2020 - Institutional and - Economic - Determinants of - Regional Public - Debt in Spain PDF

Article
Public Finance Review

2020, Vol. 48(2) 212-249
ª The Author(s) 2020
Institutional and Article reuse guidelines:
sagepub.com/journals-permissions
Economic DOI: 10.1177/1091142120901672
journals.sagepub.com/home/pfr
Determinants of
Regional Public
Debt in Spain
Mar Delgado-Téllez1 and Javier J. Pérez1
Abstract
We analyze from an empirical point of view the evolution and determinants
of Spanish regional public debt. Spain offers an interesting case study because
of its high level of fiscal decentralization, implemented gradually during the
past four decades, the parallel entry into force of a number of national fiscal
rules in that period, and the heterogeneity of its regions, both in terms of
economic fundamentals and some institutional features. Our main findings
are the following: (i) on average, over the sample of study, regional govern-
ments’ fiscal policies reacted to public debt increase; (ii) fiscal rules played a
limited role in controlling debt surges; (iii) a higher degree of regional fiscal
coresponsibility tends to be linked to more subdued debt dynamics;
(iv) market disciple has encouraged some fiscal restraint at the regional level;
and (v) increases in public commercial debt have affected the standard debt.
Keywords
regional public debt, fiscal rules, fiscal federalism, market discipline
1
Banco de España, Madrid, Spain
Corresponding Author:
Javier J. Pérez, Banco de Espana, C/. Alcalá 48, Madrid 28014, Spain.
Email: javierperez@bde.es
Delgado-Téllez and Pérez 213
The analysis of subnational public debt developments has gained relevance

worldwide, due to the rising share of subnational finance in the overall
financing needs of the General Government in a number of countries,
following a continued process of fiscal decentralization (see Canuto and
Liu 2010; European Commission [EC] 2012). In this context, we study the
evolution and the determinants of Spanish regional governments’ net
financing needs (measured by the change in public debt). We estimate
empirical models exploiting the pool structure of the data (17 regions over
the period 1995–2017).
The Spanish case is interesting from a global perspective, given (i) the
decentralization process carried out over the past four decades, which
entailed a significant degree of fiscal devolution to heterogeneous regions,
and (ii) the adoption of several national and supranational fiscal rules. In
addition, Spanish regional governments have been traditionally very active
in capital markets and have accumulated significant levels of public debt
during the recent economic and financial crisis. The latter led to substantial
fiscal stress during the euro area sovereign debt crisis, which resulted in a
partial bailout for several regions.
Against this framework, this article contributes to the literature that
analyzes the determinants of regional public finances. We build upon pre-
vious analyses in the field of fiscal federalism, but focus on the determi-
nants of public debt dynamics, rather than on the drivers of primary fiscal
balances. We select changes in debt as our preferred object of study instead
of budget balances, given that the former is a broader measure of net
financing needs and debt accumulation, and because deficit-debt adjust-
ments (stock-flow reconciliation) were significant in Autonomous Commu-
nities (AC henceforth, Spanish regions) in the period 2010 to 2013.1 Some
papers that precede in certain aspects our work are Argimón and Hernández
de Cos (2012), Simon-Cosano, Lago-Peñas, and Vaquero (2013), and Mus-
sons Olivella (2017), among others.2 In addition, we explore a more up-to-
date period of time and include a number of less standard fundamental
variables, in particular those related to the structure and composition of
public debt, the definition of fiscal rules’ variables, and the interaction of
both regional public enterprises’ debt and regional governments’ commer-
cial debt with regular public debt.
Among the set of determinants, we pay special attention to (i) institu-
tional factors such as fiscal decentralization and fiscal rules including self-
correcting mechanisms like the reaction to past debt and past deviations
from targets, (ii) political factors related with the electoral cycle and the
214 Public Finance Review 48(2)
composition of government, (iii) market-disciple indicators such as the

change in the implicit interest rate and the structure of debt itself, and
(iv) non-EDP debt3 focusing on public corporations controlled by AC and
regional commercial debt.4 We find that the evolution of AC debt has been
mainly determined by past debt levels, the business cycle conditions,
market-induced fiscal discipline, and the presence of vertical fiscal imbal-
ances among public administrations.
This article is organized as follows. In Stylized Facts on Regional Public
Debt section, we provide some stylized facts on regional public debt in
Spain. In Institutional Framework section, we focus on two relevant insti-
tutional issues: the process of fiscal decentralization since the early 1980s
and the evolution of fiscal rules affecting regional governments in Spain. In
turn, in Empirical Analysis section, we perform the empirical analysis of the
article, covering first the standard approach of papers on fiscal federalism,
and moving next to a deeper look at the role of fiscal rules and market
discipline indicators to end up with some results on the link between
regional governments’ standard debt and other public debt concepts.
Finally, in Conclusions section, we provide some conclusions.
Stylized Facts on Regional Public Debt

Some Trends
Spanish General Government debt increased in the period 2007 to 2017 by
nearly sixty-three points of gross domestic product (GDP; see figure 1). By
subsectors of the General Government, the increase in debt was mainly due
to the surge in central and regional debt. In nonconsolidated terms, the debt-
to-GDP ratio in these two subsectors increased from 29.5 percent and 5.7
percent of GDP, respectively, at the end of 2007 to 86.9 percent and 24.8
percent of GDP in 2017.
The surge in public debt came hand in hand with raises in other liabilities
not covered by the standard definition of debt, but that are close complements,
namely public corporations’ debt and regional commercial debt. Regional
public corporations’ debt peaked in 2010, with a maximum of 1.2 percent of
GDP, whereas commercial debt increased steeply up to the beginning of 2012,
reaching 5 percent of GDP, and decreasing sharply afterward (see figure 2).
A Standard Decomposition of Debt Changes

It is worth looking at the evolution of debt in the period under scrutiny
through the lens of the government budget constraint. Let Yt be the nominal
In % of GDP In % of GDP
100 30 25
80 25 20
60 20 15
40 15 10
20 10 5
0 5 0
95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17
Central Government Sub-national governments Autonomous Communities Local Government
Figure 1. The evolution of public debt in Spain, by subsector. Source: Banco de España.
215
In % of GDP
5 1,2
4 1,0
3 0,8
2 0,6
1 0,4
0 0,2
95 97 99 01 03 05 07 09 11 13 15 17
Other accounts payable Public Companies Debt (rigth axis)
Figure 2. Other Autonomous Communities’ liabilities not included in the standard

definition of public debt. Source: Banco de España.
GDP at t and let Dt be the nominal value of government debt. The govern-
ment budget constraint accounts for how nominal interest rate it, net inflation
pt , net growth in real GDP gdpt , net-of-interest deficit as a percent of GDP
deft , and deficit-debt adjustment DDAt combine to determine the evolution of
the government debt-to-GDP ratio,
Dt 1 þ it Dt1 DDAt
¼ þ deft þ ; ð1Þ
Yt ð1 þ pt Þðgdpt Þ Yt1 Yt
where the nominal yield it and the real stock of debt Dt are averages of
pertinent objects across terms to maturity. Its linearized version, suitable for
accounting decomposition of the fundamental determinants of debt, takes
the standard form:
Dt Dt1 Dt1 DDAt
¼ ðit pt gdpt Þ þ þ deft þ : ð2Þ
Yt Yt1 Yt1 Yt
With this decomposition at hand, it is possible to analyze the determinants

of changes in the debt-to-GDP ratio. In figure 3, we decompose these
determinants for each year over the period 1996 to 2017 for the aggregate
of AC (see figure 4 for the evolution of individual AC’s debt). AC reduced
marginally their stock of debt in the period till 2007, with positive factors
In % of GDP
6
-2
-4
1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
Primary Balance Interest Payments Real GDP growth GDP deflator growth Deficit-debt adjustment Change in debt
Figure 3. The determinants of changes in regional governments’ debt (changes as a percent of gross domestic
product) in the period 1995 to 2017. Source: Banco de España; Ministry of Finance.
217
218
Andalusia Aragón Asturias Baleares
In % of GDP In % of GDP In % of GDP In % of GDP
6 6 4 6
4 4 4
2
2 2 2
0
0 0 0
-2 -2 -2 -2
95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17
Canarias Cantabria Castilla y Leon Castilla-La Mancha

4 8 6 10
6 8
4
2 6
4
2 4
2
0 2
0
0 0
-2 -2 -2 -2
95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17
Figure 4. Autonomous Communities’ (regional governments) debt evolution in the period 1995 to 2017: Changes as a
percent of gross domestic product. Source: Banco de España.
Catalonia Valencia Extremadura Galicia
6 10 4 4
8
4
6 2 2
2 4
2 0 0
0
0
-2 -2 -2 -2
95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17
Madrid Murcia Navarra Basque Country

