo
Ch tán
sca tán
Ag Gue as
Ag Yu as
is P o
Qu asco
orn los
ja C Zac relos
ja C M ango
Mo ngo
uil Guan cruz
Co Ba na uz
sca ero
an o
Na as
orn as
Oc che
Hidche
San Ta taro
mp o
So ua
So ua
int alg
Tam Pueb
cán Cam lipas
as
xca
xca
Ba Zara ato
uila al to
Lu basc
Ja n
Ja n
no
no
int rrer
ya
ya
Ca rétar
ihu ia
Zar ia
iap
iap
Oa o
DuRoo
a
Du tes
Qu Guetes
Sinsí
Sinsí
Gu eracr
Nu Moéx
Nu Moéx
a
c
alif atec
Leó
Leó
x
x
alif ore
l
ca
u a ca
de ulip
ah ja C jua
ah
ah
o
ua rr
ZiaacSur
Nuar
ate
Ch iforn
de iforn
ra
pe
e
oto
oto
ra
e aju
al za
ré
ihua
aR
P
c
c
r
lien
lien
upeo
is P o
ia S
u
Chagoz
éxi
éxi
a
Luamp
ja C go
aQm
a
eM
eM
San Oc
dd
dd
de
ad
da
da
Ciu
Ciu
Ba
Ba
ah
oa
oa
Co
ch
ch
Mi
Mi
Source: INEGI, Estadísticas Oportunas (SHCP)
Low levels of tax effort at the subnational level

27. The system is characterized by low levels of subnational tax collection. Subnational tax
revenue collection is low relative to other OECD countries (Figure 3.11). In 2020, SNGs collected less
than 1 percent in tax revenues, of which state and municipal taxes totaled 0.7 percent and 0.3 percent,
respectively. There are also large disparities in own-source revenue collection between SNGs in the
north and center-north and SNGs in the south.
Figure 3.11: Share of tax revenue, by level of government – OECD countries
100% Taxing power of SNGs, 2018 (% of GDP)

90%
State/province Local SNG revenue
80%
18
16.5
70%
15.4
Share of tax revenue, by level of government
60% 16
50%
14
12.4
40%
12.0
11.3
30% 12
10.3
10.3
20%
9.6
10
10%
8.7
8.5
0%
7.4
8
6.5
ly
6.2
e
Lit e
y
ia
da
Ne xico
m
en
Co ia
l
ark
Re a
De y
6.1
bia
Ne tuga
Ita
rke
nc
eec
d
s
lan
str
ni
an
al
c
lgiu
na
nd
ed
an
5.7
bli
Fra
nm
ua
lom
str
rm
Me
Tu
Gr
Au
Fin
5.4
Ca
Sw
rla
eal
r
Be
pu
6
h
Po
Au
Ge
the
wZ
4.9
4.5
4.5
ech
Cz
3.4
4
2.5
2.4
2.3
2.2
2.2
1.9
1.8
1.7
1.7
1.4
1.4
2
0.9
0.9
0.6
0.5
0.4
0.4
0.3
0
ITA
EST
ISR
IRL
ICL
LTA
T
F
SLV
FIN
A
A
HU T
DN E
L
SLK
A
JP N
ESP
E
L
S
R
L
L
R
R
C
N
R
D
X
K
N
BE
CZ
ZA
NZ
PR
PO
SW
AU
AU
CH
CH
LV
FR
US
ME
BR
LU
TU
KO
GR
GB
NO
DE
CA
NL
Central State Local
Source: OECD
28. Low subnational revenue collection reflects a limited tax base, limited institutional and
technical capacity, weak collection efforts, and disincentives introduced through other transfers and
agreements. Major tax bases such as the personal income tax, the corporate income tax, and broad-
based indirect taxes are administered at the federal level, mostly due to economies of scale and
efficiency. States and municipalities are left with minor tax bases, including payrolls, motor vehicle
registration, lodging taxes, property taxes. However, even when considering revenue from these taxes,
revenue collection performance is below regional peers and OECD countries (Figure 3.11).
118
29. The most important source of tax revenues for the local governments are property taxes .
Mexico’s revenue collection in property taxes (predial) is modest at best, reflecting limited institutional
and technical capacity, and outdated property registries (cadastros). In 2019, Mexico collected 0.3
percent of GDP in revenues from property taxes. Low collection of property taxes considerably reduces
own resources of municipal governments. Despite several transfers that try to incentivize tax collection
effort by subnational governments, the positive effects are dissipated because of powerful disincentives
to collections introduced through other transfers and agreements, such as the Fondo de Estabilización,
Ramo 23 and several ad hoc decided Convenios. These transfers set a de facto “soft budget constraint”
for the states incentivizing perverse behavior of collecting less, borrowing more, and spending more.
30. A reform of property taxes can be a simple and efficient way to increase tax collection for
local governments. First, property taxes have a relatively stable tax base since the value of real estate
does not fluctuate as much as the income of companies or individuals. Second, it is a progressive tribute
since the amount of the tax increases according to the commercial value of the property. Third, since
the property tax is collected by the local government, the execution of public spending translates into
goods and services at the local level, which should decrease the resistance of taxpayers to pay the tax
(Unda and Moreno, 2015). Furthermore, property taxes are less harmful for growth than taxes on
incomes: recurrent taxes on immovable property are considered least distortionary relative to
transaction taxes (Arnold et al., 2011).
Table 3.1: Intergovernmental Transfers and Fiscal Effort, 2000-2016
(1) (2) (3) (4)

Fiscal Fiscal Fiscal Fiscal
effort effort effort effort
Total Transfers (% total revenues) -0.03 *** -0.03 ***
0.01 0.01
Participaciones (% total revenues) 0.03 0.03
0.02 0.02
Aportaciones (% total revenues) -0.10 *** -0.10 ***
0.02 0.02
Others (% total revenues) 0.00 0.00
0.02 0.02
Debt (% GDP) 0.05 0.00
0.05 0.05
Debt ( % IDL) 0.00 0.00 **
0.00 0.00
Living in urban areas (% population) 0.00 0.00 *** 0.00 ** 0.00 **
0.00 0.00 0.00 0.00
Intercept 0.39 *** 0.38 *** 0.54 *** 0.54 ***

0.04 0.03 0.05 0.05
Observations 448 448 448 448
F test 6.52 6.23 9.16 9.18
F test all ui =0 415.56 440.78 314.44 340.41
Overall R-sq 0.09 0.11 0.26 0.27
*p<0.10, **p<0.05, ***p<0.01
Source: World Bank staff calculations
31. In addition, a high degree of dependence on intergovernmental transfers weakens incentives

for SNGs to increase their own-source revenue collection. The international evidence shows that SNG
tax effort is negatively correlated with the share of intergovernmental transfers in their total revenue. 80
A stochastic frontier analysis using panel data for all 32 federative entities from 2000-2016 confirms that
80
See Blanco (1996) for Brazil, World Bank (2016) for Peru, and World Bank (2018) for Argentina.
119
the negative correlation between SNG tax effort and the share of intergovernmental transfers in total
revenues holds true for Mexico (Table 3.1 and Figure 3.12).81
Figure 3.12: Own-Source Revenue, Fiscal Capacity, and Tax Effort among SNGs (2003-2016 average)
$12,000 70%
Own Revenues Pesos Per Capita
$10,000 60%
50%
% Own Revenues Effort

