18 India Infrastructure Report 2003
2 EXPENDITURE ACCOUNTABILITY AND
THE SOCIETY
2.1 EXPENDITURE ACCOUNTABILITY IN INDIA: THE INTERLINKAGES
THE CHALLENGE OF PUBLIC EXPENDITURE
ACCOUNTABILITY
The challenge of public expenditure accountability in
developing democracies is greater than in other societies
that are already developed or have chosen to grow without
firm commitment to democratic political processes. Political
processes in these countries are in many ways more complex
than in advanced democracies. Being still on the road to
transition all ideologies, including fascist libertarian,
autarkic, communist, anarchic, and others, are possible.
Their richness of extant political ideologies is similar to
that which existed in Europe in the 1890s, a little before
many countries of Europe completed their transformation.
Additionally, new ideologies like market socialism, mixed
economy, and ‘green’ have also emerged. But much of
Europe obviously did not give itself the luxury of being
fully democratic while the process of the industrial
transformation was on. Monarchies, constitutional and
absolute, and empires and republics without universal
franchise constituted the great variety of political systems
prevalent.
Late Industrialization and Accountability of the
Executive
Advanced countries having been successful would now have
to contend with few competing ideologies. Differences are
marginal and differ only on ‘small’ issues—size of particular
subsidies, degree of commitment to the environment, the
degree of resistance to import penetration, the desired size
of social security, and attitude to in-migration. So the need
Sebastian Morris
to be accountable to the legislature is accommodated without
much difficulty, because few believe that there are major
strategies (new things to do collectively) in the offing; for
which the state has to take the initiative. Elites of trans-
forming democracies, though, could have an agenda that
goes beyond what can be voted and is acceptable to the
legislature. Thus it would be difficult for a legislature to
endorse the vast subsidization, or more correctly, the
incentivization of exports, or the enormous financial re-
pression that went on in South Korea or Taiwan. These
measures are known to have speeded up the industrial and
technological developments. Additionally, selective policies
that included inter alia the patently unfair ‘selection of
winners’, that South Korea adopted were instrumental in
the country’s quick transformation. Executives in power
possibly do realize the value of such agenda, which, by the
standards of the mature democracies, would be unfair to
the idea of equality of opportunity and hence not achievable
through the consent of a legislature. Similarly, elites of Latin
America, used the state to further their own agenda of
‘dependent development’1. These have taken the form of
the elites’ integration with global (western) capital even if
it meant slower growth, and vast capital flight from these
economies. More importantly, democracies that have still
to carry out their industrialization are also weak in the sense
that many of their citizens (probably the majority), despite
having the power to vote, are handicapped by their poor
endowments. We had earlier in India Infrastructure Report
1 A term due to Peter Evans (1984), but the idea of the
dysfunctional dependency of Latin American societies on world
capitalism has an earlier origin. Cf. Frank Andre Gunder (1967).

IIR 2002 argued that good governance itself can be considered
as a public good whose value increases with development,
and as people’s incomes rise beyond a certain critical level
above which they value good governance. Thus the legislature
itself may not be truly representative, and the scope for
pursuing an agenda though, not in the true interest of the
poor, but, nevertheless, addressing them at a token or
symbolic level, that is, rhetorically, may be large.
‘Off-budget’ Expenditures are Common
‘Off-budget’ or non-voted expenditures in developing
democracies are more common than in mature ones, or
even autocracies. Contingent liabilities, cross subsidies,
highly inefficient modes of subsidization, multiple tax
rates, many deductions and exceptions in rates,
administrative mech-anisms for allocation of land and
scarce public goods are all vehicles for by-passing legislatures.
Many of the above are initiated in the name of the poor,
but could actually benefit many others than the poor.
Despite an active opposition, the opposition itself when
in power would derive much use from these ‘off-budget’
measures so the opposition and discussion is not against
‘off-budget’ tran-sactions per se but their specific uses and
objectives. Thus left-wing parties would like to use the state
in a manner that benefits their constituencies without
having to get the same voted. They would end up supporting
food or agricultural subsidies that are implicit rather than
explicit, because once in place such subsidies are most
difficult to remove. An industrial lobby could similarly
push through administered cost-plus pricing regimes for
products of mass consumption with built-in cross subsidies
rather than clean budgeted subsidies. Middle classes could
support allocations through administrative mechanisms
rather than the market, or through universal rationing,
because they could tilt such schemes to their advantage
either because the really poor are not even in the market
for the good2, or because they are confident about
influencing the actual allocation process3 given queues or
shortages.
In India just before the reform of the 1990s, there was
probably no group which did not enjoy (at least on paper)
some benefit as a result of state controls and distortionary
policies, while the economy as a whole bore the costs4.
The measures of implicit subsidies at 12 per cent of gross
domestic product (GDP) by the National Institute of
Public Finance and Policy (NIPFP) is only a partial measure
2 This is, for instance, true of the subsidies on higher education,
cooking gas, fertilisers, and even food in India.
3 Examples of these would include drinking water in urban areas,
already discussed in the last IIR.
4 The dead weight losses rather than the transfers may have been
far the greater.
Expenditure Accountability and the Society 19
of the off-budget activities of the government as a whole.
They do not include the aspect of the guarantees and
leveraged borrowings indulged in by the government.
Despite their vast and pervasive nature, formal efforts at
review/accountability, etc. that have emerged thus far may
amount to mere ritual efforts. Why is this so? Democratic
govern-ments have the task of not only carrying out the
functions of government (and, in less development countries,
the additional one of taking the country through the industrial
transformation), but also of being fair, or at least appearing
to be fair. Being fair may not be completely consistent with
the process of bourgeoisie development that brings about
the industrial transformation, so the task of appearing to
be fair has a special place. The discovery of hidden budgets
and whom they subserve does not automatically lead to
change on that count alone. Significant political shifts or
strategies for systemic change are required, and little is
achieved by the mere public knowledge of hidden budgets.
Hence there is a tendency to not correct major distortions
by eliminating or removing them, but to ‘work around
them’ sometimes through counteracting distortions. Cases
in point are the electricity subsidies or the oil subsidies,
which were discussed at length in the IIR (2002).
Politics is Not Everything
The above is rather well known and easily admitted. But
that should not deter attention from the major gains that
can be made through better and more efficient processes,
and reforms to correct many of these ills, even within the
constraints of the current state and politics. As was brought
out in the IIR (2002), the vested interests of today may
have been only an unintended consequence of past policy
errors. Reform has the potential to unlock forces in favour
of reform especially in democracies, when the strategically
correct steps are taken. It means that reformers have the
added task of making and pushing for options that are
politically feasible and enhance the size of the economy5.
Moreover, as the competition for rents increases, the
returns to competition and control decreases, and that in
turn can release strong political process for reform and
better management of the state. It is our belief that the
reform of the early 1990s have resulted in forces (including
political) for the second round of reforms and change, but
policy-makers have not been able to live up to the task
of providing alternatives that are sufficiently contextualized
and workable. The abundance of doctrinaire ‘solutions’,
and the blanket blame of the politicians, only exposes the
inadequacies of the policy-makers today.
Accountability of government expenditure is an aspect
of good governance in the conventionally understood sense.
5 See Ashima Goyal, Section 2.2, in this report on the political
economy of expenditures, and investment by the state.
20 India Infrastructure Report 2003
We now discuss the case of the central government’s
expenditure processes.
DIMENSIONS OF EXPENDITURE ALLOCATION
We may usefully develop certain axes for the understanding
of both central government expenditures and transfers to
state governments and parastatals. These we may identify:
1. The spatial dimension: that is, where the money
is spent
2. On what the money is spent: viz. the sector.
3. Social optimality: is the money spent on the best
possible project that is on the project with the highest
social return given feasibility?
4. The performance dimension: do the project/prog-
ramme managers perform in the best possible manner?
And does performance feed back into the budgetary
(allocative) decisions?
The spatial and sectoral dimensions can have important
correlations and may not be considered as being entirely
independent. Thus an emphasis on coal usage and coal
development in national policy and its support through
central projects and sponsorship implies that regions with
coal resources are recipients or beneficiaries of government
expenditure.
Planning and ‘Objectification’ of Expenditure
With planning and the fiscal process in the post-independence
period, in both form and in content, the first three dimensions
were sought to be institutionalized and taken out of the
political process. The immediate result and basis of the same
was the creation of the Planning Commission.The Planning
Commission’s role is worth recalling in the context of
government expenditures. It allowed the objectification of
important expenditure decisions since a large part of
government expenditure (other than on routine adminis-
tration) was related to the Central (and State) Plans. And
even when not formally designated as Plan expenditure had
much to do with the carrying forth of previous Plan objectives,
and the assets and organizations they created.
Interestingly, the Mahalanobis Plan’s four broad sectors
evolved to more than a hundred sectors, so that planning
became an exercise in intersectoral consistency, worked
out through elaborate input–output models, which were
later expanded to bring in consumption demands.
The Planning Commission went beyond the sectoral
allocations, when it carried out project evaluation in the
Project Investment Board (PIB). The PIB only considered
the projects of the government and its public enterprises.
Private projects which were also part of the Plan were
merely incorporated, and funded in part through
development finance institutions. The budgetary contri-
bution to the Plan constituted the governments direct role
in investment. This non-discretionary way of spending
proved functional and kept the political pulls and pressures
at bay for quite some time, till the Plan collapsed in 1965.
The politicking and the pressures for investment as we see
today began only after the collapse of the Plan. Nevertheless
even in the Plan period upto 1965 there was one open
dimension wherein non-economic factors could have been
brought to bear—the spatial dimension. But this too may
have been severely curtailed because the big investments
in heavy industries and social overhead capital or
infrastructure were to a large extent location specific.
Marginal Political Role
Some leeway for pulls and pressures in the choice of head
offices of public sector corporations, investments in urban
infrastructure, investments in softer areas like high-cost
technical education institutions, would, of course, have
been there even when Nehru was alive. And sometimes
may have been even perversely exercised.6 Determination
of social optimality involved discretionary choices—taking
the form of choice of technology and equipment and
sometimes location when there were many possible locations.
It would be facile to say that all the investment expenditures
in such industries and the choice of equipment and
technology were taken in a completely apolitical manner.
Nevertheless, the overwhelming role of politics, which
came about in the 1970s and the 1980s, would have been
quite absent then. The blatant direction of investments to
the minister’s constituency that we see today, or the
locational politics played by Indira Gandhi in 1966 was
nonexistent.7 If the cases of the setting up of steel plants
or that of the oil industry in the 1950s and the 1960s were
typical, then a strategy to build self-reliance and the long-
term interests of the country were more overriding than
the politics of location, and the locational tournaments, we
see today.
The earlier objectivity that was derived from the Plan
could also then be vitiated. The objectivity if at all was then
entirely reduced to the sectoral allocations; and not to the
details of the projects or investment nor the locational choices.
6 A large factor in taking investments away from the spatial
optimal was the intrusion of ‘balanced’ regional development into the
Plan itself, through its policy of freight equalizations, which shifted
the advantage of steel and coal-using industries out of eastern India.
7 Around this time the systematic rejection of Talcher (even
today the world’s best site for a steel mill) for a second time, to
counter the influence of the old Congress in Andhra Pradesh and
Karnataka led to the sanction of the Vijaynagar and Vizag steel
plants. Both locations were far inferior than Talcher. See K. Das
(1997) and K. Das (1999).