4 8 6 4
6
4
2 2
4
2
2
0 0
0
0
-2 -2 -2 -2
95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17 95 97 99 01 03 05 07 09 11 13 15 17
La Rioja
In % of GDP
4
-2
95 97 99 01 03 05 07 09 11 13 15 17
Figure 4. (continued)
219
(real GDP growth and inflation) broadly compensating the debt-increasing

effect of interest payments and, to a much lesser extent, primary deficits.
The most recent crisis generated a significant contribution of public deficits
to debt accumulation. Finally, deficit-debt adjustments exerted a significant
impact, in particular over the period 2010 to 2013.
Institutional Framework
The Process of Fiscal Decentralization in Spain
After a gradual transfer of responsibilities from the central government to
the regions since the 1980s, Spain has now become one of the most decen-
tralized countries in the European Union. Indeed, 50 percent of General
Government expenditure was carried out by subnational governments in
2017, with about 36 percent and 14 percent in the hands of regional gov-
ernments and local governments, respectively (see figure 5, second panel).
In particular, AC were responsible for more than 90 percent of health-care
and education public expenditures in 2017. The transfer of those expendi-
ture responsibilities has been asymmetric both in pace and intensity. The
main differences concern the time at which the various AC took over
education and health competencies.
In parallel to this devolution process, a financing system for the subna-
tional governments was also progressively developed (see figure 5, first
panel, on the extent of revenue decentralization). Again, the process was
not completely homogeneous across regions. In particular, a distinction
should be drawn between the ordinary-regime AC (all except the Basque
Country and Navarre), with partial fiscal autonomy, and the specific-status
AC (the Basque Country and Navarre), which have full fiscal autonomy
with the exception of customs tariffs.
In essence, the Basque Country provincial authorities (Álava, Guipúzcoa
y Vizcaya) and Navarre’s regional government have the power to maintain,
establish, and regulate, inside their territory, the tax regime. However, there
are some coordinating provisions with the central government, which basi-
cally imply that the effective overall tax burden arising from their regula-
tory power must not be lower than the existing in the rest of the country.
The Fiscal Rules Framework

AC have always been subject to some constraints and limitations on their
capacity to borrow and/or generate budget deficits. Specifically, credit
operations with term to maturity of less than one year can only be used
A Ratio of subnational governments' revenue to general government revenue (%)

80
70
60
50
40
30
20
10
0
AT
LV
FI
NL
MT
UK
IT
LT
DK
EL
CY
IE
HU
SI
LU
HR
EE
DE
PL
FR
SK
CZ
PT
BE
SE
RO
BG
ES
EU28
Regional Local
B Ratio of subnational governments' expenditure to general government expenditure (%)

70
60
50
40
30
20
10
0
AT
LV
FI
MT
NL
UK
IT
LT
CY
DK
IE
EL
HU
SI
LU
HR
DE
PL
FR
EE
SK
BE
CZ
PT
SE
RO
BG
ES
EU28
Regional Local
Figure 5. International comparison of subnational governments’ revenues and

expenditures (2017). Country nomenclature following International Organization for
Standardization 3166/2. (a) Ratio of subnational governments’ revenue to General
Government revenue (percent). (b) Ratio of subnational governments’ expenditure to
General Government expenditure (percent). Source: Eurostat.
to cover temporary treasury requirements, while credit operations at a lon-

ger term should meet some conditions including in some instances the
authorization of the central government. In particular, one condition resem-
bles a “golden rule.” Advocates of this rule claim that governments’ bor-
rowing should only finance productive public investment that has potential
to pay for itself over the long term (Köppl-Turyna and Pitlik 2018). Mus-
grave (1988) justifies the golden rule as a matter of intergenerational equity
as investment in fixed capital may benefit not only the current generation
but also those that follow.
In Spain, a number of fiscal rules constraining AC deficits have existed
since the beginning of the fiscal decentralization process described in the
previous section. The framework was reinforced recently on the wake of the
European sovereign debt crisis. In particular, in 2012, a more stringent and
comprehensive fiscal rules’ framework was approved by the Spanish Par-
liament (Budgetary Stability and Financial Sustainability law5). The main
elements of this legislation are (i) a debt limit for the General Government
as a whole (a debt-to-GDP ratio of 60 percent) and for each subsector (13
percent of GDP for AC), (ii) a (structural) balanced budget rule, (iii) a
public spending growth limit, and (iv) detailed rules to guarantee the timely
payment to government’s private sector providers.
Empirical Analysis
In this section, we analyze empirically the determinants of regional public
debt.
Data and Hypotheses to Be Tested

In line with the extant literature, we include a large number of potential
explanatory variables grouping them in economic, political, and institu-
tional factors. We follow closely the definitions and variables of Argimón
and Hernández de Cos (2012). Apart from the standard ones used in the
literature, we also include a number of less conventional variables linked to
the structure of public debt and market discipline measurement (see Table 6
for the summary statistics).
Economic variables used as controls. Economic theory has highlighted the

economic cycle as a fundamental determinant of budget balances and, as
a consequence, of changes in public debt. Therefore, we include in our
analysis the lag of real GDP growth for each AC (variable Economic cycle),
taken from the Annual Regional Accounts published by the Spanish Statis-
tical Office (INE). The channels through which economic growth affects
public debt are two-fold: first, through the denominator of the debt-to-GDP
ratio and, second, via its impact on debt accumulation. The former is based
on the idea that even high debt ratios can be sustainable in a framework of
healthy economic growth, while in a situation of low or negative growth,
even low debt ratios can turn out to be nonsustainable. The latter is due to
budget deficit increases in economic downturns, generated either through
the operation of automatic stabilizers or through the impact of counter-
cyclical discretionary fiscal policies designed to stabilize the economy,
while the opposite occurs in expansions.6 We also include a variable that

captures budget compliance (normalized by GDP).7
Among the set of economic factors, we also consider as a control variable
the deviation of each region’s GDP per capita from the regional average as a
proxy for the degree of each region’s economic development.8 A priori, it is not
obvious what would be the expected relation of this variable with the evolution
of debt. Ceteris paribus, while a richer AC can obtain private financing more
easily due to its wealth status (leading to a tendency to register higher debt
levels), at the same time, its cost of financing would be probably lower,
decreasing its financing needs and pushing debt levels down. The literature
does not provide clear evidence on which effect dominates.
Overall inflation is a factor that is typically considered a key driver of the
debt-to-GDP ratio. It may exercise opposing forces over debt accumulation.
On the one hand, inflation indirectly affects tax revenues (depending on the
degree of wage indexation), which may be compensated by the increase in
price-indexed expenditures,9 while on the other hand, it impacts directly on
debt through its deflating effect on the debt-to-GDP ratio. We include the
deviation of each region’s consumer price index with respect to the national
mean in such a way that possible common trends are taken into account.
Finally, a branch of the literature emphasizes the role of asset prices in
affecting fiscal outcomes. In particular, it is stressed the latter may affect
fiscal outcomes through the tax system (taxes on capital gains and losses,
taxes on transaction, and tax relief, in particular, in the Spanish case, for
house purchases). In the case of Spain, financial and nonfinancial assets
form the basis of certain taxes managed and collected by AC. In order to
capture this determinant of debt, we include housing prices because of their
relevance in the recent Spanish boom period (1995–2007). Housing prices
might be a good proxy to capture the incidence of assets on regional public
finances. We define the variable as follows: deviation of the change in each
region’s index of housing prices with respect to the national mean.
Fiscal rules. We use a measure of the strength of fiscal rules developed by the
EC, the so-called fiscal rules index, that is available for the whole period of
study.10 Public debt developments may be affected by the presence of
different types of fiscal rules insofar as they supposedly pose a permanent
constraint on fiscal policy. One of the main hypothesis that we intend to
contrast is whether fiscal rules have subdued the surge of regional debt, that
is, whether the rules have had the expected dissuasive effect. Thus, if fiscal
rules are effective, we would expect that an increase in strength of rules
should entail a moderation of debt. We also include the interaction of this
index with three variables: lagged economic growth, the lag of the deviation
of budget balances with respect to the average for all AC, and the lag of the
deviation of the debt-to-GDP ratio to the average.11 The main idea behind
the selection of these variables is to test whether fiscal rules are more
effective depending on the economic and fiscal situation of the AC.
There are no previous insights about the effectiveness of fiscal rules
contingent on the strength of economic growth. Nonetheless, it would be
easier for the central government to be stricter in the application of rules to
those AC that present strong economic growth. Thus, we would expect a
negative impact for the interacted variable. A similar reasoning may be
applied to the other interacted variables. If the budget balance ratio is worse
for an AC in t 1, that is, it has a higher deficit or a smaller surplus, fiscal
rules would be effective if in the current year the debt level does not grow as
much. This would mean that the rules are effective for those AC with less
sound finances. In the case of the debt interaction variable, when an AC has
a higher debt, fiscal rules would be effective if the impact is negative,
meaning that the higher the debt, the less would it grow on the following
period. It may be noted that in the previous studies, such as Argimón and
Hernández de Cos (2012) and Delgado-Téllez, Lledó, and Pérez (2017),
fiscal rules do not seem to be very effective in deterring deficit noncom-
pliance. Therefore, we would not expect a high effectiveness of fiscal rules
in controlling debt evolution.
Political and institutional factors. The literature has proven the necessity of
including political and institutional factors to explain the persistence of
budget deficits and the accumulation of debt in advanced economies. In
our analysis, we include a number of political variables: (i) ideology, mea-
sured, first, by the percentage of left-wing MPs over the total seats of
regional parliaments and, second, by the weight of regionalist parties’ MPs
(regional parties that do not belong to a national party) over the total
number of seats of the regional parliament; the first variable aims at captur-
ing potential pro-spending biases depending on the ideological orientation
of the regional government, while the second could be instrumental for
testing the existence of different attitudes toward public debt accumulation
depending on the scope of the objective function of the regional govern-
ment; in both cases, there are no specific assumptions; (ii) a dummy for the
political concordance of national and AC governments that measures the
political alignment between both governments; (iii) electoral cycle: instead
of the standard election dummy that displays a value of 1 in an election year
and a 0 otherwise, we use a transformation of the original variable to
measure proximity to elections, computed as a continuous variable of the