$8,000
40%
$6,000
30%
$4,000
20%
$2,000 10%
$0 0%
r
es
o
a
e
ón
án
Su
la
a
a
o
la
n
go
Ro
hu
ch
at
lo
tá
im
nt
sc
ic
ca
eb
Le
ac
al
a
ju
na
pe
éx
ba
ca
ua
ie
a
ni
ax
id
Pu
ho
Co
na
vo
an
M
al
Yu
Si
Ta
m
or
ih
Tl
ic
Ba asc
ue
ua
nt
Ca
Ch
lif
ui
N
G
Ca
u
Ag
Q
ja
Own Revenues Fiscal Capacity Effort (right)
Source: World Bank staff estimates
Low levels of spending efficiency

32. The system suffers from low levels of spending efficiency. This is largely due to the high vertical
fiscal imbalance and the lack of accountability and responsibility for much of the funds spent. The lack of
spending efficiency is also reinforced by the design of several participaciones and aportaciones transfers,
which introduce perverse incentives. For instance, among the Participaciones, the “Fondo de
Participaciones por el 100% de la Recaudación del ISR” is a matching grant from the federal government
to the states to hire and expand the number of their public employees (the more public employees are
hired, the larger the transfer). In the case of the Aportaciones, the current conditional grants for health,
education and other sectors are distributed with formulas based on capacity (supply side) measures
(number of clinics, hospital beds, number of teachers etc.) as opposed to being distributed in terms of
needs, such as the size of the client base. Such distribution rules provide clear perverse incentives to
overspending and building excess capacity.
33. In summary, Mexico’s intergovernmental transfers system does not effectively devolve fiscal
resources to the regions where they were collected, mitigate regional disparities in the provision of
public goods and services, promote policies that accelerate local economic growth, or encourage
greater tax effort among SNGs. The devolutionary component of the transfer system outweighs its
equalizing component, and thus transfers do little to offset horizontal fiscal imbalances or improve the
overall equity of public spending. However, the equalizing component is strong enough to undermine
the relationship between local economic growth and the amount of revenue-sharing transfers received
by each SNG, preventing transfers from effectively encouraging pro-growth policies. Meanwhile, the
81
Estimating tax capacity and tax effort is a three-step process. First, stochastic frontier analysis is used to estimate each state’s
tax capacity. The tax effort of each state is then calculated as the ratio between observed and potential tax collection. Finally,
the tax effort is regressed to account for the share of transfers in total revenue.
120
large share of transfers in state revenue weakens incentives for SNGs to boost own-source revenue
collection.
Subnational borrowing
34. Mexico’s subnational debt framework was reformed in 2000 to support the development of
markets for SNG borrowing. The revised regulatory framework for SNG debt included: (i) the
establishment of Master Trust Funds (MTFs), which enabled SNGs to borrow against future income from
revenue-sharing transfers and provided a secure source of funds for SNGs to repay debt; (ii) addressing
supply-side constraints, which limited the exposure of private creditors to SNG debt, with the
introduction of subnational credit ratings into capital-risk weighting formulas for bank loans, enabling
lenders to more accurately assess idiosyncratic subnational risks; and (iii) the mandatory registration of
subnational loans with the SHCP and their conformity with financial transparency requirements.
35. The establishment of MTFs significantly reduced credit risks and expanded the market for SNG
borrowing during the 2000s. MTFs’ regulations define common debt-contracting and repayment
procedures, under which federal revenue-sharing transfers are deposited in the MTFs via an irrevocable
instruction to the federal treasury. The use of MTFs greatly reduced the likelihood that a state
government might attempt to use federal revenue-sharing transfers to finance current spending rather
than servicing debts contracted through the MTF. In addition, credit ratings and supply-side regulations
reinforced the market-based framework for subnational borrowing. By 2010, all 32 federative entities
and about 50 municipalities had received credit ratings from at least one recognized rating agency, and
thus subjecting SNG finances to growing surveillance and scrutiny by private markets and rating
agencies.82
36. While these reforms facilitated the robust expansion of the market for SNG lending, important
systemic weaknesses were still present:
 Incentives for lenders to focus on MTFs rather than on the borrower’s fiscal situation:
Given the critical role of the MTFs in ensuring debt repayment, the fiscal position of SNGs was
given limited attention when assessing credit risk, with creditors focusing mostly on the legal
strength of the MTFs while thorough assessments of the fiscal situation of SNGs were relegated
to a complementary role or even dispensed with altogether. Thus, SNGs had little incentive to
improve their fiscal positions to obtain credit on better terms.
 A disconnect between credit risks and borrowing costs: Relying on federal revenue-
sharing transfers to fund MTFs –and debt service repayments by SNGs— did not fully eliminate
the repayment risk premia for MTF-secured subnational debts relative to federal debt. The cost
of borrowing for SNGs continued to be subject to high risk premiums.
 The proliferation of trust funds and the excessive commitment of federal transfers and
other forms of non-earmarked revenues (NER) to finance debt service: Encouraged by the
success of MTFs in expanding access to private credit, SNGs and creditors created new trust
funds for specific operations. These funds were secured by future state tax revenue or NER
inflows, and in some cases, by the use of already committed federal revenue-sharing transfers
to an existing debt liability.
 Weak fiscal management by SNGs: The limited availability and reliability of fiscal data,
inadequate fiscal reporting by public-sector agencies, and the absence of mandatory registration
for short-term SNG obligations and other subnational liabilities, including those arising from
82
The National Banking and Securities Commission (Comisión Nacional Bancaria y de Valores, CNBV) also regulates SNG lending
operations by banks based on credit ratings and expected losses.
121
public-private partnerships (PPPs), prevented an accurate assessment of the financial situation
of SNGs.
37. The stock of subnational debt remains low and do not pose a systemic risk. Yet several
Mexican SNGs have experienced episodes of fiscal distress, including rising debt levels, shrinking fiscal
space, and liquidity problems. These episodes have threatened their ability to provide essential services,
invest in public goods, foster economic growth, and promote the welfare of their populations. The stock
of SNG debt amounted to 1.5 percent of GDP in 2005. In the wake of the global financial crisis, the
Mexican economy suffered a sharp contraction that negatively affected subnational fiscal and debt
dynamics. Between 2008 and 2013, faced with rising spending obligations and a limited capacity to raise
own-source revenues, SNGs increased their borrowing to finance widening deficits. The aggregate SNG
debt stock rose from 1.7 percent of GDP in 2008 to 3.0 percent in 2013, when it peaked and declined
thereafter. As an aggregate, SNGs have been running a surplus between 2016 and 2020, contributing to
a steady decline in the level of indebtedness both as a share of GDP as well as a share of non-earmarked
revenues. In 2016, 20 states had debt-to-non-earmarked revenues above 50 percent, and among those
8 had levels above 100 percent. By 2019, that number declined to 17 and 4, respectively, pointing to a
general improvement in debt dynamics over recent years. Aggregate subnational debt stood at 2.5
percent of GDP in 2021.
Figure 3.13: SNGs debt (current MXN), 2005- Figure 3.14: SNGs debt (% of GDP), 2005-2021
2021
800,000 3.5
700,000 3.0
600,000
2.5
500,000
2.0
Millions of current MXN
Percent of GDP
400,000
1.5
300,000
1.0
200,000
100,000 0.5
- 0.0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
States Municipalities Total States Municipalities Total
Source: INEGI, SHCP Source: INEGI, SHCP
Figure 3.15: SNGs aggregate fiscal balance (% of Figure 3.16: SNGs aggregate debt (% of non-
GDP) earmarked revenues)
122
71.8
67.8
0.20 65.1
61.9
58.5
0.15
0.12
0.09
0.06
2016 2017 2018 2019 2020 2016 2017 2018 2019 2020
Source: SHCP Source: SHCP
Figure 3.17: Spending and public investment by SNGs

10 9.43 9.35 0.60
% of GDP (spending by SNGs)
% of GDP (public investment)