SCB at Variance with Planning
The task of social cost benefit (SCB) analysis or cost
benefit (CB) analysis was not entirely consistent with the
derivation of investment targets and expenditures thereof
from the Plan. If the Plan warranted certain investments
then neither SCB nor CB could say that they were not
socially desirable or financially unviable. Private or
commercial viability is not an issue at all in such an
intersectoral exercise. All that SCB/CB ought to do is to
select the best out of possible projects to achieve the
investment and output targets. In this limited sense, therefore,
SCB /CB would amount to locational, scale, equipment,
and technology choices. In the latter the quality of the
technology transfers, and the possibility of linked credit
and the concessions thereof would then have to be assessed.
It is interesting that as long as the role of project evaluation
was restricted to this role of choice of the best project,
rather than investment choice, the Plan was functional. In
the 1970s the introduction of the financial return criteria,
that is, CB, would have acted to frustrate the Plan. The
requirement could contradict with the target investment
specified by the Plan model, and strict adherence to the
target return could mean underinvestments. That would,
through the input–output relations, create inconsistency,
and supply and demand linkages would then work in a
‘domino’ fashion to derail the Plan. Additionally, since the
Plan was largely closed to trade (and the exchange rate was
mis-aligned by a wide margin) the traded sector could not
have been used to overcome supply and demand limitations.
IRRs Compromised
What concerns us here is the ambiguity and ritualized
assessment of projects, particularly the SCB/CB analysis
that emerged out of this contradiction, and the ‘solution’
that was found. Rather than conceptually contest the
intrusion implied by commercial viability, the same was
accommodated as a ritual. The way the system got around
the problem was to report all desired and ‘predecided’
projects with internal rate of returns (IRRs) above the cut-
off rate of 12 per cent! There was, of course, no ranking
of projects based on financial IRR, since the ‘objective’
was to ‘maximize’ the social return ‘known better’ to the
bureaus, subject to a satisfactory financial return as revealed
by the IRR for CB. With such ‘accommodation’ then, it
was a simple matter to frustrate serious project evaluation.
If the IRR could be vitiated at the margin, then it was a
small step to violate it so completely that in effect the cost-
benefit analysis was a mere ritual that the state went
through. Going through the form of SCB and CB analysis,
laying out the planned increase in sales, etc. could have
at best constituted a hurdle, where the check was whether
Expenditure Accountability and the Society 21
the proposers could put down the project in the stated
form.
Equally important, the very framework of SCB and
financial analysis may not have been appropriate for the
country. It is well known that environmental costs are not
taken into account in the governments’ SCB. Similarly, the
costs of land used, and of displacement and rehabilitation
or compensation were not built into projects. Border prices
for tradables were not used systematically. Without border
prices but with social prices for goods and services the
planning and expenditure patterns would have been consistent
only either with autarky8 or with a large weight for the
Indian economy in world trade. And the presumption of
either is absurd. Equally importantly, by not using referral
prices elsewhere, inefficiency or administered pricing, which
took the prices far from the social cost, could be fed back
into investment decisions in a major way if border prices
were rejected for tradables. Such inconsistencies had deeply
distorted the economy. Other social prices, such as that
of labour, were not used systematically.9
Plan Targets and Competition for Expenditure
The above discussion in no way means that cost-benefit
analysis or project evaluation was downplayed. Not only
did entities such as public enterprises carry out detailed
project analysis but the Planning Commission’s PIB virtually
duplicated the process when it examined every item at
great length and over much time—looking at it from all
possible ‘angles’ including security, foreign exchange
conservation, promotion of indigenous technology, and
use of local equipment, credit linkage, protection of small
industry, employment, and such like, before approval was
granted. The processes were long drawn out, wasteful, and
prone to being manipulated10. One could not exclude the
possibility of projects being caught in the loop where so
much time was being taken for approval that it was already
time to revise the estimates, and hence for re-appraisal.
Since 1965, the plan targets may have served principally
to define the resources available for particular constituents,
not necessarily that going into a particular sector, unless
it was a narrow one like power with its own department.
8 The Soviet Union with its vast land and natural resources may
have come as close to realizing planning under autarky as was
possible for any real economy.
9 In any case the possibility of government (or enterprises) being
able to use labour intensive technology was seriously affected by the
schism in the labour market. But the government’s non-use of social
price of labour could have seriously affected significant opportunities
that were there especially during the 1970s. It would have also
reinforced the bias against exports right through to the present.
10 See Morris (1987) for a detailed analysis of the processes and
in what manner they were dysfunctional even from the logic of
planning and active state enterprise leading to industrialization.
22 India Infrastructure Report 2003
The total Plan investments included own resources of
public enterprises, market borrowings, and part of private
investments, besides the budgetary contributions to the
Plan. It is the budgetary contributions to the Plan and the
allowed market borrowings (since these were usually made
on the guarantee of the centre) that contributed public
resources, for which enterprises and departments on their
behalf competed.
Sooner rather than later, the process of planning became
one of departments /ministries taking back with them the
target estimate of the Planning Commission (Perspective
Planning Dept) of growth rate of the economy and the
sector and the investment requirements, and working out
the amount that can be assumed to come from the private
sector and the budgetary requirements. Also incorporated
were other extension, promotion and other infrastructure
that were seen as important in the upgradation of quality
and technology and overall development of the sector.
Departments such as Small and Village Industries,
emphasized the latter a great deal. For others where the
private sector was absent, and which were investment heavy
the key aspect were the investments into the sector. Besides
inter sectoral consistency, the formal division of tasks
(sectors) between the centre and the states constituted
another basis to determine the sectoral allocations. If transfers
(as in Power and Urban Development) from the centre to
the states are included, then the constitutional division may
not have been as important in delimiting the sectors, as the
formal list of state and central functions would indicate.
Programmes and Growth of the Bureaucracy
The Plan process was based on investment targets
operationalized through projects setting up enterprises anew,
and existing enterprises setting up projects. The taking up
of the ‘HYV Strategy’ and, following from it, the constitution
of programmes such as SFDA (Small Farmers Development
Agency), MFDA (Medium Farmers Development Agency),
the IRDP (Integrated Rural Developement Programme)—
the employment programmes, since 1965, brought in an
entirely new dimension to the allocation process of public
funds. As other programmes were added, the capital part
of Plan expenditure declined over the period from 1965
to 1980. More importantly, they brought in discretion in
the hands of the government, and the politician. Indira
Gandhi’s Garibi Hatao at the level of public expenditures
must be seen as creating the space for significant government
expenditures outside investment for ‘direct targeting’, to
create political support and for providing the bases for the
party in power to reward its vast numbers of field workers.
For the bureaucracy it provided a major avenue for expansion
and spread across sectors and regions not constrained by
the need for investments as they emerged out of the
intersectoral consistency modelling. The ideology of the
‘socialistic development’ of going directly to the poor, of
poverty alleviation, of creating employment, etc. provided
the intellectual basis for these ‘sop’ programmes. The rate
of growth of the government and the bureaucracy should
ideally have slowed down after 1965, since, by then, the
basic institutions (control and regulating organizations),
and the productive enterprises had already come into place.
But that rate was kept up to a steady 2 to 2.5 per cent
per annum due to the ‘space’ provided by these programmes.
Coordination Failure and Ritualized ‘Evaluation’
The ritualized assessment of financial viability that did not
correctly assess the risks would have reinforced the lack
of risk mitigating actions within government. The point
is not that the state should look at risks the same way the
private sector does. States can bear much larger risks
especially when the levers of the economy, such as monetary
and fiscal policy are with the state. Actually, the very
purpose of state involvement in production is to overcome
the risk barrier as private industry would see the same.
Given the vast coordination economies that are waiting to
be picked by states and very large corporations attempting
investments on a large scale, the Plan itself would have
reduced the risk of investments for society as a whole, and
for the private sector, and passed (communicated) this risk
reduction to individual enterprises through policies such
as licensing and import restrictions. But due to improper
project analysis, the state, which bore a large part of this
contingent risk, may not have been aware of the nature
and extent of the risk it was bearing. We may usefully call
this contingent risk the economic coordination risk. And,
therefore, mitigating efforts may not have been as vigorous
as they may otherwise have been.
Thus a very large part of the delays and cost overruns,
especially in the 1970s and the early 1980s, before the
environmental factor became very large, was due to ‘lack
of coordination’ between various sectors or enterprises
implementing projects. Failures in coordination, arising
inter alia out of bureaucratic slackening, hierarchical
processes and structures which disallow project-level
interaction between persons implementing the projects,
barriers created by rank and department, have imposed
severe penalties on projects. But this aspect of value arising
out of coordination would not have been clear to senior
officers and managers at the departmental and project
levels so that adequate efforts would not have been made
to bring about the required coordination.
Coordination always has a free rider aspect, and even
for governments this need not be false, unless consciously
there is an agent whose primary task is to bring about the
coordination. Or the positive externalities would have to