distance to elections as in Franzese (2000, 2002) and Mink and de Haan
(2005).12 This variable takes the value of 1 in the electoral year while in the
next period, it would take 1/number of years left until the next election.
A similar reasoning applies to the subsequent years.13
There are two possible expectations about the impact of the electoral
cycle. First, the year after elections governments may start new investment
projects, with the intention of finishing them before the next electoral date.
Thus, debt would increase most in the first years of the electoral cycle. On
the other hand, expenditure tends to increase when the electoral date
approaches, implying a surge of debt to finance the new expenditure. More-
over, governments have incentives to increase debt rather than tax rates to
finance the additional expenditure because of the fiscal illusion channel,
that is, taxpayers may not consider an increase in debt as an increase in tax
payment in the future.14
Fiscal federalism–related control variables. The territorial organization of a

country has also been signaled by the extant literature as a further determi-
nant of the fiscal situation, either measured by the fiscal balance or by the
stock of debt. In particular, the responsibilities assumed by the regions, the
instruments for financing them, and the relationships between regional and
central governments are all factors that certainly affect the aggregate fiscal
outcomes of a given country and, more specifically, the distribution of fiscal
outcomes among the different layers of government. Notably, the literature
has devoted some effort to analyze the existence of a so-called soft budget
constraint problem whereby a regional government may have incentives to
conduct an undisciplined fiscal policy under the expectation that the central
government will intervene in case of trouble (see Qian and Roland 1998;
Kornai, Maskin, and Roland 2003; Sorribas-Navarro 2012; Delgado-Téllez,
Lledó, and Pérez 2017).
Following the literature, we include in our analysis some alternative
measures of fiscal coresponsibility, measured by (i) tax autonomy, the ratio
of taxes over which the regions do have normative power, over their total
nonfinancial revenues;15 (ii) the ratios of expenditure on education and
health over regional GDP, main expenditure items leading the evolution
of expenditure decentralization over the last twenty-two years.16 Also the
ratio of public investment over total expenditure is used as a proxy of
voluntary spending (not under legal minimum requirements). (iii) Vertical
fiscal imbalances that are proxied following Eyraud and Lusinyan (2013) by
the difference between nonfinancial revenues and expenditures, both net of

transfers with other public administration levels.
The literature argues that there should be correspondence between the
extent of a given region’s spending responsibilities and its fiscal autonomy
(fiscal coresponsibility), the latter being understood as the ability of the
regions to generate income to finance that spending. Otherwise, vertical
fiscal imbalances could emerge in the regions that would be usually filled
by federal transfers. These transfers distort the relationship that should exist
between the level of taxes and the benefits obtained by citizens in a partic-
ular region, creating a common pool problem. As regards the impact of own
revenue decentralization on fiscal balances, Governatori and Yim (2012)
discuss that theory does not provide clear predictions. On the one hand, a
high level of tax autonomy means that regional governments have more
own resources to cover a given amount of expenditures, leading to better
fiscal balances. On the other hand, one has to acknowledge that this type of
variable conveys no information on the relative size of regional own rev-
enues compared to their expenditures, which is probably a better way to
capture regional governments’ incentives to behave in a financially respon-
sible way. In addition, the impact of revenue decentralization may differ
depending on the share of transfers/taxes in AC’s revenues.
Market discipline, structure of debt. Beyond the factors analyzed in the pre-
vious paragraphs, the ability to increase debt by a given level of adminis-
tration is fully determined by its ability to raise the necessary funds. In
addition to increasing taxes or decreasing expenditure, AC need external
financing (national or international). Thus, either investors buy the debt of
AC or banks grant loans to them. One possible conjecture is that market
pressure might be a key determinant on the change in public debt. The case
of Spain is not one in which there is full reliance on capital markets to
contain subnational borrowing as in the cases of Canada, Switzerland, and
the United States. The latter are cases in which the central government does
not set any limits on regional governments’ borrowing, so that these levels
of government are free to decide the form of borrowing, and may decide by
themselves to adopt a fiscal rule in an attempt to enhance their credit
standing in the market.17 Before 2012, Spanish regional governments were
constrained by upper-level rules while they were also subject to strict mar-
ket scrutiny. Nonetheless, in response to heightened fiscal stress levels due
to the European sovereign debt crisis, the central government constituted
the so-called Regional Government Liquidity Fund in 2012 which provided
funds directly to AC. Conditionality and financial control were reinforced
to compensate the weakening of market discipline for those AC that decided

to enter this financing program.18
To approach the influence of market discipline, we explore the following
control variables: (i) the implicit interest rate, as a measure of the market
pressure;19 (ii) the aggregate average AC issuance cost rate, as a market
distress indicator; and (iii) a number of variables linked to the composition
of debt, as follows: First, the ratings in last period may impact the debt
evolution because the lower the ratings, the harder should be to issue new
debt. Second, the ratio of short-to-long-run debt, being short-term debt the
one with a maturity less or equal to twelve months. Short-term debt could be
associated with the reaction to sudden changes in market sentiment. Thus, it
is not always a negative indicator because a government may expect a
reduction of interest in the near future, and therefore, it would prefer
short-term debt until costs decrease.20 In a context of worsened perception
about a given sovereign, though, increased reliance on short-term debt can
lead to heightened vulnerabilities as worsening perceptions of a given
region’s creditworthiness can quickly feed into higher interest costs. Third,
the ratio of securities-to-loans, with the prior in mind that loans could be
more easily obtained in somewhat “captive” markets versus open compe-
tition to capture investors in securities. In the particular case of the regions
of Spain, regional savings banks (“Cajas de ahorros”), based on given
regions and with a governance structure prone to regional government’s
influence typically assumed a role as AC bankers. Finally, the ratio of debt
held by nonresident versus that held by residents offers an alternative per-
spective of markets stress as resident investors probably have better infor-
mation on AC’s economic situation, whereas nonresidents will probably
react quicker and shift portfolios toward more secure assets than residents
when AC’s credit risk surges. Meanwhile, from AC point of view, resorting
to foreign financing helps reducing the possible crowding out effect of the
debt surge (Broner et al. 2019) and the doom-loop problem or nexus
between sovereign and bank credit risk (see Brunnermeier et al. 2016).
Additional control variables: Pressure from units accounted for outside the
boundaries of the General Government sector. In particular, we consider the
two non-EDP debt measures described in Stylized Facts on Regional Public
Debt section: the debt of public corporations owned by a given region and
regional commercial debt. Indeed, the related literature would suggest that
(i) under tight budgetary rules, a government may try to circumvent its
constraints by cutting transfers to public corporations that, in turn, can
finance the same spending by issuing debt that is not computed by means
of the same accounting standards used to define the rule (typically as in

national accounts); (ii) an excessive level of nonstandard debt may end up
generating pressure on the responsible government to bailout the external
indebtedness vehicle.21 (iii) In turn, commercial debt may be used as a
buffer in times of crisis, encouraging governments to delay payments to
their suppliers, as it actually occurred in Spain up to 2011. We also include
an informed dummy of the Supplier Payment Financing Fund (FFPP) pay-
ments in percentage of GDP. This includes all the payments of the suppliers
of the AC made by the central government funds.
The Empirical Model