9.01 9.02 8.80
9
0.50
8 0.50
0.49 0.48
7
0.40
6 0.39
0.36
5 0.30
4
0.20
3
2
0.10
1
0 0.00
2016 2017 2018 2019 2020
Non-earmarked spending Earmarked spending

Total spending Public investment
Source: SHCP
The Fiscal Responsibility Law for Subnational Governments

38. To ensure the sustainability of SNG finances and enhance fiscal planning and expenditure
execution at the subnational level, a new institutional framework for controlling SNGs indebtedness
was adopted with the approval of the Fiscal Responsibility Law for Subnational Governments (Ley de
Disciplina Financiera de las Entidades Federativas y los Municipios of 2016, LDFEFM). In 2015, a
Constitutional amendment was passed in Congress to authorize federal regulation of SNG borrowing. 83
This was followed by the approval of the LDFEFM the following year. Complementary regulations issued
in 2016 and 2017 further elaborated the new institutional framework for regulating SNG indebtedness.
The new regulatory framework for SNG debt established hierarchical controls on subnational
indebtedness as well as monitoring and control mechanisms, which complemented the market
mechanisms developed in the 2000s. The framework consists of: (i) the LDFEFM itself; (ii) its three
complementary regulations: the “traffic light” system, a public debt registration system, and regulations
for contracting debts at the lowest cost; and (iii) access to guaranteed federal debt (deuda federal
83
Per the provisions of the Mexican Constitution, this amendment required the approval of Congress and the legislatures of all
32 federative entities.
123
garantizada) to be used in debt-restructuring operations between SNGs and creditors under fiscal-
adjustment agreements.
39. To ensure the fiscal sustainability of SNGs, the LDFEFM includes a fiscal-balance rule that
imposes financing caps according to a set of debt thresholds defined in the “traffic light” system.
Based on three debt indicators—the debt-to-NER ratio, the debt-service-to-NER ratio, and the short-
term-obligation-to-total-revenue ratio—the “traffic light” system classifies SNG indebtedness as either:
sustainable (green), in observation (yellow), or high (red). 84 Each classification imposes a ceiling on net
financing, which is equivalent to the overall balance. For SNGs with sustainable levels of indebtedness,
the ceiling is set at 15 percent of NER; for SNGs with debt levels in observation, the ceiling is set at 5
percent; and for highly indebted SNGs, the ceiling on net financing is set at zero, which implies a
balanced budget (Figure 3.18)85. Because these ceilings are inversely related to debt levels, compliance
with the “traffic light” system is expected to automatically prevent excessive indebtedness.
Figure 3.18: LDFEFM Fiscal Rule and the “Traffic Light” System
Source: World Bank staff
40. The LDFEFM also establishes the following rules for SNG fiscal management: (i) a “golden
rule”: borrowing should finance productive public investment rather than current spending; (ii) a cap on
the annual growth of personnel expenditures equal to either the real GDP growth rate or 3 percent in
real terms, whichever is lower; (iii) a provision that when actual revenues exceed the levels projected in
the budget law, the excess should be used to finance anticipated debt repayments or accumulated in a
natural disaster fund; (iv) a provision that when actual revenues are below the level projected in the
budget law, spending on non-productive expenditure items should be cut; and (v) an escape clause that
enables the authorities to temporarily suspend the application of the fiscal rules, which can be triggered
in the event of a natural disaster or a severe economic slowdown.
41. The system for recording subnational debt is designed to enhance transparency by providing
detailed, timely information on SNG debt stocks, borrowing operations, interest rates, maturities, and
debt-service flows. The system’s governing regulations oblige borrowers and creditors to register all
borrowing operations in an SHCP database. The system also contains provisions for registering other
SNG liabilities, including PPP obligations and associated contingent liabilities. A standardized record of
the public debts of federative entities helps ensure that taxpayers, firms, subnational officials, and the
84
The debt to NER ratio has a predominant role in the final rating in case of different ratings for each of the three indicators.
SNG indebtedness indicators and classifications are updated quarterly. Classifications corresponding to the second quarter
define the financing cap for the next year’s budget.
85
The LDFEFM establishes a transition period until 2020 in which SNGs that had deficits higher than the ceiling will gradually
reduce them until reaching the net financing ceiling.
124
federal government have access to reliable, up-to-date information on the debt obligations of each state
and municipality, and it enables all stakeholders to make decisions based on the same information. The
data recorded in the system are also used to calculate the “traffic light” indicators described above.
42. To enhance competition, reduce borrowing costs, and promote transparency in lending
operations, the revised framework also regulates how SNGs contract debt. These regulations mandate
that SNGs contract debt through competitive auctions with a minimum number of participants, and that
they include guidelines for defining borrowing costs, selecting offers with the lowest costs, and
publishing both the details of the offers received and the outcome of the selection process (Figure 3.19).
43. The LDFEFM allows for the possibility of debt restructuring, provided that a fiscal-adjustment
agreement is in place. The law defines the conditions for using federal guarantees linked to fiscal-
adjustment agreements to support debt restructuring. To obtain a federal guarantee, a SNG must sign a
fiscal-adjustment agreement with the federal government that defines multiyear targets for the fiscal
balances, spending levels, and own-source revenue generation. Moreover, the LDFEFM requires highly
indebted SNGs to have their fiscal-adjustment plans approved by the SHCP. Under these conditions,
federal guarantees for debt-restructuring operations are intended to enhance the federal government’s
ability to positively influence the fiscal performance of SNGs and reduce borrowing costs. 86
Figure 3.19: Steps and Guidelines for Contracting Credit Operations
Source: SHCP
44. The LDFEFM also contains regulations to improve budgeting, public financial accounting, and
fiscal planning. First, budgets must observe the limits on fiscal balances and expenditures imposed
under the LDFEFM. Revenues and expenditure arrears accrued during previous years must be used and
recorded in line with LDFEFM regulations. In addition, the law requires the budget to include an
assessment of contingent liabilities and the actuarial balances of social security systems, and an analysis
of the budgetary impact of any new policy initiatives. Second, SNGs are required to adopt harmonized
rules for public financial accounting that encompass the entire government sector as defined by the
SHCP and the National Council of Public Accounting. Finally, medium-term fiscal frameworks are to be
86
Since 2013, several highly indebted state governments have undertaken debt-restructuring operations with private banks
based on fiscal-adjustment plans agreed upon with the SHCP. As the indebtedness framework did not provide a formal role for
the federal government in these agreements, the participation of public banks served as an indirect way for the SHCP to reduce
restructuring costs and impose targets to strengthen the fiscal position of the states.
125
included in the annual budget laws, with a five- and three-year horizons for states and local
governments, respectively.
45. Overall, the adoption LDFEFM and ancillary regulations have strengthened the framework for
controlling and monitoring subnational indebtedness. The fiscal rule that underpins the LDFEFM is
simple and transparent, and it provides clear operational guidance for fiscal policy. The “traffic light”
system not only defines the financing ceilings for the fiscal rule, but also provides information to
creditors and taxpayers on the fiscal solvency of SNGs. Likewise, the public debt record system is
designed to offer reliable, regularly updated information to market participants. Regulations requiring
that debt be contracted under the best market conditions help improve transparency and lower
financing costs, and the rules regarding federal guarantees for debt-restructuring operations are
expected to enhance the federal government’s leverage over the fiscal performance of highly indebted
SNGs and reduce the frequency of debt restructuring. Finally, the LDFEFM also contains important
budget-preparation and execution rules, such as the mandatory inclusion of medium-term fiscal
frameworks in SNG budget laws, which is expected to improve fiscal planning and enhance expenditure
efficiency.
Main objectives of the reform of the transfer system and other considerations
46. Key objectives to be considered in a potential reform of Mexico’s intergovernmental system:
 Reduce the existing large VFI by providing state governments with significant tax