be correctly internalized through very large corporations,
which presumably are integrated enough to garner these.
The potential of very large projects, whose viability itself
depends upon the success of coordination, would go
unrealized if such organizational structures do not exist or
have been vitiated, even when they are taken up by the
state. The industrial and economic structures may still
change and development (diversification) take place because
the public sector is investing, but at an unnecessary and
additional cost. This is because the coordination risk and
its lack of mitigation (failed coordination) works its way
as large time and cost overruns. These would, of course,
slow down the growth, rates of the economy11.
PERFORMANCE LINKAGE
A major deficiency in public expenditure processes, that
has been there from the very beginning, is on the performance
dimension. If an authority/ministry/parastatal or state
enterprise did poorly in a project or programme, then there
was little in the process that reduced its chance to get
projects or programmes (expend money) in the next round;
or made it expensive for that entity to do so. While the
devolved funds through the Finance Commission formula
could not have had a performance link imposed by the
central government, the Plan funds to the states and others
could easily have had performance linkage, which could
have positively affected the allocation and spending processes.
A Major Failure
The only ‘performance linkage’, if one may consider it that
way, may have been in putting forth projects and programmes
especially in the latter, when a certain component was to
come from the state. Many of the administratively and
fiscally weaker states not being able to make out the formal
proposal, or put in its share of contribution would have,
therefore, lost out in getting funds for projects and
programmes from the central government. Poor project
implementation did not penalize the agency implementing
the same. There were systematic differences between
divisions of state Public Works Departments (PWDs)
across the country in implementing highway projects but
the central ministry of road and surface transport, until
recently, never distinguished on that criteria between these
divisions. It was always the right of a certain state PWD
division, by virtue of its location, to oversee and maintain
a particular portion of the National Highway. Funds,
therefore, for National Highway Development were routed
through these PWDs irrespective of their performance!
11 The relationship between cost and time overruns in public
investments, and the growth rate of the economy is brought out in
Section 5.1 by Sebastian Morris in this report.
Expenditure Accountability and the Society 23
Public sector organizations that were better than others in
expenditure management and efficient use of funds did not
get more plan funds than others that were more wasteful.
‘Achievement Equals Expenditure’
Processes may have even worked in a perverse manner to
give more public money to poorly functioning organizations!
Thus in the wake of structural adjustment, there was little
central funding available to the National Thermal Power
Corporation (NTPC) acclaimed as both an efficient
implementer of projects, and a manager of power businesses.
Instead large funds were still available to the National
Hydro Power Corporation (NHPC) whose operational
efficiency and capacity in projects was nowhere near that
of NTPC. The logic was that ‘anyway NTPC would be able
to manage’ through market borrowings and internal
resources. There is not a sector or a field of funding where
similar considerations did not hold. Under the cover of
this approach lies a major failure in public expenditure
management. Surprisingly, the idea that state funds are for
the weak, continues to have much circulation even today.
The analogy of the family is often quoted despite its
obvious irrelevance. A good part of public money even
today maintains parts of the state organizations and arms
of government including its productive apparatus,
irrespective of their contributions, need, or performance.
It certainly contradicts the idea of public expenditure for
the benefit of society! In this false idea the vehicle for
achievement of a task is being mistaken for the task.
No Learning from World Bank
Despite the example of the World Bank whose lendings for
projects were typically accompanied by close monitoring
and conditionalities relating to performance, there was no
attempt on the Planning Commission or Finance Ministry’s
part to impose conditionalities relating to performance,
pricing, etc. Had such measures evolved, the extremely
poor performance of both projects and programmes could
possibly have been avoided. The public expenditure system
could then have evolved in the direction of being ‘incentive
compatible’. It is this lack of performance linkage in allocation
of state funds that finally broke the back of the budget. No
country could have borne a steady and system-wide cost
and time overrun in the range of 50 to 60 per cent on half
(virtually the entire public sector part) of its capital formation,
and leakages in the same range or more out of its non-
capital forming but programme-related expenditures, and
not collapsed. Why has performance linkage been so difficult?
1. The Planning Commission in its expenditure
allocation and management aspect was entirely an extension
of the bureaucracy and therefore evaluation and examination
would simply have been broken down into particular tasks,
24 India Infrastructure Report 2003
ignoring their compositional aspect. In other words, there
was no organ responsible and able to translate the overall
plan tasks for other organizations as their specific tasks
and hold them responsible. This would have been important
because the total cost of non-performance of a part went
beyond that part as we mentioned in the discussion on
coordination economies and contingent risks.
2. Unlike a development bank, there is no possible
aggregate measure for the Planning Commission’s
performance as a whole that is meaningfully derived from
the subtasks, or projects’ and programmes’ performance.
For a development bank that does not have a monopolistic
position in the economy the surplus, the total lendings, the
recovery rate, and such other measures, are generated in
the natural course of the operations, and can be used for
measurement of its performance. The Planning
Commissions’ own money (capital or funds) are not at
stake. It is merely chanellizing public funds. Thus even
when disbursing the budgetary support, had the loans/
grants from the government been routed through as the
Commissions’, then certain appropriate measures of
performance could have been generated. Thus deviations
from standard times and approved costs for projects could
have been one measure of the Planning Commissions’
performance. This could have been done at least when
funds were rolled over as the productive enterprises it
created were able to get an increasing portion of their
requirements from the markets and internal resources.
3. Follow through in government bureaucracies to
moneys given is difficult, and the Planning Commission
was no exception. Typically, governments take a great deal
of time and spend much resources examining proposals
for spending but, once the money is given, there is little
ability or motivation to follow through to ensure that the
primary purpose or task for the money has been achieved.
The monitoring would be more in the form of the insistence
on periodic reports, or a completion certificate. These
reports need not be actionable as management control
systems of corporations are, and could at best, be diagnostic.
Most traditional government tasks like defence and general
administration can only be measured in terms of expenditure.
So even in the case of projects and programmes it is easy
and ‘consistent’ to let the task be measured by expenditure,
relative to allocations, even if they are contrary to the task
of adding to economic and productive capacity. Traditional
audit in government which is examination of ex-post
consistency with processes and rules could then continue
much as before.
4. Plan fund grants were not sought to be sharply
distinguished from devolved funds. So the idea of the
centre as a mere router of funds could have spilled over
on to the Planning Commission’s role.
5. There was incentive incompatibility in bundling
the planning and project /programme roles together. Had
the project role been carried out through an organization
structured like an investment bank, the effectiveness of
projects and their implementation could have had some
performance linkage. It is interesting that when the NTPC
used World Bank funds, it found the Bank’s conditionalities
useful in ‘disciplining’ the Indian government! It could
pressure the government to allow it to implement projects
quickly without the dysfunctional interference and
interventions by government departments. It also used the
conditionalities to persuade the government to allow it
rational prices for its output. The World Bank
conditionalities, even when its lending constituted a small
part of the total funds of projects, and even when guaranteed
by the government, put much weight in the hands of an
organization that was attempting to orient itself towards
its primary task. Thus the value of the World Bank assistance
or project funding may have been much more than that
measured by its lending to development projects in India.
Had the same or similar procedures been adopted for
much of Plan Funds by the Commission especially from
1965 onwards, it is quite possible that things would have
been very different, and growth would have been much
faster to unambiguously bring about the industrial
transformation in India. Growth in India was about 1.5
per cent per annum slower than what it would otherwise
have been due to the delays and cost overruns in public
sector projects. A large part of the responsibility for the
same should go to the Project Appraisal Division (PAD)
of the Planning Commission and the PIB. (Morris, 1987).
ORGANIZATIONAL PROCESSES AND ACCOUNTABILITY
While the political reasons for the poor state of public
expenditure accountability are recognized (but perhaps not
understood well enough), the bureaucratic and systemic
forces that result in slack and waste are not admitted as
independent causal factors. Dilatory processes, very low
work loads on administrative staff, multiple objectives that
also tend to be ill defined, all these arise in part out of
the nature of the bureaucracy in India. The general belief
though is that political interference is the cause of
administrative failure and inefficiency.
There is virtually no discussion (contextualized to the
specific social, political, and economic environment) of the
government in India as an organization that is prone to
failure. Thus, the apparently innocuous system of reporting
to the Parliament imposed on the Public Sector Units
(PSUs) have contributed a great deal to the dysfunctional
interface they have with the government. Their dysfunctional
interface is the primary reason for the failure of the PSUs—