The incidence of the different determinants on the changes in public debt
mentioned in the previous section will be tested by means of a standard
econometric model that can be specified in quite general terms as:
Dit XN
D ¼ ai þ bj Ojit þ eit : ð3Þ
Yit j¼1
Under the proposed approach, the change in public debt of each regional
government, i, at time t, D DYitit , depends on a set of control variables, O,
encompassing the economic, political, institutional, market-induced, and
non-EDP factors mentioned above. Following the traditional fixed effects
model, ai in equation (3) aims at capturing all the unobservable AC effects
that are time-unvarying, while eit is an error term assumed to be white noise.
As for the estimation method, and in order to avoid any biases stemming
from the possible correlation between the individual effects and the regres-
sors, we estimate model (3) in first differences. Moreover, given the possible
simultaneity of some of the control variables and the dependent variable, the
estimation is carried out by the generalized method of moments (Arellano
and Bond 1991) using lagged regressors as instruments.
Although this type of models were first conceived for a sample with a
large number of subjects and a small number of periods, it has been grow-
ingly used in the literature for macroeconomic analysis purposes, where T is
relatively large and N is small. Nevertheless, when estimating this model for
a number of years (T) not small and number of subjects (N) not large (T ¼
23 and N ¼ 17), some problems may arise. First, we cannot use all lags
available as we would have a problem of proliferation of reduced form
coefficients (see Arellano 2016). Therefore, we have restricted the number
of lags to two.
Moreover, as the dynamic panel model methodology was developed for

a sample with a large N and small T, a small bias may appear in our model
estimation. Thus, coefficients may change slightly when changing the spe-
cification of the model. Therefore, for the sake of robustness, we have also
estimated a fixed effects model. As we are not trying to estimate the exact
impact of a specific variable, the small bias should not be a problem in our
conclusions. Therefore, in interpreting the results, we will be cautious in not
considering them as a fiscal reaction rule. In addition, turning back to the
fixed effects model, and following the Hausman test criteria results, the fix
effects model is chosen over the random effect.
Results
The results are shown in tables 1 to 4. The two first tables include results
generated with two different samples, the whole sample (1995–2017) and a
sample excluding the recent crisis period (1995–2007), and two fixed
effects models’ estimation for robustness sake.
In table 1, we explore the role of the main economic variables and the
fiscal rules. In table 2, we focus on studying the institutional and political
factors. The effect of the different measures related with the market disci-
pline is shown in table 3 while in table 4, we consider the impact of the non-
EDP debt (public corporations and commercial debt) on the AC’s debt.
Beginning with the baseline model shown in table 1, there are six models
included in this table. The first two columns show the generalized method
of moments (GMM) for the whole sample, the second two the GMM with a
restricted sample from 1995 to 2007, and finally, the last two columns show
the fixed effects model for the whole sample. The dependent variable is the
change in debt-to-GDP ratio. The first conclusion we draw from the results
is that AC debt does not seem to follow an explosive path as both the lagged
dependent variable and the lagged level of debt show a negative effect on
the change of debt, which in the case of the level of debt is highly signif-
icant. This means that whenever the level of debt or the change in debt
increases in the previous year, the change in debt will be subdued in the
current one. This is probably due to a combination of factors, first, if AC
debt increases, the central government would control more thoroughly the
finances of the AC, and second, it would be harder for the AC to finance this
surge of debt at a reasonable cost, encouraging a stronger fiscal effort to
reduce the deficit. Interestingly, the stock of debt impact is stronger for the
shorter sample, probably indicating a looser fiscal discipline during the
crisis.
Table 1. The Determinants of Regional Governments’ Debt Changes (Changes as a Percentage of GDP): Baseline Models.
230
Dependent Variable: D EDP Debt GMM (1995–2017) GMM (1995–2007) FE (1995–2017)
Lagged dependent variable 0.12* 0.13* 0.13 0.23* — —

(0.06) (0.07) (0.16) (0.13) — —
Economic cycle (t 1) 0.16*** 0.21*** 0.05*** 0.09*** 0.29*** 0.34***
(0.02) (0.03) (0.02) (0.04) (0.02) (0.02)
Budgetary compliance (t 1) 0.08 0.03 0.04 0.10 0.30*** 0.16**
(0.05) (0.06) (0.14) (0.13) (0.06) (0.07)
EDP debt (t 1) 0.15*** 0.13*** 0.65*** 0.55*** 0.04*** 0.03***
(0.01) (0.02) (0.17) (0.12) (0.01) (0.01)
FFPP 0.12*** 0.11*** — — 0.12*** 0.11***
(0.01) (0.01) — — (0.01) (0.01)
Inflation deviation 0.06 0.04 0.16 0.21* 0.16 0.25
(0.16) (0.16) (0.13) (0.12) (0.18) (0.18)
GDP per capita deviation 0.07* 0.07* 0.08*** 0.09*** 0.00 0.00
(0.04) (0.04) (0.03) (0.02) (0.01) (0.01)
House price deviation 0.02 0.01 0.01 0.01 0.00 0.01
(0.01) (0.01) (0.01) (0.01) (0.02) (0.02)
Fiscal rules balance deviation(t 1) — 0.03 — 0.10** — 0.10***
— (0.03) — (0.05) — (0.03)
Fiscal rules growth deviation (t 1) — 0.09*** — 0.05 — 0.12***
— (0.02) — (0.04) — (0.02)
Fiscal rules debt deviation (t 1) — 0.03*** — 0.05 — 0.00
— (0.01) — (0.03) — (0.00)
Fiscal rules index (t 1) 0.07 — 0.04 — 0.06 —
(0.10) — (0.07) — (0.07) —
Number of observations 357 357 187 187 374 374
Hansena 0.50 0.43 0.95 0.75 — —
m1 0.02 0.09 0.63 0.48 — —
m2 0.36 0.22 1.00 0.71 — —
R2 — — — — .72 .75
Hausmanb — — — — 0.00 0.02
Note: Instrument set in all models includes the second and third lag of the endogenous variable. EDP ¼ excessive debt procedure; GDP ¼ gross
domestic product; GMM ¼ generalized method of moments.
a
Hansen is the p value of the test of the overidentifying restrictions (see Hansen 1982), which is asymptotically distributed w2 under the null
hypothesis that these moment conditions are valid. A p value equal or higher than .05 indicates that the instrument set is valid, which is confirmed
under all models, m1 and m2 are the p values of serial correlation tests of order 1 and 2, respectively.
b
Hausman test p value under the null hypothesis indicates that the random effects model is preferred due to higher efficiency.
*significant at 10 percent level.
**significant at 5 percent level.
***significant at 1 percent level.
231
232
Table 2. The Determinants of Regional Governments’ Debt Changes (Changes as a Percentage of GDP): Political and Fiscal
Decentralization Variables.
Dependent Variable: D EDP Debt GMM (1995–2017) GMM (1995–2007) FE (1995–2017)
Lagged dependent variable 0.12* 0.13* 0.15 0.25 — —

(0.06) (0.07) (0.17) (0.17) — —
Economic cycle 0.16*** 0.14*** 0.05*** 0.05** 0.27*** 0.22***
(0.02) (0.02) (0.02) (0.02) (0.02) (0.02)
Budgetary compliance (t 1) 0.08 0.06 0.05 0.00 0.40*** 0.21***
(0.06) (0.06) (0.15) (0.16) (0.06) (0.06)
EDP debt (t 1) 0.15*** 0.14*** 0.59*** 0.50*** 0.02** 0.04***
(0.02) (0.01) (0.14) (0.15) (0.01) (0.01)
FFPP 0.12*** 0.12*** — — 0.11*** 0.13***
(0.01) (0.01) — — (0.01) (0.01)
Fiscal coresponsibility — 0.04** — 0.02* — 0.00
— (0.01) — (0.01) — (0.01)
Vertical fiscal imbalances — 0.09*** — 0.07** — 0.05
— (0.03) — (0.03) — (0.03)
Education expenditure over GDP — 0.16 — 0.01 — 0.30*
— (0.21) — (0.13) — (0.16)
Health expenditure over GDP — 0.48*** — 0.25 — 0.52***
— (0.11) — (0.24) — (0.10)
Investment expenditure over GDP — 0.14 — 0.10 — 0.10
— (0.19) — (0.20) — (0.11)
Concordance center–periphery 0.17 — 0.05 — 0.17* —
(0.19) — (0.10) — (0.10) —
Percentage of left-wing parties’ MPs 0.01 — 0.00 — 0.00 —
(0.02) — (0.02) — (0.01) —
Pro-autonomy 0.03** — 0.00 — 0.00 —
(0.01) — (0.02) — (0.00) —
Distance to elections 0.11 — 0.04 — 0.04 —
(0.07) — (0.08) — (0.16) —
Hansena 0.56 0.50 0.96 0.98 — —
m1 0.06 0.02 0.64 0.86 — —
m2 0.34 0.13 0.90 0.32 — —
R2 — — — — .71 .78
Hausmanb — — — — 0.00 0.00
Note: Instrument set in all models includes the second and third lag of the endogenous variable. EDP ¼ excessive debt procedure; GDP ¼ gross
domestic product; GMM ¼ generalized method of moments.
a
b
233
Table 3. The Determinants of Regional Governments’ Debt Changes (Changes as a Percentage of Gross Domestic Product):
234
Market Discipline.
Dependent Variable: D EDP Debt GMM (1995–2017) FE (1995–2017)
Lagged dependent variable 0.06 0.05 0.06 0.06 — —