revenue autonomy: This will contribute significantly to improve the overall performance of the
system of intergovernmental finances by increasing overall spending efficiency, stimulating tax
revenue effort, and strengthening fiscal discipline in fiscal management and borrowing.
 Reduce horizontal imbalances through the introduction of a robust equalization grant:
The allocation formula of the grant would recognize the existing disparities in expenditure needs
and fiscal revenue capacity across the states. Capital grants should be consolidated and
distributed in an equalizing manner closing the existing gaps in basic infrastructure. All this will
significantly increase equitable access to basic services for citizens regardless of where they live
in Mexico, at the same time it will reduce inefficient fiscal migration and increase human capital
creation and prospect for growth.
 Eliminate perverse incentives in existing conditional transfers for inefficient spending,
lower tax effort and fiscal indiscipline. This can be achieved by the consolidation of many
transfers and the redesign of the distribution formulas for those left standing.
47. The reform should consider two overall constraints: (i) to be affordable and fiscally
sustainable; and (ii) to be politically feasible or acceptable by stakeholders.
 Affordability: the overall impact of the reforms on the federal budget can be calibrated
to be revenue-cost neutral. In addition, the reforms could also be designed to be tax burden
neutral for the taxpayers (i.e., no increase in individual taxation). Alternatively, the reforms
could allow for an increment in federal funds destined to SNGs, which could be accomplished by
allowing for an increase in overall tax burden, or by redirecting funds from other federal
spending programs. Allowing for a non-budget-neutral reform would facilitate the transition to
the new system, as part of the additional funds could be used to smooth out the transition
period to the new regime for states that are net losers vis-à-vis the current status quo. On the
other hand, increasing the tax burden or crowing out resources from other spending programs
may be difficult in a fragile post-COVID19 recovery environment or in a tight overall fiscal
situation at the federal level.
126
 Political feasibility: the reforms also need to consider the political economy effects for
winners and losers among the states and devise appropriate sequencing strategies to minimize
political opposition. Adopting a “hold harmless” strategy, so that no state would see less funds
than those received in the year previous to the reform, would help minimize opposition to the
reforms, but at the cost of elongating the transition period, distancing further into the future
the time for the full operation of the system.
Policy options
Transfer system
48. Intergovernmental transfers in Mexico are characterized by large vertical and horizontal fiscal
gaps, underscoring the need to improve their equity and enhance their efficiency, and strengthen
their contribution to the country’s development priorities. The current system’s most pressing
challenges stem from its excessive complexity and inconsistent policy objectives. The proliferation of
small revenue-sharing and conditional earmarked transfers has needlessly complicated Mexico’s fiscal
federalism framework, while the multiple objectives and conflicting allocation criteria used by individual
transfer mechanisms weakens their impact. Different elements of the system were designed to achieve
the incompatible goals of devolution and equalization, and while on balance federal revenue raised in
wealthier regions is redistributed to poorer ones, the distribution formula for the largest revenue-
sharing transfer falls far short of achieving interregional equity. Moreover, the allocation criteria for
most conditional transfers are based on supply-side factors rather than demand, and ultimately reducing
affect expenditure efficiency of SNGs. Heavy dependence of SNGs on federal transfers also creates
perverse incentives that nullify the impact of mechanisms designed to encourage own-source revenue
generation.
49. Reshaping the current array of revenue-sharing transfers into a single, pure revenue-sharing
transfer, and a single, pure equalization transfer would help better balance the conflicting objectives
of revenue devolution and fiscal equalization. This would also serve to ensure the transparent
devolution of fiscal resources, while enabling policymakers to calibrate the degree of horizontal
equalization. One option would be to split Participaciones in two main components: a devolutionary-
revenue-sharing mechanism and an equalizing transfer based on the horizontal fiscal gap to ensure
progressivity and efficiency.
50. The formula for the pure revenue-sharing component could be designed to ensure that
transfers reflect the amount of tax revenue collected within the jurisdiction of each SNG (i.e., a
derivation or origin-basis criterion). Most of the federal revenue that finances revenue-sharing
transfers comes from tax bases that are under the exclusive purview of the federal government (e.g.,
value-added tax and corporate income tax), but each state’s contribution to GDP can be used to
approximate its contribution to federal revenue. Basing revenue-sharing transfers on state-level
economic activity encourages state governments to adopt policies that boost local economic growth,
but the convoluted nature of the current formula obscures this incentive. States may also be more likely
to collaborate with the federal government to improve tax administration if a share of the additional
revenue will be transferred directly to SNGs.
127
51. The distribution criteria for a pure equalization transfer should capture differences in
expenditure needs87 and own-source revenue capacity88 across SNGs, the two fundamental
dimensions of the horizontal fiscal gap. Expenditure needs (not actual expenditures) are the amount
each SNG would need to spend to provide a standard level of public services based on its population
size, local socioeconomic conditions, and local public administration costs. Revenue capacity (not actual
revenue collection) is the ability of each SNG to raise own-source revenues based on an average level of
administrative effort, adjusted for the size of the government’s assigned tax bases and the funds
received through revenue-sharing mechanisms.
52. Estimated expenditure needs and fiscal capacity of each jurisdiction can be used to
approximate the horizontal fiscal gap. A positive fiscal gap indicates that local expenditure needs
exceed local fiscal capacity, and additional resources would be needed to make up the difference. A
negative fiscal gap implies that no additional resources would be required. Once the size of the gap has
been established, policymakers must determine how much of the fiscal gap will be covered by the
transfer system and define the criteria for allocating these funds among recipient SNGs. The criteria
should ensure that resources are transferred only to SNGs with positive fiscal gaps. Equalization
transfers need not eliminate subnational differences in expenditure needs and fiscal capacity altogether
(i.e., may or may not target full fiscal equality). Instead, a more limited equalization transfer may seek to
reduce such differences to a socially tolerable level. For example, the distribution formula could be
designed to achieve a minimum level of service delivery across all subnational jurisdictions. The more
ambitious the minimum level, the larger the equalization transfer.
53. The criteria for distributing equalization transfers should incentivize expenditure efficiency
and encourage (or at least not discourage) local revenue collection. Equalization transfers should
reflect the revenue capacity of SNGs and not their actual revenue levels, which could encourage
subnational authorities to neglect revenue collection. Similarly, assessments of expenditure needs
should be based on clearly defined quality standards and costs rather than actual expenditure levels,
which could encourage inefficient public spending. The numerous small revenue-sharing transfers
currently used to promote subnational tax collection could be consolidated into the equalization
transfer, as its allocation criteria would directly incentivize revenue effort by SNGs.
54. The allocation formulas for conditional transfers could be reformed by replacing the current
supply-side criteria with demand indicators. Conditional transfers for education, for instance, could be
distributed on an expenditure needs basis, taking into consideration local cost of service delivery.
Moreover, an equalizing factor could be included to accelerate the convergence of relevant indicators
across states.
 The Conditional Transfer Fund for Education Payroll and Operational Expenses (Fondo de
Aportaciones para la Nómina Educativa y Gasto Operativo, FONE) could be distributed with a
stronger emphasis on a per-student and needs-based allocation. The modified proposed formula
could reflect a core per-capita component that is adjusted for the costs of delivering a standard
87
Two methodologies to estimate SNGs’ expenditure needs involve: (i) estimating the amount of spending per capita necessary