their lack of accountability and task orientation. This had
been brought out at length in the IIR (2002). The more
general phenomenon of sub-units of government, irrespective
of function or original design, to have very similar processes
and work practices—rule orientation to the point of losing
sight of the task—is acute in India. Undue dependence on
leadership is another problem, Even when there is
commitment it is to the sub-tasks or to the functions and
sub-units rather than to the task as such.
Preference for ‘Patch-work’
Similarly, government procurement is also known to be very
problematic with no clear solution. The dysfunctional
relationship between the elected executive and the officers
is perhaps the most serious problem of all. We had argued
why responsibility embedment is difficult with a cadre-
based bureaucracy that is not under the control of the
elected executive, especially at the lower levels of government.
The political executive cannot promise task performance to
his clients (the people) because he has little control over the
cadre and officials in general, and so there is little or no
scope for him to make political capital. That leaves him with
the option of corruption and graft as a pay off. With the
passage of time only individuals with corruption and power
on their minds take to politics. Change has little chance
of coming from within the system. Only severe external
pressures would have a chance of leading to success. The
system’s capacity and motivation to correct mistakes and
distortions is weak, and ‘patch-work solutions’ are preferred.
We had made these and related arguments in the last IIR,
so there is no need to repeat them. We had also brought
out the deep-seated reasons for a large part of the
misgovernance and corruption—absurd policies, control
and subsidization measures and extension targets for
governments. With better design of policy, especially subsidy
delivery mechanisms, and with incentive compatible
structures for the same, many problems can be overcome.
Hence we had claimed that having the right policy and legal
framework would amount to attempting change with the
advantage of acting ‘at a leverage point’ for change. We had
suggested that organizationally another crucial point of change
is to make all top bureaucratic posts competitive, so that
experts outside the cadres could come in with new ideas
and lateral thinking. Additionally, we had also argued that
the key to embedment of responsibility and accountability
of local bodies, is to make the salaried employees truly
accountable to the elected executive, by changing the working
rules, and reporting relationship, rather than continuing
with the unbroken chain of command within the bureaucracy.
These issues have an obvious and direct bearing on
public expenditure accountability. Rather than restate the
obvious, we will cover the ground of public procurement,
and the lack of decentralization in government, which we
Expenditure Accountability and the Society 25
had not discussed in the last IIR, and also review findings
of the Expenditure Reforms Commission.
ERC Exposes Waste
The government had in 1997 set up the Expenditure
Reforms Commission (ERC), which has now submitted its
various reports. The report pertaining to the Ministry of
Urban Development and Poverty Alleviation12, is somewhat
more expansive and provides some measures of the waste
in government. Having approached the problem from the
financial angle, there are few, if any, discussions on the
dysfunctionalities in the working of the ministry. The
object, anyway, was expenditure reforms, as conventionally
understood, and the Committee adopted the rather simple
approach of identifying and recommending the removal of
surplus manpower and expenditure. So large and obvious
were the waste and surpluses that without much detailed
considerations it was possible for the Committee to
recommend rather major downsizing. And in reading
between the lines it was possible to gleam some of the
crucial problems with governmental processes today.
Ignoring the Limitations of Governments
The most glaring drawback is that there are too many
schemes, and an inability to design these schemes in a
manner that is coherent with the economic, social, and
political processes in society13. Widely recognized as an
inappropriate mode are discretionary controls like quotas,
licences, and banns that were rampant in the Indian system
till the economic reform of the 1990s. Similarly, with
regard to the design of programmes and schemes of the
Planning Commission and central ministries, is the
assumption that every problem14 can be overcome through
direct administrative action. Typically,the assumption has
been that a subsidy given to beneficiaries identified
continually, a low interest loan, or the special allocation
of an input considered as scarce or necessary for starting
an economic activity could be provided by the government,
directly to the intended beneficiary to solve the problem.
Besides the incoherence with markets and the economy
there was little or no recognition of the constraints that
emerge from the nature of bureaucracies that are inherently
ill designed to perform complex ‘enterprise’ tasks.
12 For a more detailed review of this report see Section 5.3 by
Biju Varkkey.
13 One aspect of this inadequacy, viz. the market-unfriendly
ways of control by the state, has been widely commented upon now
for over two decades.
14 Be it of poverty, unemployment, underdeveloped skills,
illiteracy, lack of education or nutrition, women’s ill health,
marketing by petty producers, underdevelopment technology, social
deprivation or exclusion.
26 India Infrastructure Report 2003
The greater the task the more the number of posts and
the deployment, has been a standard assumption of most
bureaucracies. This in India has created a fast growing
bureaucracy, that until the 1990s grew at rates faster than
the overall growth of the labour force in India. The problem
of integration, of interdepartmental incongruity or conflict,
of duplication, etc. are only now being seriously discussed
to find solutions15.
Cadres Create Dysfunctional Pressures
Additionally, cadre pressures create the demand for too
many posts. Cadres have the dysfunctionality of being
difficult to hold accountable to the executive. This aspect
had been discussed in the IIR 2002. The existence of
multiple cadres creates a very significant problem. Thus
there are the Indian Administrative Services (IAS), Indian
Police Services (IPS), Indian Revenue Services (IRS), Indian
Ordnance Factories Services (IOFS), Indian Supply Services
(ISS), etc. While the IPS and specialized services like the
Indian Audit and Accounts Services (IAAS), have meaning
in being separate, many others like the IAS, IOFS, Indian
Economic Services (IES) and ISS are not particularly different
in terms of the tasks the officer is called upon to perform.
But the separate identities create a strong pressure to continue
within the government activities for which these cadres
were created to provide the space for the growth of the
cadre. The Ordnance Factories are top heavy because officers
of the IOFS would otherwise not have had options. The
same is true of the Indian Supply Services. Their impact
on overall employment growth, wage, salary, and general
administration expenses are severe. No wonder then that
the ERC in its report on the Ministry of Urban Development
and Poverty Alleviation reports huge ‘surplus’ of posts.
Little Delegation
Too little delegation within the government is one of the
reasons for the massive over-manning. True delegation
would, at the top, imply reduced power to accommodate
dysfunctional tasks that could be thrust upon the bureaucracy
by corrupt politicians and others. Delegation implies that
there is better responsibility embedment since then the
tasks have greater coherence with the ability and power
of the person/position responsible for the same. When
there is little delegation then at the top there is the absolution
of responsibility of ‘small tasks’ which should anyway have
15The depth of the morass within the Central Public Works
Department (CPWD) is difficult to imagine. The contract value per
man month that the department handles is a mere Rs 0.5 lakh when
its norm (itself a highly watered down one) was Rs 5 lakhs. Private
builders and others securing construction services do at least Rs 10
lakhs per man month. Draftsmen in the department made 1.5
drawings per month against the norm of 10 per man month!
been done at the lower levels’, or which require the attention
of multiple functionaries for their performance. Victims
and senior civil servants then end up blaming the juniors
and lower levels of the bureaucracy. ‘Lower-down’ staff can,
of course, blame everything on ‘lack of powers’. The lack
of responsibility embedment is enhanced by the tendency
to have committees take decisions. Committees which
disperse responsibility horizontally obfuscate the same.
The lack of vertical interaction in the committee form, or
the splitting of a task into various processes, and the lack
of decentralization results in little or no accountability in
government despite hierarchical systems.
Dysfunctional Oversight
The poor performance of various projects, schemes, and
programmes have in turn led to the creation of monitoring
cells and sub-organizations within the government. Their
systems and procedures are no different from the rest of
government and typically do not have the independence
from the line officers. Instead of performance of this task,
the task itself has got modified as ‘reporting’. It has been
ritualized and become an aspect of misgovernance. This
was true of the Department of Programme Implementation
(DPI). The periodic examination of government
organizations and processes can itself get ritualized and
bureaucratized to a point, so that it gets nullified. Then
mere formal changes are recommended, when the need is
for major overhaul, decentralization and incentive compatible
systems, simplification of the task, and ascription of
responsibility. Thus even the ERC whose reports are a great
improvement over earlier reports is itself a victim of the
approach that does not recognize the deep-seated nature of
the problems of the government in India.
Reform recommendations too surprisingly do not
recognize the constrains on governments, or what it can
and cannot do. Governmental structure does not allow the
performance of complex or ill-defined tasks. Part of the
reason for the committees, and the little or no delegation
lie in the poor design of schemes and programmes. Attempts
to do the impossible such as implement a price-based
subsidization with no leakages, or ensure that despite low
prices for ‘priority’ goods there is no under-investment in
the same, create unadministrable situations. Then the task
itself gets displaced to suit the bureaucratic convenience.
When rules cannot give direction, there is the natural
tendency to decide within the `security of a committee.’
The hard task of working against markets rather than with
them, have over the long period, especially from 1965,
onwards, made the battle against rents hopeless. These,
more than any inherent characteristic of Indian culture or
the politics, have tended to corrupt the administration and
render its processes and functioning wasteful.

Public Procurement in India and the Creation of
Market Failure
We had in a different context16 argued that the unnaturally
large space for shoddy goods in the country arises from
the nature of the government’s procurement processes.
Government procurement is based on the belief that contracts
can be complete, and the lowest price is the best way at
which to purchase, irrespective of the nature of the product
or the service. Low price does not automatically translate
itself into lowest costs to the procurer and herein lies one
of the major sources of waste in government expenditure.
As more aspects of the product are discovered, government
agencies tend to include them in the specifications on the
contract, but the contract can never be complete, since in
any complex product or service there is functionality to the
whole that cannot be so easily specified and/or verified.
Then opportunities exist for rents, through formal or ritual
adherence to the specifications, but in the actual supply of
poor quality or less quantity of the service or product. The
usual bureaucratic response to such corruption in India has
been to introduce even more sequential processes and to
involve many more officers laterally placed in the decision
process. This could actually allow rents to continue, despite
the (formal) adherence to the processes. Responsibility
displacement occurs. The huge firm side costs which these
processes then impose, result in good firms abandoning the
queue to sell to the public authority. The public authority
then sees only firms that are unscrupulous and have
developed ways to overcome the costs of going through
cumbersome processes through corruption and influence.
Thus the administrative (state) failure in inappropriate
procurement breeds market failure and that combination
imposes large costs on the economy. Beyond lowest prices
there are interesting methods that can reduce the failure
and give incentives to losing firms to monitor the process.
Thus performance clauses, performance guarantees,
specified evaluation process, independent certification
engineers, revelation of names and identities of bidders
post contract, are just a few of the methods that could be
adopted. There is much that theoretical insights from the
contracting and purchase literature can contribute17.
Private Information Is Important
Blacklisting of poorly performing firms which is part of
any procurement process, even the simplest, is problematic
16 In discussing government purchases of small scale industry
products, and the vendor development policies of many state-owned
enterprises, we had brought out the dynamic nature of state failure.
See Morris (2001).
17 For a discussion of the same and their extension to the case
of public procurement (by governments), see Section 4.2 by Ajay
Pandey in this report.
Expenditure Accountability and the Society 27
in India. Blacklisting is vitiated by the legal loophole that
allows the same person to operate multiple and ‘benami’
firms, and the formal identification of the bidder, as a
particular firm, being allowed to be hidder. Therefore
bringing in private information on the evaluation table,
especially about the true identity of the bidder and his past
performance, into the evaluation process can act as a
major determinant against wilful mis-performance, provided
the same is done transparently.
BUDGETING
As mentioned earlier, the expenditures incurred by a
government have the dimension of location/ and sector of
economic/ social activity. They additionally have the
dimension of the organizational form and the item type.
Thus an item of expenditure could be spent in Gujarat
(location), on the chemical industry (sector) by a central
government public sector enterprise (organization) on wages
and salaries (item). Additionally, it could have the purpose
of supporting R&D (an activity) and may have been granted
by the Department of Science and Technology (allocative
ministry or the department of the government), as part of
the Plan Expenditure.
Plan and Non-plan Expenditures
The economic and functional classification of the budget
is really a two dimensional classification over the item
(economic aspect) and the purpose cum authority (the
functional classification). Budgets with both dimensions
being simultaneously shown are of great value in analysis.
The other dimensions are implicit in the disaggregation
by territory and around the entity making the expense.
These dimensions are lost in the process of aggregation
so that, for the nation as a whole, only the economic and
functional dimensions remain. An aspect of the purpose
though in terms of Plan and non-Plan, is also available.
This classification came about with the Planning that was
put in place in the post-independence period. Plan funds
are those that emerge out of the approved Plan as budgetary
support for the Plan. They would have been originally
whetted and approved by the Planning Commission18.
Thus the funds required to set up a technical institute
would come from Plan funds, when such an institute is
part of the Plan. Once the Plan funds have been expended,
and the institute has been built, the further expenditures
18 In the period before redistributive programmes such as the
Integrated Rural Development Programme (IRDP) etc. came into
being, that is, in the period up to 1965, they were largely derived
of the Plan model, essentially from intersectoral and growth
considerations.
28 India Infrastructure Report 2003