(0.06) (0.06) (0.06) (0.06) — —
Economic cycle 0.15*** 0.13*** 0.16*** 0.15*** 0.29*** 0.30***
(0.02) (0.02) (0.02) (0.02) (0.02) (0.02)
Budgetary compliance (t 1) 0.02 0.03 0.03 0.06 0.24*** 0.26***
(0.09) (0.08) (0.07) (0.07) (0.06) (0.06)
EDP debt (t 1) 0.25*** 0.20*** 0.18*** 0.18*** 0.04*** 0.04***
(0.04) (0.02) (0.03) (0.02) (0.01) (0.01)
FFPP 0.12*** 0.12*** 0.12*** 0.12*** 0.12*** 0.13***
(0.01) (0.01) (0.01) (0.01) (0.01) (0.01)
Rating (t 1) 0.12 — 0.00 — — —
(0.08) — (0.06) — — —
Ratio short-/long-term debt 0.00 0.01 0.00 0.01 0.01* 0.01**
(0.01) (0.01) (0.01) (0.01) (0.00) (0.00)
Ratio securities/loans 0.02** 0.02** 0.03*** 0.02*** 0.00 0.00
(0.01) (0.01) (0.01) (0.01) (0.00) (0.00)
Ratio debt nonresidents/residents 0.00 0.00 0.00 0.00 0.02*** 0.02***
(0.00) (0.00) (0.00) (0.00) (0.01) (0.01)
Implicit interest rate 0.39*** 0.32*** — — 0.10*** —
(0.10) (0.06) — — (0.02) —
Issuance interest rate — — 0.04 0.05** — 0.07**
— — (0.02) (0.03) — (0.03)
Hansena 0.28 0.17 0.43 0.43 — —
m1 0.03 0.01 0.01 0.02 — —
m2 0.20 0.29 0.94 0.62 — —
R2 — — — — .75 .74
Hausmanb — — — — 0.00 0.00
Note: Instrument set in all models includes the second and third lag of the endogenous variable. EDP ¼ excessive debt procedure; GMM ¼ generalized
method of moments.
a
b
235
Table 4. The Determinants of Regional Governments’ Debt Changes (Changes as a Percentage of Gross Domestic Product): Non-
EDP Components (Public Corporations Owned by Regional Governments and Commercial Debt).
236
Dependent Variable: D EDP Debt GMM (1995–2017) FE (1995–2017)
Lagged dependent variable 0.28*** 0.28*** 0.03 — — —

(0.06) (0.07) (0.08) — — —
Economic cycle 0.15*** 0.16*** 0.17*** 0.19*** 0.18*** 0.39***
(0.02) (0.03) (0.03) (0.02) (0.02) (0.02)
EDP debt (t 1) 0.08*** 0.07*** 0.15*** 0.02*** 0.02** 0.04***
(0.03) (0.02) (0.03) (0.01) (0.01) (0.01)
Public companie’ debt (t 1) 0.08 — 0.00 0.14 — 0.60***
(0.56) — (0.68) (0.12) — (0.13)
D Public companie’ debt — 0.24 — — 0.36* —
— (0.25) — — (0.20) —
Commercial debt (t 1) 1.65*** 1.63*** — 1.37*** 1.46*** —
(0.29) (0.28) — (0.11) (0.11) —
D Commercial debt (t 1) — — 1.34*** — — 1.18***
— — (0.27) — — (0.16)
Hansena 0.10 0.12 0.03 — — —
m1 0.01 0.00 0.00 — — —
m2 0.59 0.59 0.36 — — —
R2 — — — .68 .69 .62
Hausmanb — — — 0.00 0.00 0.03
Note: Instrument set in all models includes the second and third lag of the endogenous variable. EDP ¼ excessive debt procedure; GMM ¼ generalized
method of moments.
a
b
Hausman test p value under the null hypothesis that the random effects model is preferred due to higher efficiency.
The second main result from the estimations is that, as expected, the
business cycle affects significantly the evolution of debt.22 In turn, the
variable GDP per capita deviation has a negative but weak impact, signaling
that those AC that have a higher GDP per capita tend to have less accumu-
lation of debt. Neither consumer price inflation deviations nor the house
price inflation deviations seem to have any significant impact on debt
evolution.
As regards the variable measuring budgetary compliance, it displays a
puzzling outcome because it is only statistically significant in the fixed
effects model. If the sign of the coefficient is negative, as it is in the fixed
effects and the long sample GMM, it means that fiscal noncompliance in the
past does not restrict regional governments in the present. Regarding the
variable measuring the “Fund for Suppliers’ Payments,” it has a significant
and direct effect on debt evolution as expected because the funds received
since 2012 were directly recorded as standard debt.
Still in table 1, the last four variables are the ones related with fiscal
rules. First, the fiscal rules strength index in the previous period does not
have a significant impact on debt variation one year after. This is a very
striking result that is nonetheless in line with previous studies such as
Argimón and Hernández de Cos (2012) and Delgado-Téllez, Lledó, and
Pérez (2017): current fiscal rules are not very effective in coping with fiscal
noncompliance (deficit control). In Spain, most of the central government
control until recently was concentrated on deficit targets compliance as
there were no specific debt targets until 2012. Notwithstanding, there are
some significant results when interacting the fiscal rules index with GDP
growth, the budget balance, and the debt-to-GDP ratio in deviations from
the AC average. As it appears, fiscal rules would tend to be more effective
when economic growth is lower, when debt levels are higher, and when
budgetary balances are worse. Thus, the two last results are as expected, but
it is slightly puzzling that fiscal rules are more effective when economic
growth is more dimmed. This could be rationalized if one considers that in
bad economic times public finances can be at stress, and tight rules might be
necessary to cope with potential sustainability problems.
In table 2, we focus on political and institutional factors. Political factors
seem to have a weak effect. Curiously, the only political variable that has
some effect is the weight of regional parties’ parliamentarians but neither in
the short sample GMM nor in the fixed effects model (FE) it has any
significance. More interesting is the result of the institutional factors’
impact. Revenue autonomy plays a key role in debt evolution as expected.
When AC have a larger share of their revenues coming from taxes and other
sources of revenue that are directly controlled by the regional government,