to finance a core suite of programs and activities in each jurisdiction; or (ii) using a weighted index of relative need to establish
an average or common standard of service delivery and then dividing that amount by the number of beneficiaries. This would
require data on population, geographic area, poverty rates, and other factors that determine differences in spending needs to
create weighted indexes, which are then multiplied by the total population of each jurisdiction to obtain the adjusted
population. This methodology does not account for differences in the cost of service provision across jurisdictions. Whichever
technique is used, expenditure needs must always be estimated after accounting for the impact of conditional transfers, which
finance a large share of the public services delivered by Mexican SNGs.
88
Estimating fiscal capacity would require data on tax bases. The necessary information may not be available in all cases.
128
service across the country (favoring poorer regions). It could also include an equalizing factor
designed to accelerate the convergence of education indicators of lagging states.
 The Conditional Transfer Fund for Health Services (Fondo de Aportaciones para los
Servicios de Salud, FASSA) could be distributed on per uninsured person basis. While Seguro
Popular is gradually improving both the coverage of the uninsured population and the regional
equalizing impact of federal transfers to the health sector, new distribution criteria for FASSA
transfers may accelerate progress in these fronts. As a transition, the federal government could
distribute the same amount of resources in nominal terms in a given year (harm hold clause)
and the annual increase in the transfer could be distributed on equal per uninsured person
basis.
55. The government could reduce or re-shape existing discretionary transfers to subnational
entities such as Ramo 23. Ramo 23 resources could be transformed into a capital grant which can be
used by the federal government to finance investments by SNGs that have strategic national or/and
regional spillovers. Using formula based matching grants instead of ad-hoc transfers would enhance
accountability at the subnational level. Other options would include consolidating resources under
Ramo 23 and FAIS and/or creating a capital transfer with clear equalizing criteria.
56. Prospective adjustments to other conditional transfers could yield additional efficiency gains,
but designing effective reforms would require a more detailed analysis of each mechanism. Merging
FISE with FISM89 would help reduce the fragmentation of conditional transfers. FISE represents a very
small share of FAIS, and municipal governments implement most social infrastructure projects, which
mainly deliver local benefits with limited spatial spillovers. Merging the two transfers could yield a
modest efficiency gain without weakening their overall impact. However, the authorities would need to
simplify the two-stage distribution rule to prevent two municipalities with identical indicators from
receiving different amounts merely because they are located in different states.
57. Finally, Ramo 23 subsidies and individual agreements could be strengthened to prevent the
discretionary and nontransparent distribution of federal resources. For example, Ramo 23 subsidies
and individual agreements could be consolidated into FAIS to establish a comprehensive capital
transfers category as part of the updated fiscal federalism framework.
Main building blocks of the new system
58. Overall, main building blocks would be comprised of a new piggyback PIT tax, a new revenue
sharing system, a new equalization transfer, and a new capital transfer. 90 The piggyback PIT would
represent a new state tax in the form of a surcharge on Personal Income Tax. Revenue sharing
represents a new system of revenue sharing formula based on population. While the largest
aportaciones for health and education would remain unchanged in size, these transfers would allocate
resources according to reformed criteria. The equalization transfer refers to a newly designed transfer
aimed at limiting horizontal inequalities among states. Capital transfers will be designed to reduce
infrastructure gaps among the states and foster economic convergence.
59. The piggyback PIT is aimed at enhancing the revenue autonomy of SNGs. The new state level
piggyback PIT would give the states the discretion to select a flat rate between a minimum rate and a
maximum rate applied to the taxable income in the federal tax return. To avoid increases in taxpayers’
tax burdens, the federal PIT would be adjusted in rates so create fiscal space so to leave the new state
89
See supra note 4.
90
For illustration purposes, assuming a revenue-cost neutral reform to the federal budget (i.e., the overall pool of funds
available to carry out the reform remains within the budget envelope that currently represent the revenue sources for the
state).
129
piggyback PIT, not only as revenue/cost neutral to the federal government but also tax burden neutral
for taxpayers. The new state piggyback PIT would still be administered by the federal tax administration
authorities, but it would be a state tax because the states would have to determine the tax rate, getting
all its revenues (net of a potential administration fee paid to the federal tax administration for collection
costs). Negative state tax competition with a potential “race to the bottom” would be avoided by setting
minimum rates that states can apply.
Box 1. International experience with design of the piggyback PIT
International experiences regarding the actual structure of piggyback taxation varies along several dimensions.
Perhaps the most important is the variety of approaches to enhance the revenue autonomy of subnational
governments. Subnational governments may levy tax on the national government tax liability, taxable income,
or even just on wage and salary income. Although there are good arguments in favor of each of these
alternatives, there are several advantages of requiring subnational governments to levy tax on the national PIT
liability. Specifically, it allows subnational governments to mimic the progressivity of the national tax with a
single rate; it reduces the administrative and compliance costs of the subnational tax; and it ensures
coordination on the definition of the tax base between the subnational and national governments. An example
of how it could work in Mexico is as follows: (i) the rate setting authority for the state PIT is vested in the states;
(ii) the states levy a single sur-tax rate on the central government tax liability; (iii) the maximum sur-tax rate
would 25 percent and the minimum 15 percent; and (iv) the central government will repeal any PIT revenue
sharing with subnational governments, and reduce all PIT statutory (national) rates by 20 percent.
In this manner the single state sur-tax rate applied to the central government tax liability allows the state
governments to mimic the progressivity of the central government PIT tax rate schedule (approximately ranging
from 1 to 7 percent) and the combined effective tax rates (federal plus state) would be quite similar to what the
federal tax rate schedule currently is. Several other approaches are also possible, with the federal and state
governments applying the rates to a shared taxable income. For example, to the extent that a delta change in
overall taxes is feasible and desirable, the states could share 30 percent of the federal taxable income base, use
the same number of brackets and follow the same progressivity pattern as those of the federal government but
be allowed to implement PIT rates different and higher than the federal government ones (for some or all
brackets).
Source: Authors’ elaboration
60. The new state tax would be paid in the same federal tax form (as before), but now will have a
clear separate section for the determination of the state’s tax based on the net taxable income for the
federal tax component. The allocation of revenues will be based as much as possible on the place of
residence of taxpayers (information readily available in the federal PIT return). Being a state determined
tax, state revenue autonomy will increase, hence reducing VFI. This would help increase spending
efficiency, tax effort, and fiscal discipline. The revenue neutral constraint could be relaxed in the context
of a tax reform. The new state PIT could be integrated with the current state payroll tax, and thus
eliminate the well-known distortionary effects of that tax. This is so because the current state payroll tax
only covers certain sources of income that would also be covered with other sources on income in the
new state PIT tax. In addition, tax effort and collections with the new tax would be guaranteed, as it
would have a minimum or floor tax rate and collections would be enforced by the federal tax authority.
The new state piggyback PIT would amount to about 1.3-1.9 percent of GDP, depending on the design.
Under all reform options, the piggyback PIT would be created as part of “participaciones”. Such a design
would respect the spirit of the 1978 Pacto Fiscal between the federal and state governments. The new
state piggyback PIT could be considered a form of revenue sharing, but one that also offers states the