(unless incorporated into the Plan again) would be out of Developmental Expenditure
non-Plan expenditure. The same is true for programmes
Another dimension of government expenditure is in the
that are part of the Plan. Commentators on government
developmental, non-developmental categorization. It is
expenditure patterns have generally decried the decline in
assumed that expenditures on all economic and social
Plan expenditures, as the ratio of Plan to total government
services and on certain programmes will lead to
expenditures have declined. But this may not be entirely
development, and hence, are categorized as developmental.
justified. To the extent that many of the sop programmes
Non-developmental concerns the traditional functions of
are planned but have vast leakages, Plan expenditures could
state—police, defence, interest payments, pensions, and
be quite wasteful when they largely generate rents. Even
food subsidy. Thus many of the sop programmes and some
if rents are not generated, but the design of the programme
producer subsidies are categorized as developmental. The
is ineffective or wasteful, the Plan expenditure as such need
decline in the share of developmental expenditures have
not always be a good thing.
also been bemoaned in academic discussions but there is
But in the sense that there is little discretion in Plan
little basis for the same given the poor efficacy and efficiency
funds spending, decline in the proportion of Plan expenditure
in government expenditures. Equally important is the failure
is indicative of the decline of planning, and increase in
of the state in its vital state functions which no other
the discretionary part of government expenditure.
entities can perform. This is not unrelated to the
underfunding in crucial activities such as the conduct of
Capital Expenditures in the Plan
justice and the operations of the police force. Herein poor
A larger proportion of Plan expenditure (than of non-Plan) performance has imposed vast transactions cost on the
is capital expenditures, and these are seen as having value. economy. Thus, there is really little meaning in discussions
Despite the vast delays and cost overruns, which increase that infer quality of government expenditures purely on
the waste in capital expenditures, and the misdirection of account of share of developmental expenditure.
significant amount of capital formation, it is still true that
‘at the end of day there is usually a “productive” asset19’. Constitutional Transfers
Contrarily revenue expenditures could include maintenance
Transfers from the central government constitute another
expenditure of capital works. The functionality of physical
significant item. Transfers are largely determined by the
capital assets in a transforming economy is without doubt,
finance commissions and their formula /method and criteria
except when there are gross violations of allocative
have generally allowed transfers to state governments in a
efficiency20. This is not to deny the value of human capital
transparent way. The possible vitiation at this level is small
formation. In most areas of human capital formation except,
though the dysfunctionalities or waste would be a feature
perhaps, education, the programmes and expenditure
at the level of spending by the state governments. The same
processes do not result in any significant gain—the so-
can be said of transfers from state governments to local
called special employment schemes, sop schemes such as
bodies. The recommendations of the various state finance
the Prime Minister’s Rozgar Yojna (PMRY), etc., are no
commissions (SFCs) following Constitutional Amendments
doubt part of such waste, arising out of improper spending.
73 and 74, and further developments in decentralization
In education, though, spending on both revenue and capital
have as yet to be institutionalized. In many states the SFCs
result in direct benefits, subject, of course, to the off-item
recommendations have to be notified. On state transfers
leakages, and the efficiency of the education system in
the principal issues relate to the proportion of non-
most parts of the country21. It is debatable whether in
discretionary transfers that is, those that arise out of
areas like Bihar, with near total state failure, even spending
constitutional process—and the ad hoc support that has
on education would result in positive benefits.
increased over the 1980s and the 1990s. The FC-based
19 In Bihar, where the system has systematically vitiated the transfers have been robust to changes in government, and
difference between capital and revenue expenditures to misreport to different parties existing at the state and central levels.
expenditures, a significant part of which would be rents, this cannot So in transfers the mechanisms and criteria as outlined in
be said at all. the FC reports and adopted would be the central basis.
20 These have taken place. An example would be the setting up
Perversities are in the design of the devolution formula.
of the central business district of New Bombay with an occupancy
of 10–12 per cent close to 15 years after construction! Or gauge
conversion with traffic load less than before. Competition for Discretionary Funds
21 For a state’s presentation of the trends and patterns in social
Ad hoc grants are another matter. They, no doubt, have
expenditures see Section 6.2 by Mahendra Dev; and for the
discussions and thinking that precede the Central Budget see Section functionality in disaster relief and in urgent expenditure
6.3, also by Mahendra Dev. requirements which are difficult to foresee. They are not

tied to programmes and are not as deeply embedded in
the expenditure process as the spending out of Plan funds.
There is much competition for these funds. In the best
of cases these funds can be crucial for recovery out of
disasters, and there is scope for making political capital
for the party of the day to claim credit in usage of these
funds. More importantly, they can be more quickly spent
until the bureaucratic processes finally catch up to introduce
delays and the formal processes. At worst, they are seen
as free and discretionary transfers, constituting a way of
favouring the government of the state. State governments
generally tend to exaggerate the damage and destruction
in order to gain more central funds.
Once the disbursement process begins there would be
the reverse tendency to assess the losses more realistically.
As time passes and the real—unpublicized or uninfluential—
victims’ turn comes, the purse would inevitably tighten, as
the bureaucratic processes would have by then caught up.
Given that much of government expenditure is caught in
an interpenetrating quagmire of planning allocation and
disbursal processes, audit and ritual impact assessment,
the distance between the initiating politician or group and
the realization of the objective for the spending is usually
so disjointed, that there is little scope for identification
and, hence, of deriving credit for good public actions.
Shorter political cycles (or life of elected officials and
politicians) and the more distant levers for initiating
expenditures, and the fact that much of it is already derived
out of plans and programmes, has resulted in much attention
and competition for the little ‘open’ funds that exist.
The literature too has been critical of the rise in ad hoc
expenditures. To the extent that through these allocations
particular governments in states are favoured or
discriminated, it is a creeping dysfunctionality. But they
do have the merit of being expended quickly without the
delays and ritual associated with regular government
expenditures. Only with reference to the ideal process of
project and programme expenditure could this distinction
be maintained. It is only in capital expenditures on projects
and social expenditures in areas with vast externalities like
education and health or highly constrained infrastructure,
(despite all the weaknesses of the state) that one can say
that money spent therein is better than money spent
elsewhere. The by far more important problem of the
efficacy and efficiency of government expenditure has been
typically downplayed.
MP’s Special Funds
There is much attention and expectation with regard to
the special funds which the Members of Parliament (MPs)
can disburse for their constituencies. This is not surprising
since there are ‘open’ funds. Here the MP’s discretion is
Expenditure Accountability and the Society 29
far greater, and he does not have to go through the usual
processes of justification and ritual for the projects that
he initiates. It is interesting that despite the small size of
just about a crore per MP, a few MPs have been able to
use the same for initiating significant projects to garner
much political capital without violating processes and
creating rents. These not too numerous examples illustrate
the potential that is there for the political and economic
system to bootstrap out of the mess it is in currently.
Administrative reform that cuts wasteful ritual, vestigial
processes derived out of past plans and needs now defunct,
or out of programmes over the design of which nobody
seems to have any control is the need of the hour. Such
reform would create the space for performance of the
politician and for appropriation of the same as political
capital. Corruption can then reduce considerably.
THE CENTRAL BUDGET AND TRANSPARENCY
The annual budget of the central government attracts
attention as no other economic event does. The presentation
of the budget has become a ritual, a major media event
that lasts for days, and details are discussed and commented
upon in the media in innumerable groups and meetings
that follow. The tax side of the budget receives a lot of
attention. Citizens and associations try to calculate their
disposable incomes and gross revenues. The market also
reacts, although, since the major announcements may have
been anticipated, the reaction is spread over a longer
period. Since the reform process began, the fiscal deficit
has become a household term. Surprisingly, despite all the
limelight, the expenditure side is rarely discussed except
in rather simplistic ways that may amount to mere
verbalization of the figures. The diagnoses of expenditures
are very rudimentary at best.
Bundling with Policy Announcements
The budget comes bundled along with major and minor
policy announcements, as well as announcements of
programmes for spending, subsidy and concessional credit,
of funds to support an activity, or the setting up of institutions
and organizations. These proposals when they are new draw
some attention, but it is the ‘overall thrust of the budget’
that hogs the limelight. Thus the budget is presumed to be
developmental if the outlays on social expenditures especially
education, or on rural development go up significantly. Or
it is considered as pro poor if subsidies are maintained and
administered prices do not rise. The large number of
administered prices on public services and other products
like energy which are heavily taxed makes the budget
proposals a major influence on prices. The fact that the
economy has liberalized a great deal, and the role of taxes
30 India Infrastructure Report 2003
and administered prices have come down22 implies that
the budget should not be as dominant as it used to be.
Other major economic events such as the busy and lean
season credit policy announcements of the Reserve Bank
of India (RBI), or the stance of monetary policy, have in
reality become far more important than the budget in
determining the course of the economy, but do not attract
the attention that the budget does. In Parliament despite
the discussions (which tend to be polemical) there is little
diagnoses of the budget, and in-voice voting on the strength
of the majority of the government in the Lower House the
budget is passed. There are almost no discussions on the
deviations of the actuals from the budget proposals of the
previous years and the reasons for the same. It is only now
that the data has begun to be presented to allow even a
rudimentary discussion on this most vital aspect of fiscal
performance, and expenditure accountability.
Need for a Budget Model
The focus on the fiscal deficit by the financial institutions
and the investing community abroad has recently begun
to draw attention to expenditures and to the prospect of
the government being able to meet its fiscal deficit targets.
Attempts are made to make sense of past slippages,
projecting the trends in the slippage onto the future, so
that the ‘realistic’ likely fiscal deficit can be inferred. Such
discussions are common in academic and financial circles.
Unfortunately, any really meaningful discussion is precluded
because the budget rarely reveals the growth (and inflation)
assumptions (targets) that have been made. Similarly, the
relevant sub-sector growth and targets (assumptions) are
not explicit—exports, imports, industrial, and agricultural
growth. It is high time (and eminently feasible) to make
the budget process and reporting truly transparent, and
allow meaningful discussion to take place. Thus presentation
of a model of the economy of the principal items of the
budget (let us say of 10 major items on the revenue and
expenditure side) is the starting point. The presentation
would have to separate the framework, or the model, from
the estimates of the parameters or the structural variables
used, and the assumed growth rates of the economy of real
gross domestic product (GDP), prices, exports, imports,
etc. Forecasts/assumptions regarding key variables such as
interest rates, exchange rates, would also have to be explicitly
revealed.
With this information it becomes possible to judge the
‘performance’ of the government at least in terms of it
being able to achieve its budgetary targets, including the
22Nevertheless, the informal influence of government on PSUs
and other enterprises keeps the deviation from economic prices
large in some sectors.
fiscal deficit, in a meaningful way. Such analysis is standard
and routine, in management accounting for control and
budgeting in large firms. Herein, the deviation of costs
from budgeted costs or of total sales from its target would
be routinely broken down into variances or factors for
better understanding and control.
It would then be possible to differentiate between lack
of achievement of expenditure or the fiscal deficit, targets
that arise because of regime shifts, and that due to non-
realization of growth targets for example. More generally
ex-post diagnoses and holding the government to account
would be feasible, since the regime that the government
set out to reach and the realized regimes could be inferred.
‘Off-Budget’ Items
Similarly, the changes in the framework or the changes in
the parameters of the model, brought about by the structural
changes in the economy would have to be mentioned and
explained whenever these are accommodated. When final
outlays are announced, the basis of their allocation, and
the macro variable with which the item in question is
expected to have a relationship if any, would have to be
made explicit. For example, the introduction of a new
credit subsidy for small industries, would result in the
expenditure on that count to be expected to be related to
the capital formation in small firms.In other words the
‘budget-makers’ need to share their assumptions, beliefs,
and expenditure frameworks. The Committee on Fiscal
Transparency23 makes similar observations.
Off-budget items like outstanding guarantees, contingent
liabilities and their details, with appropriate tabulations,
would help to focus the legislature’s attention on such items
which have typically been used by governments (the
executive) to get over the need to have the authorization
of the legislature.
REVIEWING THE TRENDS IN EXPENDITURE
Interest payments and debt servicing show the fastest rise.
See Fig. 2.1.1. Budgetary subsidies have had a slower
growth largely because export subsidies were eliminated
over the period up to 1992–3. Since then subsidies have
grown rapidly. Another dampener has been the removal
of universal food subsidies as ‘better’ targeting of the same
was attempted. Social service expenditures show a somewhat
slower growth. See Fig.2.1.2.
23 B1 (2002), ‘Report of the Standing Committee on
International Financial Standards and Code’, under the chairmanship
of Y.V. Reddy, Deputy Governor, RBI. The same has been reviewed
in this report in Box ‘Fiscal Transparency and India’ in Section 4.1
by Ajay Pandey.