debt seems to be more stable. This is because first, they can better anticipate
the amount of revenues they will have in an specific year in contrast to a
situation of heightened dependence on central government transfers, and
second, regional governments would have more margin of manoeuver
because they could raise tax rates instead of relying on new debt. In the
same vein, vertical fiscal imbalances have a direct impact on debt increases
as expected. Finally, only the health expenditure variable displays a direct
and significant impact on the dependent variables. This may have to do with
the rigidity of health expenditure, with enormous fixed costs and the exis-
tence of strong social resistance to decrease this specific expenditure. These
results are robust to the exclusion of the two AC that have a differentiated
financing regime (the foral system), namely Navarre and Basque Country,
which enjoy a higher level of autonomy in tax collection.
Table 3 includes the analysis of the role of market discipline–related
variables. The following additional results in this table can be underlined:
(i) the two proxies for the cost of financing, the implicit interest rate and the
issuance rate, behave similarly. An increase in the implicit interest rate or in
the issuance rate discourages increases in debt; thus, higher financing cost
may have a deterrent effect on AC level of indebtedness.23 (ii) The indica-
tors that measure market discipline through the evolution of the ratings are
not significant in the estimations. (iii) Neither the ratio of short-to-long-
term debt nor the ratio of loans held by nonresidents versus by residents has
a significant effect. In the first case, it matches with our hypothesis of
ambiguity of the short-to-long-term ratio. In the second, it may have to
do with the restrictive nature of the variable as it only includes holders’
nationality structure for loans and omits securities’ holders information
due to nonavailability. (iv) Finally, the ratio of securities over loans pre-
sents a negative significant sign. This variable captures the extent to which
regional governments have access to less “captive” investors (those buy-
ing securities) that tend to request more fiscal discipline from the issuer.
Consequently, it would seem that some degree of market discipline is in
place.
Lastly, in table 4, we show some estimated models to assess the linkages
between regional governments’ standard debt, on the one hand, on two
other types of debt linked to their activity, namely, the debt of regional
governments’ public corporations and regional governments’ commercial
debt.24 Only commercial debt presents a direct and significant impact on
debt evolution, that is, the dynamic behavior of commercial debt has some
predictive power over standard regional debt. Even though we do not show
it in the table, the results for this variable are alike for the short sample
period 1995 to 2007.
Finally, table 5 shows some elevated correlations among specific vari-
ables which may lead to multicollinearity problems. In order to ensure that
multicollinearity is not affecting our estimates, we replicate all models
considering separately those variables that show a strong correlation in an
attempt to confirm that the results are robust. In addition, we relax the strict
exogeneity condition considering the lag of the stock of debt as a prede-
termined variable, obtaining similar results with the exception of the sta-
tistical significance of nondisciplinary fiscal rules.
Conclusions
In this article, we study the determinants of regional public debt accumula-
tion process over the period 1995 to 2017. We find that it depends mainly on
stabilizing factors such as the economic growth and also on the previous
stock of debt. Other factors affecting public debt are market discipline,
proxied by the financing cost, and “the fiscal decentralization design,” that
is, the distribution of revenues and expenditure competencies. On the con-
trary, political variables and fiscal rules strength do not seem to have a
noteworthy effect on AC debt. Finally, commercial debt has been a source
of standard debt expansions, especially during the recent economic crisis.
These results highlight the need for some policy actions to ensure
regional debt sustainability. First, the lack of effectiveness of fiscal rules
might be read in relation to the economic literature emphasizing that a set of
features are crucial to achieve a certain incidence of any fiscal rule on the
behavior of governments. In this regard, issues such as transparency, the
credibility of penalties for fiscal noncompliance, and the existence of inde-
pendent institutions responsible for monitoring compliance appear as deter-
minants of the success of the fiscal rules. Second, as argued by Oates
(2008), fiscal discipline may be enhanced when subnational governments’
spending competences are funded to a larger extent with their own reven-
ues. Reviewing the current design of Spanish fiscal decentralization system
considering ways of improving efficiency and transparency in the govern-
ance could improve AC fiscal responsibility.
In addition, market discipline may be fundamental in enhancing regional
governments’ fiscal prudence. The recent bailout has deeply damaged this
source of discipline, rendering essential to recover investors’ trust by
improving their solvency image in order to resume financing their activity
with private funding. Finally, regulation has been adapted to prevent the
240
Table 5. Correlation Matrix.
Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
1. Debt change 1
2. Economic cycle .69*** 1
3. Inflation deviation .03 .02 1
4. GDP per capita deviation .07 .06 .28*** 1
5. House inflation deviation .05 .06 .26*** .09 1
6. Fiscal rules .43*** .41*** 1
7. Revenue autonomy .03 .02 .09 .46*** .01 .19*** 1
8. Vertical fiscal imbalances .14** .16** .08 .57*** .01 .03 .80*** 1
9. Health expenditure .14* .16** .26*** .80*** .09 .07 .47*** .65*** 1
10. Education expenditure .34*** .36*** .25*** .77*** .12* .39*** .31*** .57*** .80*** 1
11. Investment expenditure .09 .09 .44*** .07 .29*** .20*** .35*** .40*** .41*** 1
12. Electoral date distance .04 .08 .01 .01 .01 .02 .06 .04 0 .09 1
13. Concordance .16** .11* .09 .13** .01 .01 .11* .14** .06 .10* .14** .05 1
14. Percentage of left-wing MPs .14** .18*** .01 .22*** .03 .10* .09 .13** .11* .13** .16** .01 .05
15. Percentage of regionalist .06 .01 .17*** .56*** .05 0 .31*** .31*** .31*** .35*** .17*** .02 .38***
16. Short-/long-term debt .08 .02 .05 .02 .09 .24*** .16** .04 .05 .09 .07 0 .01
17. Securities/loans debt .19*** .17*** .06 .31*** .06 .16** .26*** .26*** .16** .34*** .12* .02 .16**
18. Nonresident/resident debt .02 .02 .14** .15** .08 .07 .09 .16** .04 .08 .12* .01 .1
19. Implicit interest .32*** .31*** .01 .04 .03 .73*** .09 .04 .07 .36*** .02 .01 .02
20. Issuance interest .13* .11* 0 0 0 .66*** .17*** .06 .06 .30*** .08 .04 .02
21. Rating .38*** .20*** .03 .16** .03 .74*** .01 .03 .06 .30*** .52*** .08 .05
22. Budget compliance .55*** .35*** .07 .05 0 .33*** .01 .11* .08 .20*** .08 .02 .20***
23. Debt/GDP .49*** .27*** .04 .08 0 .72*** .10* .01 .08 .32*** .41*** .03 .03
24. Fund for suppliers’ payments .67*** .35*** .02 .15** .02 .43*** .01 .04 .12* .25*** .23*** .08 .07
25. Public companies’ debt .50*** .64*** 0 0 0 .37*** .12* .11* .11* .33*** .21*** .09 .12*
26. Commercial debt .19*** .13* .07 .29*** .05 .05 .23*** .19*** .27*** .19*** 0 .06 .02
(continued)
Table 5. (continued)
Variable (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25)
1. Debt change
2. Economic cycle
3. Inflation deviation
4. GDP per capita deviation
5. House inflation deviation
6. Fiscal rules
7. Revenue autonomy
8. Vertical fiscal imbalances
9. Health expenditure
10. Education expenditure
11. Investment expenditure
12. Electoral date distance
13. Concordance
14. Percentage of left-wing MPs 1
15. Percentage of regionalist .30*** 1
16. Short-/long-term debt .04 .14** 1
17. Securities/loans debt .03 .41*** .13** 1
18. Nonresident/resident debt .09 0 .15** .28*** 1
19. Implicit interest .01 .06 .24*** .16** .17*** 1
20. Issuance interest .02 .01 .27*** .06 .1 .77*** 1
21. Rating .05 .14* .06 .44*** .06 .40*** .34*** 1
22. Budget compliance .04 .13** .20*** .11* .09 .27*** .14** .32*** 1
23. Debt/GDP 0 .01 .03 .21*** .1 .46*** .44*** .88*** .42*** 1
24. Fund for suppliers’ payments .16** .1 .06 .21*** .05 .21*** .05 .54*** .42*** .48*** 1
25. Public companies’ debt .12* .02 .05 .02 .13** .43*** .25*** .18** .31*** .06 .07 1
26. Commercial debt .04 .27*** .12* .25*** .12* .16** .09 .21*** .23*** .01 .06 .33***
Note: GDP ¼ gross domestic product.

241
242
Table 6. Summary Statistics.
Variables Observations Mean Standard Deviation Minimum Maximum
Main variable Debt change 391 0.82 1.51 1.48 9.46