capacity to use their newly acquired tax autonomy to change the size of the funds received.
130
61. A single, unconditional,91 revenue sharing transfer would be distributed in transparent manner
of equal per capita basis. This simple distribution formula would have some equalizing effects in effect
distributing resources collected by the federal government from relatively richer to relatively poorer
states. It would be possible to consider still a simple formula but with stronger devolution effects—
benefiting those states that generate more revenues—using an index with two criteria: population and
state GDP. The higher the weight given to the second criterion, the higher the devolution and the less
equalizing the impact will be.92
62. Another interesting possibility of reforming the revenue sharing mechanism would be to set
apart some of the available funds to introduce a single purpose grant incentivizing revenue collection
effort at the state level. Although there is no good theoretical public finance basis for the states to
collect and spend more, there is an argument that can be made for stimulating tax effort in those
subnational governments that for historical or political economy reasons have lagged in terms of
revenue mobilization effort. One way to do that would be to offer a temporary transfers stimulus
proportional to the difference between the average state potential revenue effort and the actual
potential revenue effort of the state.
63. Conditional grants or “aportaciones” for education and health (about 2.5 percent of GDP)
would be preserved in terms of the size, but their distribution formulas would be amended.
Specifically, the reform would focus on abandoning supply or capacity factors-- to include only demand-
driven or need criteria.93
64. In sum, the proposed fiscal autonomy for the states in the form of a piggyback PIT tax may be
“financed” either through two mechanisms: from (i) the elimination of the state payroll tax (Impuesto
sobre nómina), resources of the PIT fund (Fondo del Impuesto Sobre la Renta) and the (proportion of) PIT
share in participaciones; or (ii) the PIT fund (Fondo del Impuesto Sobre la Renta) and the PIT share in
participaciones would be converted in a piggyback PIT, while no changes would be made to the payroll
tax. 94 Under all options, the residual participaciones (net of funds “used” to create piggyback PIT tax)
would be converted in a new revenue sharing mechanism. Aportaciones (net of those destined for
education and health) as well as funds from different types of convenios will be transformed into a new
equalization transfer and capital transfer.
65. A new robust fiscal equalization transfer would help limit horizontal fiscal disparities among
the states. While some “aportaciones” have an explicit equalizing objective, their impact on regional
91
Any current conditions in the use of revenue sharing funds would be removed.
92
Given the problematic current distribution of transfer resources, its regressivity, this alternative formula for revenue sharing
should only be considered if the new equalization grant discussed in what follows below is sufficiently powerful to offset those
regressive effects.
93
In 2014, the conditional grant for education FAEB was replaced by the Education and Operational Expenditure Contribution
Fund (FONE), but its modified distribution formula still includes a supply-side component (the operating costs of the teachers'
payroll), besides the demand-side component (the number of students in each state). Like the FONE, prior to the 2007 reforms,
the Health Services Contribution Fund (FASSA) distribution formula was based on supply elements, including the number of
decentralized federal health workers and operating expenses of federal health facilities that had passed for state
administration. While the FASSA also includes equalization criteria, the distribution of funds is still largely determined on the
basis of historical budget allocations, which in turn reflects the supply-side conditions that prevailed when health services were
decentralized. Because richer regions tend to have a larger number of health personnel, more qualified health personnel, and
better health facilities, the distribution of FASSA tends to replicate regional disparities in health services. But note that besides
FASSA there is the Universal Health Program, Popular Insurance, which distributes about 0.4% of GDP among states, based on
the size of its uninsured inhabitants.
94
Following initial interactions with the Government, the team would prepare simulations of the proposed reform scenarios. As
part of the calibration process, the proposed size of each element might change.
131
fiscal disparities is very limited (if any) at best. 95 The fiscal gap can be calculated by estimating
expenditure needs and fiscal capacity across the states. On fiscal capacity, the revenue potential of own
taxes (including the new state PIT) and revenue sharing would be included. 96 On expenditure needs, the
computation could include (or exclude) the needs for those services already financed with conditional
grants, such as education and health. The symmetrical inclusion of the needs and the conditional grant
funds into fiscal capacity would be justified if the conditional grant only finance a part of the actual state
needs in that sector. Only states with larger estimated needs than estimated fiscal capacity would be
eligible to receive equalization transfers, a feature that would increase the equalizing power of the
grant.97 The new equalization transfer would be quite sizable (up to 1.8 percent of GDP, depending on
the size of the capital transfer). The exact calibration of the size of the transfer would only be possible at
a later stage of the reform preparation and would also depend on the calibration of other elements of
the decentralized fiscal architecture. This would be affected by whether the revenue sharing formula
depends only on population, and not state-level GDP: the need for equalization might be reduced
compared to a scenario where state GDP plays a role in the revenue sharing formula (the size of the
piggyback PIT tax).
66. The size of the capital grants will need to be determined taking into account other elements of
the subnational fiscal architecture. The main contribution from the reform effort would be to amend
the distribution formula of capital transfers. One proposal would be to dedicate capital grant funds
exclusively to state infrastructure. Another option, replicating the recent reforms of the EU structural
funds, would also provide grants to companies for investments in lagging regions and seeking regional
economic convergence. The new capital grant transfer would apply a formula driven allocation with
strong equalization features, which would also contribute to the economic convergence of the states.
Two aspects of equalization would be highlighted. First, a component that addresses the equalization of
historical differences in the stock of capital infrastructure across the states, with the objective of
reducing it over the coming years. This would require taking a stock inventory of infrastructure facilities
across the states and dedicating a share of the available capital funds to help close some of the gap
every year. The second element would recognize the ongoing capital infrastructure needs of all states
as well as their capacity to finance those needs out of current revenues and ability to borrow. This
would involve developing a fiscal gap formula with the calculation of public infrastructure needs based
on demand factors and financing capacity.
Table 3.2: Intergovernmental Transfers in Mexico, 2016-2021
95
In particular, the Fund of Contributions for the Strengthening of Federative Entities (FAFEF) which distributes 1.4% of federal
taxes (or 0.2% of GDP) to the states according to the inverse of GDP per capita and therefore has a clear equalizing character.
However, because FAFEF's resources finance subnational government debt obligations, pension obligations, and institutional
and technical training, the fund has no direct impact on regional fiscal disparities.
96
The states currently use very little of their taxing powers and therefore calculating tax revenue potential will be a complex
task. Of the 18 taxes state can collect, on average only 6 of these taxes are actually collected, although the numbers vary by
state (Baja California collects most taxes (9) and Puebla the least (3)). The four taxes most collected by the states are the payroll
tax (accounting for close to three-fourths), lodging tax, lotteries, and vehicle tenure taxes. The three least collected taxes are
commercial activities, industrial activities and for mining. There is quite clearly a need for state tax reform, but that is topic
beyond the scope of this note.
97
The equalization power would be further enhanced if surplus states (those with larger fiscal capacity than expenditure needs)
were required to make additional contributions to the pool of foods (fraternal or Robin Hood contributions). Such type of
equalization feature, however, does not have much tradition in Mexico or in other Latin American countries.
132
Source: SHCP
Local government revenue collection: property taxes