Fig 2.1.1 Index of Plan and Other Expenditure of the
Central Government at Current Prices (1991–2 = 100)
Wide Differences
The adjustments of expenditure on account of Plan
expenditure were far more significant than on non-Plan.
When viewed as capital or revenue, capital has shown the
Fig. 2.1.2 Index of Economic Services and Expenditure
of the Central Govt. at Current Prices (1991–2 = 100)
Expenditure Accountability and the Society 31
Fig. 2.1.3 Index of Revenue and Other Central Government
Expenditure at Current Prices (1991–2 = 100)
largest relative decline. See Fig. 2.1.3. The same is true
of total capital Plan expenditures. Capital expenditures
have fallen as a proportion of Plan expenditure. See Fig.
2.1.4. Despite all odds the primary deficit has been under
control having fallen to less than 1 per cent.
Fig 2.1.4 Ratio of Certain Items of Expenditure
32 India Infrastructure Report 2003
Fig. 2.1.5 Deficits of the Central Government and Interest
Payments to GDP at Market Prices
The revenue deficit continues to remain high at around
3.5 per cent of GDP despite the fall in the primary fiscal
deficit. This is largely because of interest payments which
has risen from 4 per cent of GDP to about 4.5 per cent
of GDP. See Fig. 2.1.5. As a proportion of GDP non-
interest revenue expenditures have not risen. Debt service
has kept pace with interest payments because government
has had to reroll cheaper debt taken earlier with more
expensive debt, as the SLRs on banks were sought to be
reduced in a situation of tight monetary policy24. Non-
plan expenditures have grown faster than Plan expenditures.
See Fig. 2.1.1. This in itself is considered as resulting in
poor ‘quality’ of the fiscal deficit. However, because a large
part of Plan expenditures have leakages and waste and
capital plan expenditures involve vast cost and time overruns,
from the point of view of reduction of waste and rents this
cannot always be claimed as poor quality of adjustment.
Fifth Pay Commission Award
The postal deficit rose from around Rs 243 crores in
1991–2 to around Rs 1440 crores in 2001–2, to become
an important item of government expenditure. The sharp
rise since 1996–7 is reflective of the wage and salary
increases on account of the implementation of the Fifth
Pay Commission Award, since much of the expenditures
24 Since the nominal interest rate has begun to fall, this has
begun to reverse.
Fig. 2.1.6 Detrend of Log of GCF in Public and Private
Sectors (GCF at current prices deflated by the GDP
deflator)
in the Postal Department arise on account of salaries and
wages. The rise in the postal deficit would mirror the rise
in expenditures on wages and salaries of the government
as a whole. Prices of postal services rose but not on main
items like inland letters and postcards. Pension payments
too show a sharp rise from 1997–8 onwards. Grants to
states have risen steeply over the last three years. These
as we know have been made to the states to support their
implementation of the Fifth Pay Commission Award, and
to meet their rising deficits as they too began to face the
problem of rising interest rates.
Plan Expenditure and Transfers
From Fig. 2.1.3 notice also that at current prices, there
is a sharp decline in capital expenditures from 1998–9
onwards. Thus the attempt to continue to reduce the fiscal
deficit has also been marked by sharp slow down in capital
expenditures. Since private investments have not shown
any significant rise from that year either, the demand
depressionary effect and the brake on the growth of capacity
output that this imposes is obvious. Fig. 2.1.4 tells us that
the share of capital expenditure in the Plan expenditure
of the centre has gone down sharply. This precipitous fall
from a level close to 50 per cent to around 35 per cent
means that capital formation (even if of the wasteful variety)
is down. The central assistance to the states’ overall plan
expenditures which had been squeezed over the period
 Expenditure Accountability and the Society 33

when the centre itself had made the adjustment (1991–2

to 1996–7), recovered but not fully, from 1997–8 onwards.
Share of central Assistance in Plan expenditure of the
centre had grown very slowly.

Private Capital Takes Over

From Fig. 2.1.6, which gives the detrend of the log of real
gross capital formation (GCF), as a whole and in the public
and private sectors, we see that the period of recovery of
the 1980s was accompanied by a revival of public investments
in the mid-1970s, which went back to the average growth
rate for the period (1950–1 to 2000–1) till 1986–7, after
which the growth rate had slowed down considerably. In
the 1990s the budgetary squeeze on capital expenditures
resulted in the slow down continuing and deepening. Private
investment revival is seen from the early 1980s which
continued up to 1995–6 (very sharply from 1992–3), after
which it has slowed down. Fig. 2.1.7 brings out the shares
of the private sector (the economy other than the public
sector) GCF, and of GCF in the economy. The latter is
at constant prices, and the former ratio at current prices.
Observe that the high growth in the 1980s was also a Fig. 2.1.8 Share of Public Administration and Defence in
period of high rates of investment, and by rising share of GDP at 1993–4 Prices
private investments. Post stabilization the investment rate
rose sharply to reach a level of 29 per cent of GDP! The Recession
rise encompassed the foreign and the domestic, including
This is the point that we had brought out in the IIR
the unregistered sector. After 1995–6 it collapsed sharply
2002, and as early as in 1997 we had called attention
remaining stagnant from that point onwards.
to this stagnation in private investments. (Morris,
1997)25. There is still no recovery of private or public
investments. The latter as much as in 1997–8, still awaits
the regulatory and policy clarity to enter into areas like
power, water, and urban services. Only in roads and
telecom has that clarity come26.

Fifth Pay Commission ‘Boosts’ GDP

Only services have grown, though a significant part of the
growth have access out of the Fifth Pay Commission
Award, which has steeply increased the GDP on account
of Public Administration and Defence. See Fig. 2.1.8,
which brings out the share of Public Administration and

Fig. 2.1.7 Ratio of Real GCF to GDPFC at 1993–4
Prices and of GCF in the Private Sector to All
GCF at Current Prices
25 The point then made remains valid even now: The high
growth of the 1990s, till about 1996–7, had three very important
causes (i) higher growth of exports at 20 per cent per annum in
dollar terms from 1993–4 onwards for four years; (ii) better
agricultural performance at 34 per cent per annum; (iii) reviewing
private investments which more than covered the fall in public
investments, and brought in an additional efficiency factor. Since
1997–8 the exports declined due to the exchange rate getting out
of line. The exchange rate still awaits correction.
26 Manufacturing has little space any more for private capital
investments.
34 India Infrastructure Report 2003
Fig. 2.1.9 Detrend of Log of Macro Aggregates
(1950–1 to 2000–1)
Defence in the tertiary sector and shows a sharp rise in
the Award. See also Fig. 2.1.9. It brings out the detrended
series (of the log of real values) of GDP in the major
sectors of the economy. Observe the ‘Hindu period’ 1965–
9 during which all segments of the economy grew more
slowly than during the period as a whole. Negative slopes,
if one ignores the short period variations, imply slower
than average of growth. The economy turned in 1979 and
the rapid growth of the 1980s was maintained into the
1990s till 1995–6, with the interruption during the
stabilization interlude. Since 1995–6, very clearly there is
a recession on account of both the slowing down of the
agricultural (primary) sector and manufacturing; while GDP
maintains its trend on account of the services sector’s more
rapid growth. This as mentioned before is in a large
measure due to the growth in Public Administration and
Defence.
The fiscal deficit continues to be large because the
elasticity of taxes and revenues is largest with respect to
manufacturing which has all but collapsed, thanks to the
monetary and fiscal conservation. The monetary
conservatism is evident in the declining ratios of the
monetized part of the fiscal deficit, and the slow growth
of money supply especially since 1995–6. Real money
supply growth has gone up over the same period, as
inflation reached very low levels, without giving a boost
to the economy. So now the situation is vintage keynesian,
that would require a fiscal boost, for the economy to
revive. As mentioned in this report, the time for mere
monetary adjustments is over. Private investment is caught
in a quagmire of regulatory and policy anomie, and that
too needs correction. Otherwise the fiscal boost would
have to bring back public investments in a major way.
A more ambitious policy of money growth (interest-rate
targeting mentioned earlier) that lowers the nominal interest
rate and spurs growth (even if today it has to be kicked
off by public expenditures) can actually reduce the fiscal
deficit. It would work through high GDP growth, and
through higher growth on the revenue side; since the
elasticity of government revenue with industrial rather than
overall GDP growth is very high.27 Further expenditure
cutting to reduce the fiscal deficit, is not an alternative at
this juncture, since any politically feasible cuts on non-
interest revenue expenditures would be too small to do
anything worthwhile and would instead affect GDP growth
through the expenditure multiplier side. Cutting of
expenditure to bring down the primary fiscal deficit would
have to await major structural reform including significant
privatization, and regulatory and policy clarity in investment
heavy sectors such as power and water28, that would
guarantee further buoyancy in private investment. See
Boxes 2.1.1 and 2.1.2 for why reform efforts that tend to
bring about efficiency and reduce waste or cut public
expenditures have to worry about the demand side and be
prepared to actively bring about private investments.
SUMMARY
The challenge of public expenditure accountability in
developing democracies is greater than in other societies
that are already developed or have chosen to grow without
firm commitment to democratic political processes.
Competing ideologies in post-independent India have led
to increase in ‘off-budget’ or non-voted expenditures. Many
of these expenditures are initiated in the name of the poor,
but actually benefit many others than the poor. In India,
just before the reforms of the 1990s, there was hardly any
group which did not enjoy some benefit as a result of state
controls and distortionary policies, while the economy as
a whole bore the costs. The vested interests of such
government expenditure are an unintended consequence
of past policy errors.
Large amount of government expenditure on programmes
of the ‘Garibi Hatao’ variety must be seen as creating the
space for significant government expenditures outside
27 We had made the same point in the IIR 2002, and
developments since then have only further strengthened the logic of
this alternative strategy outlined here.
28 We had made this point in the last IIR, and the matter has
become even more urgent today.
 Expenditure Accountability and the Society 35