Economic factors Economic cycle (t 1) 391 2.08 2.56 5.60 7.10
Inflation deviation 391 0.00 0.29 0.98 0.86
GDP per capita deviation 391 0.00 3899 7422 9335
House inflation deviation 374 0.00 3.00 7.31 15.17
Fiscal rules Fiscal rules 391 2.27 0.99 1.00 3.72
Fiscal rules economic growth deviation 391 3.68 6.29 17.05 23.06
Fiscal rules budgetary balance deviation 391 0.00 2.03 12.11 6.53
Fiscal rules debt/growth deviationt 1 391 28.87 33.84 2.23 158.81
Institutional factors Revenue autonomy 391 30.53 18.25 0.07 91.22
Vertical fiscal imbalances 391 8.94 4.74 6.29 33.62
Health expenditure 391 5.72 1.32 3.25 9.96
Education expenditure 348 3.91 0.84 1.75 6.26
Investment expenditure 391 1.69 0.73 0.44 4.06
Political factors Electoral date distance 391 0.62 0.28 0.25 1.00
Concordance 391 0.47 0.50 0.00 1.00
Percentage of left-wing MPs 377 43.72 8.94 18.00 62.00
Percentage of regionalist 391 20.20 23.66 0.00 76.00
Market discipline Rating (t 1) 327 8.82 3.01 1.96 13.00
Short-/long-term debt 391 9.47 11.09 0.00 61.00
Securities/loans debt 391 33.29 25.63 0.00 100
Nonresident/resident debt 391 50.81 44.59 0.00 351
Implicit interest 391 4.57 2.64 0.27 14.86
Issuance interest 381 3.81 1.94 0.57 9.58
Budget compliance (t 1) 391 0.24 0.93 7.58 2.33
Debt/GDPt 1 391 10.11 8.37 0.99 42.69
Non-EDP debt Fund for suppliers’ payments 391 1.86 5.61 0.00 46.10
Public companies’ debt 391 0.43 0.54 0.00 2.80
Commercial debt 391 1.07 0.57 0.38 2.62
Note: EDP ¼ excessive debt procedure; GDP ¼ gross domestic product.
243
increase in nonstandard debt as it now compels governments to comply with

a maximum delay in the payment to suppliers. Nevertheless, AC may find
other loopholes in the future to circumvent maximum legal debt levels.
Authors’ Note
The motivation of this article draws significantly from Hernández de Cos and Pérez
(2013). The authors take responsibility for all the remaining mistakes.
Acknowledgments
We thank, in particular, Pablo Hernández de Cos for his comments. We also thank
the editor; the associate editor; the three anonymous reviewers; Guillem López-
Casasnovas; Alessandro Turrini; Marcello Sartarelli; Vicente Rı́os; and participants
at the ECOMOD conference 2017, the AEDE conference 2017, and the Encuentro
de Economa Pública 2018 for helpful comments and discussions and Rocı́o Prieto
and Ligia Topán for their research assistance.
Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or
publication of this article.
ORCID iD
Mar Delgado-Téllez https://orcid.org/0000-0002-8259-2737
Notes
1. For a broad perspective on the role of deficit-debt adjustment, see Campos,
Jaimovich, and Panizza (2006) for an international perspective on this issue.
2. The institutional determinants of local governments’ indebtedness have been
more widely analyzed in the literature, mainly from a less aggregated macro
perspective than the standard in papers looking at the determinants of Auton-
omous Communities’ (AC) debt. See, for instance, Cabasés, Pascual, and Vallés
(2007), Bastida, Beyaert, and Benito (2013), and the references quoted therein.
3. EDP stands for excessive debt procedure. It is the relevant concept of debt for
the purposes of the ceilings set in the European Stability and Growth Pact.
4. Commercial debt consists of financial claims which are created as a counterpart
of a financial or a nonfinancial transaction with the private sector in cases where
there is delay in the payment.
5. Organic Law 2/2012, April 27, Organic law on budget stability and financial
sustainability.
6. Some authors point out, however, that the higher revenues in economic boom
periods may generally entail pressure on the growth of public spending in such a
way that the relationship between the economic cycle and the budget balance
may be altered or, at least, evidence asymmetrical behavior over the course of
the cycle. See Morris and Schuknecht (2007) on related grounds.
7. We measure the budgetary compliance as the difference between the actual
budget in cash terms and the initial budget. Thus, a positive value would entail
compliance as the budget balance would be better than expected in the initial
budget. A possible explanation for a better-than-first-expected budgetary beha-
vior would be the willingness of the AC government to comply with a medium-
term target.
8. This deviation is measured as the GDPpc divided by the average GDPpc multi-
plied by 100.
9. Apart from the impact on nominal gross domestic product (GDP; the denomi-
nator of the debt ratio), higher inflation may increase the budget deficit through
higher nominal interest rates and higher real cost of purchases of goods and
services or investment and, in general, of those items of public spending that can
be indexed (e.g., pensions and wages). In the presence of nonindexed taxes,
inflation may also generate higher revenues if, for instance, the tax rates are
progressive (see Hernández de Cos et al. 2016).
10. See https://ec.europa.eu/info/publications/fiscal-rules-database_en. The new
index is a normalized one, with a starting value of 0.45 to an ending value
of 2.27. We have rescaled the variable so that there are no negative values that
could hinder the estimation sign interpretation
11. Budgetary balances are obtained from the national accounts (NA) data pub-
lished by the Finance Ministry. Because these data are only available at an
annual basis from 2000 onward, the beginning of the sample is completed using
the public accounts’ budgetary data (cash based). This strategy will be used for
all NA variables. We use the lag of the budget balance over GDP to prevent
endogeneity problems.
12. On electoral cycles and budgetary outcomes, see, for example, Von Hagen
(2010) or Mink and de Haan (2005).
13. Thus, if there are four years between elections, the variable would take 0.25,
0.5, 0.75, and 1.
14. Concept first developed by Amilcare Puviani (Puviani and Volpi 1973) and
broadly analyzed in the literature (see, for instance, Oates 1979; West and
Winer 1980).
15. The period 1995 to 1999 is obtained as a linear extrapolation with the execution
data as the NA only cover the period 2000 to 2017. All NA data come from the
General Intervention Board of the State Administration website.
16. Yearly disaggregated data are available separately in the Classification of the
Functions of Government publication available at the Finance Ministry, being
2016 the last year available.
17. In Switzerland, after an episode of soaring debt over the nineties, the govern-
ment introduced a debt brake rule, which is a highly strict rule, according to
which the deficit of one year must be compensated in the next years, as
explained by Bodmer (2006).
18. The Regional Government Liquidity Fund is a financial facility created by the
central government to provide funding to the AC. The central government
directly paid the AC maturities and recorded those payment as loans granted
to the AC. The entrance to this program implied a reinforced conditionality with
a stronger supervision on the AC budgetary execution.
19. Measured as the ratio of interest payments over the stock of debt.
20. Some papers have found short-term debt to be an indicator of vulnerability to
international financial crises: Borensztein et al. (2004), Rodrik and Velasco
(1999), and Bussière and Mulder (1999).
21. On a discussion about the role of public sector enterprises in Spain, see Fernán-
dez Llera and Garcı́a Valiñas (2013).
22. We use the lag of the GDP growth to avoid possible endogeneity problems.
23. In the generalized method of moments, the contemporaneous lag of implicit
interest rate is included in the estimation so that endogeneity problem may be
solved. On the contrary, we considered that the issuance interest rate does not
have the problem of endogeneity as it is an average variable of all AC, and the
data are common for all of them. Therefore, the increase in debt in one AC
should not have a very significant effect on the average issuance rate.
24. This variable is only available for the aggregate of AC.
References
Arellano, M. 2016. “Modelling Optimal Instrumental Variables for Dynamic Panel
Data Models.” Research in Economics 70:238–61.
Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel Data:
Monte Carlo Evidence and an Application to Employment Equations.” The
Review of Economic Studies 58:277–97.
Argimón, I., and P. Hernández de Cos. 2012. “Fiscal Rules and Federalism as
Determinants of Budget Performance: An Empirical Investigation for the Span-
ish Case.” Public Finance Review 40:30–65.
Bastida, F., A. Beyaert, and B. Benito. 2013. “Electoral Cycles and Local Govern-
ment Debt Management.” Local Government Studies 39:107–32.
Bodmer, F. 2006. “The Swiss Debt Brake: How It Works and What Can Go Wrong.”
Swiss Journal of Economics and Statistics (SJES) 142:307–30.
Borensztein, E., M. Chamon, O. Jeanne, P. Mauro, and J. Zettelmeyer. 2004.
“Sovereign Debt Structure for Debt Prevention.” IMF Occasional Paper, 237,
International Monetary Fund, Washington, DC.
Broner, F. A., D. Clancy, A. Erce, and A. Martin. 2019. “Fiscal Multipliers and
Foreign Holdings of Public Debt.” ECB Working paper, 2255, European Central
Bank, Frankfurt.
Brunnermeier, M. K., L. Garicano, P. R. Lane, M. Pagano, R. Reis, T. Santos, D.
Thesmar, S. Van Nieuwerburgh, and D. Vayanos. 2016. “The Sovereign-bank
Diabolic Loop and Esbies.” American Economic Review 106:508–12.
Bussière, M., and C. Mulder. 1999. “External Vulnerability in Emerging Market
Economies: How High Liquidity Can Offset Weak Fundamentals and the
Effects of Contagion.” IMF Working Paper, 88, International Monetary Fund,
Washington, DC.
Cabasés, F., P. Pascual, and J. Vallés. 2007. “The Effectiveness of Institutional
Borrowing Restrictions: Empirical Evidence from Spanish Municipalities.” Pub-
lic Choice 131:293–313.
Campos, C. F., D. Jaimovich, and U. Panizza. 2006. “The Unexplained Part of
Public Debt.” Emerging Markets Review 7:228–43.
Canuto, O., and L. Liu. 2010. “Subnational Debt Finance and the Global Financial
Crisis.” World Bank - Economic Premise, The World Bank (13):1–7.
Delgado-Téllez, M., V. D. Lledó, and J. J. Pérez. 2017. “On the Determinants of
Fiscal Noncompliance: An Empirical Analysis of Spanish Regions.” In Fiscal
Politics, Chap. 14, edited by V. Gaspar, S. Gupta, and C. Mulas-Granados,
345–76. Washington, DC: IMF.
European Commission. 2012. Workshop “Fiscal relations across government levels
in times of crisis -making compatible fiscal decentralisation and budgetary dis-
cipline.” 1–16. Belgium: European Commission, Directorate-General for Eco-
nomic and Financial Affairs (DG ECFIN).
Eyraud, L., and L. Lusinyan. 2013. “Vertical Fiscal Imbalances and Fiscal
Performance in Advanced Economies.” Journal of Monetary Economics
60:571–87.
Fernández Llera, R., and M. Garca Valiñas. 2013. “The Role of Regional Public
Enterprises in Spain: Room for a Shadow Government?” Hacienda Pública
Española/Review of Public Economics 205:9–31.
Franzese, R. J., Jr. 2000. “Electoral and Partisan Manipulation of Public Debt in
Developed Democracies, 1956–90.” In Institutions, Politics and Fiscal Policy,
edited by R. R. Strauch and J. von Hagen, 61–83. Boston, MA: Springer.
Franzese, R. J., Jr. 2002. “Electoral and Partisan Cycles in Economic Policies and
Outcomes.” Annual Review of Political Science 5:369–421.
Governatori, M., and D. Yim. 2012. “Fiscal Decentralisation and Fiscal Outcomes.”
Directorate General Economic and Financial Affairs (DG ECFIN), European
Commission, Economic Notes.
Hansen, L. P. 1982. “Large Sample Properties of Generalized Method of Moments
Estimators.” Econometrica: Journal of the Econometric Society 1029–54.
Hernández de Cos, P., S. Hurtado, F. Martı́, and J. J. Pérez. 2016. “Public Finances
and Inflation: The Case of Spain.” Occasional paper, 1606, Banco de España,
Madrid, Spain.
Hernández de Cos, P., and J. J. Pérez. 2013. “Sub-national Public Debt in Spain
Political Economy Issues and the Role of Fiscal Rules and Decentralization.” In
Estado del bienestar: sostenibilidad y reformas. In XX Encuentro Economa
Pública, Universidad de Sevilla.
Köppl-Turyna, M., and H. Pitlik. 2018. “Do Equalization Payments Affect Subna-
tional Borrowing? Evidence from Regression Discontinuity.” European Journal
of Political Economy 53:84–108.
Kornai, J., E. Maskin, and G. Roland. 2003. “Understanding the Soft Budget Con-
straint.” Journal of Economic Literature 41:1095–136.
Mink, M., and J. de Haan. 2005. “Has the Stability and Growth Pact Impeded
Political Budget Cycles in the European Union?” CESifo Working Paper Series
1532, CESifo Group Munich, Munich, Germany.
Morris, R., and L. Schuknecht. 2007. “Structural Balances and Revenue Windfalls:
The Role of Asset Prices Revisited.” ECB Working Paper 737, ECB, Frankfurt,
Germany.
Musgrave, R. A. 1988. “Public Debt and Intergeneration Equity.” In The Economics
of Public Debt, 133–48. London: Palgrave Macmillan.
Mussons Olivella, J. M. 2017. “Fiscal Decentralization and the Cycle in Spain: An
Empirical Analysis of Fiscal Policy Responses.” Politica Economica 33:305–44.
Oates, W. E. 1979. “Lump-sum Intergovernmental Grants Have Price Effects.” In
Fiscal Federalism and Grants-in-aid, edited by Peter Mieszkowski and William
H. Oakland, 23–30. Washington, DC: The Urban Institute.
Oates, W. E. 2008. “On the Theory and Practice of Fiscal Decentralization.” Institu-
tional Foundations of Public Finance: Economic and Legal Perspectives 165:176.
Puviani, A., and F. Volpi. 1973. Teoria della illusione finanziaria. Milano, Italy:
Isedi.
Qian, Y., and G. Roland. 1998. “Federalism and the Soft Budget Constraint.” Amer-
ican Economic Review 88:1143–62.
Rodrik, D., and A. Velasco. 1999. “Short-term Capital Flows.” Technical Report,
National Bureau of Economic Research Working Paper 7364, National Bureau
of Economic Research, Cambridge, MA.
Simon-Cosano, P., S. Lago-Peñas, and A. Vaquero. 2013. “On the Political Deter-
minants of Intergovernmental Grants in Decentralized Countries: The Case of
Spain.” Publius: The Journal of Federalism 44:135–56.
Sorribas-Navarro, P. 2012. “Determinants of Subcentral Governments’ Bailouts:
Evidence for the Spanish Regions. European Journal of Political Economy 27:
154–70.
Von Hagen, J. 2010. “Sticking to Fiscal Plans: The Role of Institutions.” Public
Choice 144:487–503.
West, E. G., and S. L. Winer. 1980. “Optimal Fiscal Illusion and the Size of
Government.” Public Choice 35:607–22.
Author Biographies
Mar Delgado-Téllez is a senior economist in the Research Department of the Bank
of Spain (Eurosystem). She received her PhD from Complutense University of
Madrid, Spain. Her research interests include fiscal policy, regionalism, and
macroeconomics.
Javier J. Perez is the director of International Economics in the Research Depart-
ment of the Bank of Spain (Eurosystem). He received his PhD from Complutense
University of Madrid, Spain. His research interests include macroeconomics, fiscal
policies, European integration, and textual analysis.