67. The potential to raise property tax revenues is high. One of the limitations for property
collection in Mexico is related to the federal transfer system, which does not encourage municipalities
to increase tax collection. The ability to raise revenues from property taxes varies considerable across
municipalities, with Mexico City and municipalities in Quintana Roo collecting the most property taxes as
a percentage of the state GDP, and municipalities in the south collecting the least property tax. This
suggests that a coordinated effort between levels of government can lead to large increases in
collection.98
68. There are several factors behind the differences in terms of property tax revenues, across
municipalities in different states and within a state, including the quality of cadastral system that does
not enable the constant actualization of the fiscal cadasters, and low capacity at the local level . Many
municipalities have cadastral systems and fiscal cadasters with low level of technological sophistication.
This makes updates to the valuation of existing properties in the jurisdiction and keeping up with new
construction very challenging for many municipalities. Another issue undermining collection is local
98
In addition, the political cycle of the municipalities is 3 years.
133
capacities in many municipalities. Most municipalities in Mexico, particularly those in rural areas, do not
have the capacity to effectively administer a property tax system (Unda and Moreno, 2015).
69. Local capacity is a problem faced across the world, and international experience offers some
practical solutions, including a state (more centralized) administration of the property tax without
harming municipal autonomy for the tax. The experience gathered from larger countries and
federations is that a state level collection on behalf of the municipalities can have significant positive
effect in terms of collections, reduction of collection costs, and more efficiency, effectiveness and
sophistication in terms of fiscal cadasters, all due to economies of scale and capacity gains. The system
that has been employed in several Federations or large countries (Canada, Pakistan, Australia, India—
early stages)) could work in the following way in the context of Mexico:
 Municipalities (with few exemptions for large cities in the state) could sign an
agreement with states, for the administration of the property tax (predial) but sending back all
the revenues collected within the municipality back to the corresponding municipality minus a
small fee to maintain the overall system. The agreement keeps the tax autonomy granted by law
to the municipal government in terms of the tax. In fact, some states have this kind of
agreements already in a small or larger scale in Mexico (e.g., Nuevo Leon, Guanajuato). The aim
would be to rolling this across the country.
 States establish an entity for the purpose of establishing and maintaining a modern-
technologically advanced cadastral system and update regularly each fiscal cadaster across
municipalities. Some countries have called this entity “Municipal Revenue Board” (see figure Y).
 While the so called “Municipal Revenue Board” works as part of the state technical and
administrative apparatus, it has board where representatives of the municipalities participate to
ensure transparency and accountability. That municipal participation also ensures that policies
taken are aligned with the autonomy municipalities have on the tax, as granted by law.
 These Municipal Revenue Board would typically: (i) take joint decisions on policy issues
(base and rate, as permitted by the fiscal federal legislation), (ii) purchase and establish the
cadastral system with scale to attend all municipalities, (iii) based on that system, constantly (at
least annually) update all the fiscal cadasters of the municipalities in the system; (iv) provide
uniform guidelines and rules for re-evaluation; and (v) provide training and overall capacity
building for local/municipal staff directly involved in collection duties.
 In some countries, Municipal Revenue Boards, also do the collection of the tax itself
through municipal offices. But in other countries staff of the municipality are in charge of this
activity. At the end, the system of collection, recording and reporting is the same (with a
mainframe administered by the municipal revenue board at the state level). This decision is
mostly based on political and administrative/capacity considerations across states and
municipalities.
 Local municipal authorities help in the updating process of the overall central system, in
terms of new properties, changes in quality to properties, verifications, among other. This is also
supported by aerial/satellite survey updates provided, for scale, at the state level by the
municipal revenue board.
 The fees for maintenance retained by the municipal revenue board should be small and
for maintenance purposes only to avoid eroding trust.
134
 The initial investment could be financed by the states themselves or the Federal
government as part of a transfers for incentive purposes.
 The Federal government, through its transfers system, could also provide a fiscal
incentive to establish these Municipal Revenue boards at the state level. And at later states,
provide further incentives aligned with fiscal effort displayed by these authorities.
Table 3.3. Examples of large countries with state level collection on behalf of municipalities
Country Administration.
Assessment/Valuation Collection
POLAND Local government (AREA) Local government
Queensland, Australia State government (VALUE) Local government
New South Wales, Australia State government (VALUE) Local government
British Columbia, Canada Provincial government (VALUE) Provincial government
Ontario, Canada Provincial government (VALUE) Local government
England, UK Central government (VALUE) Local government
Netherlands Local government (VALUE) Local government
Slovenia Central government (VALUE) Local government
Czech Republic Central government (AREA) Central government
Estonia Central government (VALUE) Central government
Lithuania Central government (VALUE) Central government
Figure 3.20: MRB to run as a centralized, state-level, ICT-based property tax administrator
135
Subnational borrowing
70. The amended institutional framework for controlling SNG borrowing 99 has helped effectively
strengthen fiscal discipline among SNGs. Since 2014, borrowing controls from government-initiated
reforms led the adoption of the LDFEFM and improved fiscal balances have helped stabilize subnational
debt. This trend has been broadly consistent across jurisdictions, as most state governments have
reduced their fiscal deficits and debt ratios—underscoring the positive impact of tighter borrowing
controls. Moreover, the steady increase in current spending, which had contributed to the mounting
fiscal distress of SNGs, appears to have ceased in 2017.
71. Revising several aspects of the LDFEFM could further strengthen its ability to prevent
excessive indebtedness and improve fiscal management. These include:
 Adjusting the thresholds of the “traffic light” system. The current thresholds are too
high to be binding on the majority of the states, which weakens the system’s ability to prevent
SNGs incurring fiscal deficits and contracting excessive levels of debt, eventually leading to
liquidity or solvency problems.
 Reducing the number of regulations embedded in the LDFEFM. The law’s complexity
reflects an effort by the federal government to micromanage SNG finances, and its numerous
regulations complicate fiscal management, make compliance difficult to verify, and create
overlapping or inconsistent targets.
 Clarifying guidelines for using federal debt guarantees and fiscal adjustment
agreements. More explicit regulations regarding the concession of federal debt guarantees and
the design, implementation, oversight, and enforcement of fiscal adjustment agreements could
ensure that these tools are used effectively.
the level of indebtedness of SNGs. High thresholds create large intervals, and states with dissimilar
levels of debt obligations and liquidity indicators are often grouped together. The current thresholds are
too high and the “traffic light” system may not convey the appropriate signals. States are not classified
as highly indebted are not constrained by the fiscal rule, and thus, may continue to contract debt until
hitting the threshold of 200 percent of non-earmarked revenues (NER). This would undermine the core
objective of the LDFEFM.
Figure 3.21: State Indebtedness Levels under the “Traffic Light” System, 2020
99
The legal framework in Mexico forbids subnational governments from taking loans from foreign entities. The Federal
Constitution and the LDFEFM explicitly prohibit states and municipalities from contracting debt in any way, if it is due to pay
abroad, on foreign currency or if the lender is not a national. See Article 117 of the Constitution, and Article 22 of the LDFEFM.
136
Source: SHCP and World Bank staff
the level of indebtedness of SNGs. One of the goals of the LDFEFM was to control the growth of
borrowing and mitigate financial stress among subnational governments. High thresholds also create
large intervals, and states with dissimilar levels of debt obligations and liquidity indicators are often
grouped together. The current thresholds are too high and the “traffic light” system may not convey the
appropriate signals and restrains. States are not classified as highly indebted are not constrained by the
fiscal rule, and thus, may continue to contract debt until hitting the threshold of 200 percent of NER.
This would undermine the core objective of the LDFEFM. For instance, such high thresholds result in
only five SNGs in the categories “under observation” or “high” being restrained from incurring a fiscal
deficit and engaging in net borrowing in 2021.
74. Reducing the thresholds could enhance the signaling of the “traffic light” system and
strengthen the effectiveness of the fiscal rule in preventing excessive indebtedness among SNGs.
Halving the debt-to-NER thresholds and adjusting the debt-service and short-term debt indicators
accordingly (Figure 3.22) could significantly enhance the warning system and strengthen its impact on
SNGs’ debt dynamics.100 Under the lower proposed thresholds, 4 states (from none) would be classified
as highly indebted; 6 states would be classified as having a high debt service to NER; and 3 states would
be classified as having a high short-term obligations to revenue. In other words, fewer states would be
classified as having sustainable debt levels. The proposed adjustment would enhance the impact of the
LDFEFM’s fiscal rule on SNG debt trajectories, as a larger number of SNGs would face lower financing
ceilings. Lower proposed thresholds would enhance the “traffic light” system’s ability to discriminate
between different fiscal situations. The lower proposed debt thresholds would also reduce aggregate
debt at the state level. In sum, the lower proposed thresholds would reduce both aggregate
indebtedness at the state level and the number of states with debt levels above 100 percent of NER.
Figure 3.22: Current (LHS) and proposed (RHS) thresholds for the “Traffic Light” system
Figure 3.23: Indebtedness Classifications of States under the Current and Proposed Thresholds for
the “Traffic Light” System, 2021Q4
100
When the “traffic light” system was being designed, the World Bank and SHCP teams assessed several options for the
thresholds and initially agreed on the lower proposed thresholds described in this section.
137
Source: SHCP and World Bank staff calculations
75. While the LDFEFM provides a sound basis for controlling SNG borrowing and promoting
responsible fiscal management, some provisions of the LDFEFM may be difficult to comply with and
monitor. The federal government’s limited ability to observe and enforce compliance with the LDFEFM
— combined with SNGs’ variable levels of technical and institutional capacity, data constraints,
inconsistencies in public finance accounting rules and practices, and weaknesses in public financial
management, especially at the local level—could prevent the full implementation of the new
framework. This could weaken the LDFEFM’s enforceability and undermine its effectiveness.
 Weak public financial management and reporting among SNGs. Improvements in
SNGs’ financial accounting and fiscal reporting capacity will be critical to ensure compliance with
the LDFEFM and related regulations (budgeting, fiscal reporting, and expenditure planning,
among others). The General Governmental Accounting Law (Ley General de Contabilidad
Gubernamental) has improved accounting and reporting at the subnational level and increased
138
financial transparency. SNGs continue to suffer from deficiencies in data availability and data
harmonization, weaknesses in the registration of debt and arrears, and delays in the publication
of fiscal accounts. LDFEFM’s compliance and monitoring will require further improvements in
the government’s financial reporting and accounting systems.
 Soft budget constraints. The frequent discretionary use of federal transfers to close
financial gaps weakens SNGs’ fiscal discipline and may negatively affect the credibility of the
LDFEFM. Eliminating the discretionary use of extraordinary federal transfers could enhance the
credibility of the fiscal discipline framework for SNGs.101
 Structural problems. Due to the statutory division of tax bases between the federal and
subnational levels, SNGs have very limited capacity to mobilize own-source revenues.
Consequently, rising current-expenditure obligations are expected to undermine medium-term
fiscal sustainability. Subnational fiscal rules are not a substitute for the structural reforms
necessary to develop a more sustainable fiscal federalism framework.
 Increasing difficulties to comply with the golden rule with negative effect on the
financing of infrastructure services. The compliance with the golden rule depends on the ability
to generate operating surpluses that, together with new borrowing, will finance capital
spending. The golden rule aims at ensuring that borrowing should be used to finance productive
investment and cannot be used to finance operating deficits. In this regard, the squeezing of the
fiscal space for investment, which affects most of the state governments independently of their
indebtedness levels, implies that investment needs may not be met even in low indebted states.
Final remarks
76. This chapter analyzes the intergovernmental transfer system in Mexico. The current
intergovernmental finance system exhibits large vertical fiscal imbalances. Despite the existence of
regional socioeconomic disparities, the current system does very little to mitigate them. Furthermore,
the high dependence on federal transfers undermines SNGs’ incentives to collect more or to spend
better. Mexico’s transfer system is complex, and its objectives are inconsistent, resulting in a system
that does not promote devolution or equalization. The system creates weak incentives for SNGs to
support local economic growth or strengthen own-source revenue collection. The chapter presents
some preliminary proposals to address these issues, which would require a more in-depth examination.
77. Mexico’s property tax collection is relatively low, presenting an opportunity to increase
revenue collection by improving coordination among different levels of government and enhancing
institutional and technical capacity. This would involve addressing existing challenges, such as
improving and updating cadastral systems. One major issue affecting property tax collection is the lack
of local capacities in many municipalities. Most municipalities in Mexico do not have the capacity to
effectively administer a property tax system. International experience offers some practical solutions,
including a more centralized state administration of the property tax system, while preserving municipal
autonomy for the tax. This approach could help reduce collection costs, and improve efficiency,
effectiveness, and sophistication in terms of fiscal cadasters.
78. The LDFEFM has been effective in strengthening controls and monitoring of SNGs finances in
Mexico. One of the goals of the LDFEFM was to control the growth of borrowing and mitigate financial
stress among SNGs. However, the current thresholds used in the “traffic light” system may not be
conveying the appropriate signals and restrains. Only a few states are constrained by the fiscal rule, and
101
A more transparent mechanism for the federal government to support SNGs under financial distress could be the issuance of
a partial guarantee on subnational debt within a fiscal adjustment or debt restructuring program for SNGs.
139
highly indebted states may continue to contract debt up to the 200 percent of NER, which undermines
the core objective of the LDFEFM. The “traffic light” system’s high thresholds and large intervals may
limit its ability to differentiate between the level of indebtedness of SNGs. As such, readjusting the
thresholds used in the “traffic light” system can enhance its signaling ability and help better monitor and
control SNGs’ borrowing. This will also assist in mitigating financial stress among these entities. Overall,
revising the thresholds used in the “traffic light” system is necessary to ensure that the LDFEFM achieves
its goals of controlling borrowing and reducing financial stress among SNGs.
140