Box 2.1.1
Structural Reform, Government Expenditure, and Macroeconomic Policy

The waste in government expenditure in India is visible enough for all to recognize and admit it as a major problem. Even those
politicians who actively initiate wasteful expenditures in the garb of public services or some other laudable objective implicitly
admit to the waste inherent in most government expenditure when they themselves oppose similar acts by competing politicians;
typically when out of power. The more general ‘mai-bap’ view of the ‘sarkar’ is a reflection of the fact that the society itself has
been deeply affected by government expenditures and the spoils inherent in the same.
A good starting point to understand the role of government expenditure in the context of rents and waste, is the idea of
‘directly unproductive activity’—a term due to Jagdish Bhagwati. The sudden appearance of renta opportunity in an economy
(for example, due to licensing), would result in windfall gains to the parties able to get the licences. But soon enough if this
and similar other opportunities persist, the competition for the licences (rents) emerges. The rent-seeking activities result in people
‘working’ to earn rents, and such activities grow until the returns in working for rents itself becomes equal to the ‘normal’ returns
in productive businesses, adjusted, of course, for the differences in the risk. Such activities would result in agents and others trying
to influence bureaucrats and decision-makers. Bribes, then, slowly become more generalized, as even petty officials are able to
raise impediments to economic activity. These activities which are avoidable and not necessary to production or consumption
are obviously ‘unproductive’. This is not to suggest that they do not generate incomes. They do, as, for instance, for a typist
or a lawyer hired by an agent who works to get licences or permissions or the employees of a company working in the department
that interfaces with the government to get things done.
Rents and Waste
In other words the rent opportunity, when it stays long enough, converts what was a pure transfer from the people /consumer/
or government to an income generation and earning activity. Therefore, consumption, income distribution, and employment and
production patterns are affected. A significant part of the economy then has either supply or demand linkage with the points
of activities that generates rent. But it is impossible to extricate the one part from the other. Very humble activities, personal
services, clerical work, etc. could actually result out of the competition for rents. The entire economy in its occupational,
production, and consumption pattern is deeply affected by rents.
Waste, Rents, and Transactions Cost
The line between rents and the economic activities of the government, and the genuine transactions cost that any economy incurs
is not very clear cut especially when an empirical estimation of the two are sought. Wallis and Douglass (1996) estimate that some
40 to 50 per cent of US GDP arise out of transactions cost! Not all of this is unavoidable, and the existence of rents cannot be
ruled out. Many would agree that a significant part of the income that arises in the practice of law in the US is really nothing
but returns to the avoidable complexities in the law and the barriers to entry in the legal profession. Others may argue that the
large costs of supporting a thriving legal profession in the US is a necessary price to pay for a vibrant democracy. The unproductive
activities and waste in government expenditures in India, nor the rents and the activities spawned by these can be seen as ‘transactions
cost’ of the economy because implicit in the idea of transactions cost is the notion that in some sense they are unavoidable. b
Reduction in Waste Can be Demand Depressionary
The need to eliminate or reduce wasteful government expenditure and rents is well accepted. But it is often not realized that
their elimination or reduction suddenly could have macroeconomic effects working via the expenditure multiplier. A sudden
elimination of rents that were being earned (because of competition) would in the first instance increase the incomes of the
producers or consumers by the amount of the rent. It is not certain that the spending out of this income (both investment and
consumption) is going to be as large as the spending that took place earlier out of the same when the rent existed and was earnedc.
The result, therefore, on the demand side of wasteful government expenditure, may well be depressionary! Reformers are unaware
of this demand-side effect of large systemic improvements in efficiency that result from the elimination or vast reduction in
inefficiency. The same is true when public expenditure is reduced to make room for private expenditures. When, of course, reform
is accompanied by increases in private spending, especially on account of investments, resulting in the augmentation of both
physical and human capital in such a way that the total (autonomous) spending does not decline, then the effect on the economy
a Rents not in the sense of extra returns to genuinely scarce resources like land and particular non-duplicable skills, but of rents as return
to administrative and other powers and arising out of barriers in the decision-making process.
b This is not to deny that in the very long term when the laws of the land and the institutional processes themselves can change, these can
be considered at least in part as avoidable.
c As rents decline or as wasteful government expenditure declines following a period of reform there would be some changes in the consumption
pattern, since the kind of incomes generated by the rents and wasteful expenditures and their immediate multiplier spendings need not remain
the same as spending out out of sudden rise in incomes of consumers and producers. But this is expected to be a minor effect if adequate levels
of spending are kept up, but rising overall investment rates. It is their purpose that change dramatically. Thus a clerk or a peon employed in
a ‘fixer’s’ office is wasteful but in a factory is productive, and the spending pattern of both need not be very different.
36 India Infrastructure Report 2003

would be to improve the sustainable growth rates. The point that we are making is that the expected rise in private investment
and expenditure cannot always be assumed to be automaticd. Only a hard headed classical economist will be able to wish away
the problem, by denying the autonomous role of aggregate demand in determining the level of output of the economy. Many
reform efforts have faltered or had to be given up, as when the slowdown brought on by expenditure reduction, did not change
into a healthy recovery because nothing else was done to keep up the trends in autonomous and exogenous expenditures. The
fact that most reform efforts have been pushed by economists and policy-makers of a conservative hue has meant that depressionary
pressures have typically followed reforms that attempted elimination of waste and reduction in government spending.e This is
unfortunate. The reduction of government waste and rents when accompanied by rising private real investmentsf could actually
step up the growth rate of the economy. As such the real potential of structural reformg has been masked by the resulting demand
depressionary effects of the very same changes. The key really is to pay particular attention to private investment and, more
generally, to investment. What we have said above holds with particular force in transforming economies, which naturally have
higher investment rates and which owe a much larger part of their autonomous expenditures to investmenth.
Private Investment is not Always Bouyant
It is important to understand why the seemingly reasonable assumptions made that private investment would substantially replace
public investments, could more often than not go wrong. Firstly, private investments may have been held back by policies restricting
their operations in many sectors (now opened up), crowding out may have been a dominant feature for long, so that the private
sector may have declined or stunted sufficiently to not to be able to quickly fill the gaps left by withdrawal of public expenditures
and investments or of efficiency gains caused by the elimination or reduction of rents. Similarly, while sectors hitherto closed may
have been opened to the private sector, in many areas especially infrastructure, the lack of credible policies, and regulatory clarity
would (through high risks) keep the private sector off. The picture of a private sector ever alert to fill the gap caused by restrictive
fiscal policies and structural reform that de-facto eliminates waste and hence reduces spending is a myth in most instancesi. The
story of Indian reform where the savings rate rose and then fell with the investment rate, as private investment after 1995–6 stopped
growing at the high rate required to replace declining public investment and expenditures is illustrative.

d It is interesting that finally what was well known outside orthodoxy is in the popular press, due to the efforts of Stiglitz in his writings
against the IMF. The IMF’s assumption, that, in just starting to grow countries of Africa with weakly developed industries, the net exports would
respond with depreciation, or that private investments would rise if public investments decline, have been repeatedly proven to be wrong. Similarly
their monetary approach that gives equal weight to capital and current account autonomous flows in the determination of exchange rates, does
not make them see the merit of structural undervaluation of currencies to speed up exports in highly diversified poor economies with vast disguised
employment! It is shameful that so many academicians still take the IMF seriously, when they have gone so severely wrong in their predictions.
e More often than not a large part of government expenditure in most LDCs (excluding the East Asian NICs) could be wasteful, and as
such any reduction in government expenditure and in the size of government can have vast efficiency gain effects.
f One of the interesting possibilities is when post reform the investment rates remain the same when measured in current price terms, but
the share of private investment grows. With larger incremental output capital ratios (IOCRs), the overall growth rate goes up. Another possibility
even without an ownership mix change is when the standard graft associated with public capital formation is reduced, which lowers the cost of
investment, so that in real terms the investment level is maintained even if in current price terms it declines. Additional investments to that extent
(which replaces the spending out of rents) can then improve the growth rate of the economy.
g Structural reforms have included direct expenditure reduction by the government, tax reform, removal of rent generating discretionary controls,
lower taxes and subsidies.
h The decline in investments that have followed structural reform have been read as declines in the savings rate and, therefore, of the need
to improve savings rate, when in reality savings have fallen, following the decline in (investment) expenditures.
i A lot more on the sectoral policy and regulatory aspects would have to be done for this vision to be realized.

investment for ‘direct targeting’. This expenditure creates
political support and provides the bases for the party in
power to reward its vast numbers of field workers. For the
bureaucracy it provides a major avenue for expansion and
spread across sectors and regions not constrained by the
need for public investments.
In both the central government and transfers to state
governments and parastatals, expenditure has four important
dimensions. First, the spatial scope, that is, where the
money is spent; second, on what the money is spent; third,
social optimality, that is, to ascertain that the money is
spent on the best possible project which will be useful to
the society. And lastly, the performance dimension, that
is, that the expenditures on the project are in some manner
influenced by the past performance of the entity or
organization proposing the project.
A very large part of the delays and cost overruns of large
projects, especially in the 1970s and early 1980s, before
the environmental factor became dominant, can be attributed
to ‘lack of coordination’ between various sectors or
enterprises implementing the projects. Many of the
administratively and fiscally weaker states who could not
make up the formal project proposals, or put up its
contribution to investment lost out in getting funds for
projects and programmes from the central government in
the last two decades.
 Expenditure Accountability and the Society 37