2020 - Institutional and - Economic - Determinants of - Regional Public - Debt in Spain PDF

Uploaded by

Copyright:

Available Formats

You might also like

2020 - Institutional and - Economic - Determinants of - Regional Public - Debt in Spain PDF

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

2020 - Institutional and - Economic - Determinants of - Regional Public - Debt in Spain PDF

Uploaded by

Copyright:

Available Formats

Article

Public Finance Review

Mar Delgado-Téllez1 and Javier J. Pérez1

The analysis of subnational public debt developments has gained relevance

composition of government, (iii) market-disciple indicators such as the

Stylized Facts on Regional Public Debt

A Standard Decomposition of Debt Changes

Central Government Sub-national governments Autonomous Communities Local Government

Other accounts payable Public Companies Debt (rigth axis)

Figure 2. Other Autonomous Communities’ liabilities not included in the standard

With this decomposition at hand, it is possible to analyze the determinants

Canarias Cantabria Castilla y Leon Castilla-La Mancha

Madrid Murcia Navarra Basque Country

(real GDP growth and inflation) broadly compensating the debt-increasing

The Fiscal Rules Framework

A Ratio of subnational governments' revenue to general government revenue (%)

B Ratio of subnational governments' expenditure to general government expenditure (%)

Figure 5. International comparison of subnational governments’ revenues and

to cover temporary treasury requirements, while credit operations at a lon-

Data and Hypotheses to Be Tested

Economic variables used as controls. Economic theory has highlighted the

while the opposite occurs in expansions.6 We also include a variable that

measure proximity to elections, computed as a continuous variable of the

Fiscal federalism–related control variables. The territorial organization of a

the difference between nonfinancial revenues and expenditures, both net of

to compensate the weakening of market discipline for those AC that decided

of the same accounting standards used to define the rule (typically as in

The Empirical Model

Moreover, as the dynamic panel model methodology was developed for

Lagged dependent variable 0.12* 0.13* 0.13 0.23* — —

Dependent Variable: D EDP Debt GMM (1995–2017) GMM (1995–2007) FE (1995–2017)

Lagged dependent variable 0.12* 0.13* 0.15 0.25 — —

Dependent Variable: D EDP Debt GMM (1995–2017) FE (1995–2017)

Lagged dependent variable 0.06 0.05 0.06 0.06 — —

Lagged dependent variable 0.28*** 0.28*** 0.03 — — —

sources of revenue that are directly controlled by the regional government,

Note: GDP ¼ gross domestic product.

Variables Observations Mean Standard Deviation Minimum Maximum

Main variable Debt change 391 0.82 1.51 1.48 9.46

Note: EDP ¼ excessive debt procedure; GDP ¼ gross domestic product.

increase in nonstandard debt as it now compels governments to comply with

Declaration of Conflicting Interests

You might also like

Lagged dependent variable 0.28* 0.28* 0.03 — — —