Mexico Public Finance Review - PFR - To Client División

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mexico Public Finance Review - PFR - To Client División

Uploaded by

Copyright:

Available Formats

MEXICO

Public Finance Review

GFC (t = 0 2008 q3) COVID (t = 0 2020 q1)

4. A more growth-friendly fiscal policy becomes increasingly important to support inclusive

Source: SHCP and OECD Revenues Statistics.

Figure ES6: Decentralization ratios in OECD countries, 2020

Source: OECD Fiscal Decentralization database

GFC (t = 0 2008 q3) COVID (t = 0 2020 q1)

TFP Capital Human Capital Labor Quantity GDP Total

1990-1999 2000-2009 2010-2015

Figure 1.8 Efficiency frontier Figure 1.9 Public Investment Efficiency

Source: IMF (2019)

Source: WEO April 2022 Source: SHCP

LATAM & Caribbean

Source: OECD Revenue Statistics.

Parastatales /2; 25.9

Average, 2017-2021 Average, OECD (excluding LA)

Source: WEO April 2022

3000 Transfers to subnationals

Transfers other than to sub-nationals

Compensation of employees -15

Transfers to Sub-nationals by Federal Government Contribution to Growth Federal Transfers to Sub-nationals

4900 Transfers abroad

1800000000000 15.0 4800 donations

4600 Transfers to trusts, mandates and other similar

439 Other Subsidies

Source: World Bank Mexico BOOST and WB staff calculations.

Housing and community ammenities

1000000000000 Public order and safety -2 Defense

Defense General public services

Spending in Economic Affairs Contribution to Growth, Economic Affairs

8 Science, Technology and Innovation

300000000000 4 Mining, Manufacturing and Construction

Federal Spending on Fuels and Energy

250000000000 Purchase of titles and securities

Figure 1.34: Economic Classification: Figure 1.35: Contributions to Over-Execution in

Figure 1.36: Functional Classification Figure 1.37: Functional Classification

-2 Public order and safety

Cyclicality of Spending in Mexico

JAM BGR POL OMNSAU

TGO ZMB IND HND

Procyclical spend - Countercyclical

Use of goods and services

EMEs and MICs

Source: IMF, Fiscal Monitor April 2022.

Box: Fiscal risks deriving from PEMEX and CFE

Official Excl. Transfers and Tax Benefits

CFE Financial Balance 2013-2021 (MXN billion)

2 4.1 4.1 4.1

Source: WB Mexico BOOST.

Composition of Federal Spending, by Economic Contribution to Growth in Federal Spending, by

5000 Other 15 Financial investment

Financial investment 10 Capital investment

Source: SHCP and WB staff calculations.

Source: SHCP and OECD Revenues Statistics.

Mexico vs OECD countries

Source: OECD Revenue Statistics.

Mexico’s tax structure relies heavily on income taxes (% of GDP)

Source: OECD Revenue Statistics.

Note: Marginal contributions are estimated with respect to

Marginal contribution to equity

Kakwani Direct taxes Marginal contributions to equity (right axis)

Figure 2.6. The number of brackets seems excessive