Box 2.1.2
Efficient Expenditure and Growth

There is another thread of reasoning from the development economics literature that is of particular relevance in the discussion
of wasteful expenditures. In the 1950s the idea of the vicious circle of poverty and the low-level equilibrium brought to the
fore the notion of the big push required to break out the low savings–low investment equilibrium. That justified state involvement
in industrialization, planning, and the state’s role as a mobilizer of resources for investment spending. It also drew attention to
the coordinating role of the state, and of the need for vast foreign resources (savings) to support these efforts that would constitute
the big push above. In raising the investment rate the country would break out of the vicious circlea. Arthur Lewis’b (1954)
contribution to the debate was seminal when he showed that the problem was not so much the lack of (potential) savings as
its non-realization as savings. But the potential surplus is either wasted or exported out. That is crucial to the stability of the
low-level equilibrium trap. Existence of disguised unemployment means that resources are low. The wage is determined by the
average productivity in agriculture. Therefore, investments in modern industries result in vast surplus to potential investors. And
if sustained investments take place the enclaved modern capitalist sector can grow to embrace the entire economy. The rate of
investment, therefore, rises and the economy bootstraps itself.
Investment Rather than Savings Constraint
The existence of disguised unemployment in much of the economy makes this ‘getting out of the abyss’ very fast. In the low
level equilibrium the problem, therefore, is that investments are not being made, the potential surpluses are not being productively
used. Therefore, a lack of coordination (that results in demand being unrealized), inappropriate class structures and other political
and social factors that wastes the surplus, are the problems. Access to foreign funds is not the key to the solutionc.
The point is that rising systemic efficiencies manifest themselves principally in rising investment rates, while keeping the ICOR
from rising, and hence raising the growth rate of the economy. This is particularly true of transforming economies. The point
is that in these economies the transformation phase would show a large part of the growth as arising from input use and very
little from measurable efficiency.d Thus the gains from elimination of waste and rents, and in igniting the process of economic
transformation in late industrializing societies arise from allocative efficiencies and less from efficiencies due to new technology.
They arise from the extensification of modern technology on to the entire economy. The problem is really one of measurement.
As labour productivity rises it really takes place on the back of rising capital use, so that a total factor productivity (TFP) analysis
may still show only small rises in TFP , when both labour and capital are counted as inputs (as they would be in any particular
industry). But the social price of labour is very low—being close to its marginal product which is near zero, so that the only
relevant input for the economy as a whole is capital. And the relevant model to understand the process in its detail is the Lewis
model and its variants, and in the aggregate the Harrod Domar model—not the neoclassical model.
Efficiency Gains Implies Higher Growth
Thus efficiency gains due to reduction in wasteful expenditure and reduction of rents when matched by increases in other productive
expenditure, especially investments, would show up as increases in the growth rate of the economy following from a rise in the
investment and savings rate. Savings would be highly correlated with investments, and savings would get determined by the level
of investments, perhaps with a lag. This discussion was necessary because unlike in the already transformed economy the matter
of growth is different. When growth and its basis is treated as a universal category there can arise significant problems. Growth in
developed economies arise out of technical change, whereas in late industrializers, until they catch up, growth is the result of use
of idle resources—labour. Even after all strictly idle labour is used up, labour productivity can and still does rise as more capital
accumulation takes place. In other words, the measure of a successful structural reform is a sustained increase in the growth rates
and in the investment rates, especially when expressed in real price termse. Thus macroeconomic policies that keep investment rates
high are the key to growth and transformation, and true reform is one that can combine efficiency with high growth.
a Nurkse (1953), Rosenstein-Rodan (1943)
b A little before Lewis, Maurice Dobb (1948), describing the events in the Soviet Union and its economic transition, came to a similar
understanding of the process of change in late industrializing societies.
c Lack of demand could be. Hence the observation that the countries that have used the state and chaebols/zaibatsu like structures to bring
about coordination and overcome market failures and also redistributed land to create endowments (demand) succeeded. The East Asian economies
which did all of these also used export-led growth strategy to create the additional large external demand to very quickly come out of their low-
level traps.
d In all of East Asia this is observed. Typical growth accounting/factor productivity studies reveal that as much as 80 per cent of the growth
can be accounted for by input use, and only 20 per cent at best by improving efficiencies of factor use. Unlike what Krugman (1996b) and
others have led us to believe, there is nothing wrong per se in such economies as a result. Indeed one can argue that functionality would demand
that input use should, during the phase of transformation when idle resources exist, contribute to the bulk of growth. Given that the economy
starts from a situation of vast disguised unemployment, systemic change that lead to realization of potential surplus as investment leads to capital
accumulation that absorbs labour.
e The latter could happen when investment goods prices fall relative to that of other goods, so that at current prices the ratio still remain
the same.
38 India Infrastructure Report 2003

Box 2.1.3
The Burden of Programmes

When we look back now with the benefit of hindsight it is amazing that so much public expenditure for development could
take place without a performance linkage! It is only in the mid-1980s that data on delays and cost overruns were available
anywhere within the government! The DPI brings out in its annual report the data on public projects under implementation
at the central level. The actual task of the department is best seen as one of statistical compilation and ex-post reporting. It does
not have any influence on the process of project implementation.
DPI and Monitoring
The DPI came about out of the need to have centralized information on the vast numbers of programmes under implementation
out of central funds all over the country from the block and village level upwards. The enormous waste and leakage here prompted
the government in the mid 1980s to formally and centrally monitor the same and ‘weekly flash reports’ and monthly detailed
reports became the order of the day as the Prime Minister, Rajiv Gandhi, himself, was concerned with the progress of the
programmes. It is a moot point that despite these reports and the paraphernalia of monitoring nothing corrective is possible since
information alone cannot lead to corrections without any links between the those looking at these reports and the field staff.
Since there was hardly any responsibility embedding—given a little or no incentive compatibility—within the government to
ensure project execution on time, no organization or a part of an organization, could be held to be in charge, and there was
no scope for attributing failure in any meaningful sense. It is also of great significance that nearly all the performance measures
of programmes were in terms of expenditures (inputs) rather than the actual achievement. Achievement in these reports was
nothing more than spending!. This was the way the system (read, the bureaucracy) got the better of a Prime Minister with no
political baggage, and who really wanted to ‘make government work faster’.
Many Studies, No Learning
Earlier the Planning Commission itself had carried out scores, if not hundreds, of ‘evaluation studies’ of IRDP and IRDP-like
programmes. Some of these studies did point out that leakages were large, and the programmes were ill-structured and ill-
conceived, and the multiplicity at the field level made them beyond the organizational capacity of the officials and offices at
the field level (block, taluka, and district). Yet, virtually no changes, (forget innovations in design, organization or delivery) ever
took place. The only new factor has been the attempt to involve non-government organisations (NGOs). Here, too, except when
there was a rare NGO deeply committed and willing to spend much manpower and quality leadership and organizational
resources, programmes have continued much as they were with the added expense of NGO involvement, and ritualized
assessment, capacity building , participation, and training.
Ritual Studies
The bulk of the Programme Evaluation Studies (PES) related to IRDP-like schemes were ritualized studies without a critical
framework. Indeed their ‘framework’ was such that the possibility of coming to the conclusion that performance was poor was
ruled out by the framework of the study! In internalizing the administrative categories, and in accepting expenditure per se as
a measure of the output of these programmes, they were mere tautologies, in the form of an evaluation study. It is interesting
that even the bulk of the independent and academic studies too have been similarly flawed. An entire research institution the
National Institute of Rural Development (NIRD)—is caught in the ritualized assessment of government schemes arriving at inane
and meaningless conclusions, and correlations between performance and social demographic characteristics. The organizational
economic and financial aspects were usually completely ignored or brought in, in a way that could not have led to a diagnostic
form, from which actions for correction could have been derived.
From the rare critically-oriented studies by independent and perceptual scholars we know that on the whole the employment-
oriented schemes (like the Food for Work Programme and the Employment Guarantee Scheme) had rather better success than
asset formation schemes (small businesses, bicycle repair shops, loans to buy milch cattle, etc.).
Continual Redistribution is Problematic
The whole idea of directly targeting the poor, that is, continual redistribution through the state can be fundamentally critiqued
(Morris, 1991 and 2002). We have already mentioned that countries which did a one shot redistribution through land reform
ensuring basic endowments to all, went on to industrialize in the twentieth century; and theirs were the only truly independent
and successful transformations in this century. In contrast it is easy to see why the continual redistribution would not work.
Income distribution is one of the stable characteristics of any economy and this can change only slowly (than even economic
structure). There are obviously critical minimum changes for durability of the change. A small change in income distribution—
few poor (about one in a hundred) every year being ‘lifted out of poverty’—would not do much. The inertia and logic of the
existing structure would reassert itself in many ways; reversion of those ‘lifted up’ would only be one aspect. The actions which
do not affect the existing power and economic relations would not be sustainable, since they do not make for systemic consistency.
The ‘system’ would incorporate such programmes into the existing relations through leakages, influence over beneficiaries, slow
 Expenditure Accountability and the Society 39

upward mobility of the beneficiaries, negative feedback effects as when one family below the poverty line is lifted up another
falls below! The principle of Le Chatelier would apply with vigour in this situation. Charity has never worked to reduce
widespread poverty, and the design of IRDP-like programmes, despite their apparent micro level productive aspect, need not,
at the meso or macro-levels, be consistent. The same may be true of training in activities for which demands do not exist or
are slowly growing or where barriers to entry are large. At any time what is being done is only a drop in the ocean.
‘Building’ on Weaknesses
There are other, perhaps more, weighty reasons from the supply side that would restrict the efficacy of such programmes. They
need to be continuously implemented. That means that they are crucially dependent upon an efficient and committed administration,
all the time. They cannot tolerate even small levels of state failure. Mistakes that result in leakages, modes of behaviour that are
dysfunctional continue to be reinforced. If the object was really redistribution, in a way that improves the productive potential
of the economy, there are other obviously more efficient and simpler ways of doing the same such as through food for work and
infrastructure based on food stamps that could have created employment, demand and productive assets. These at the same time
would not have demanded the efficiency and commitment of the administration continuously. Obviously, the real functionality
of these programmes is ‘sop’ and the neutralization of potential dissent among those without or with very little endowments.
Nearly all these programmes were based on continuous and repeated identification of the poor; and the task of identification
and issuance of the benefit—be it credit grant or training—being simultaneous, there was little basis or incentive to do the job
of identification thoroughly. But had the two been separated and the status of an individual /family identified once (for possibly
as long as a decade), then a queue that could self-monitor could have been built up. Transparency and public notification of
the beneficiaries could have contributed much to correct identification. Thus, even the possible gains in efficacy, and efficiency
possible within the paradigm of direct targeting through better design went unrecognized. See Appendix 2.1.1 for a very partial
but illustrative list of these and other programmes that attempt to ‘remove’ poverty and improve the living conditions of people
through schemes that have continual administrative overload. Notice that there is very little change in the form and structure
of these schemes. Schemes change names, the content remains the same. They merge into other schemes. Not just the intended
beneficiares but even few officials would be able tell or list out all the schemes that operate.
The budgetary and expenditure processes were problematic. The funds having to pass through various rungs of the
bureaucracy, those who designed the projects, were distanced from the officials who were to actually implement the projects.
The latter could hardly be held responsible for its failure, since the design was not theirs and responsibility embedment cannot
be assumed. Moreover, there are far too many schemes with different names and accounts and differing in marginal ways—the
asset or the purpose—but in most cases really amounting to some commercial credit or grant being delivered. The total number
of schemes would overwhelm the local officials and their infrastructure if they have to transparently and carefully implement
all the schemes. Failure was virtually guaranteed in the design.
Direct Transfer Schemes Fare Better
The schemes which attempted direct transfers have fared somewhat better. Typical of these have been widows’ pension, old age
pensions, and scholarships for poor people. In societies with much local awareness of policies and programmes, with somewhat
more equituous income distribution, and with the capacity among people to demand accountability from local officials (Kerala,
and, possibly, in parts of the southern and western states) a significant part of the spending would have gone to the intended.
In areas with much state failure, and no great local-level movements or consciousness among the people, such sums would have
been spirited away almost entirely as rent. In still others with state failure but with strong local consciousness there would be
a better chance for the same to reach the people since the activity being a mere transfer is not dependent upon the capacity
of the state to design and execute a complex task (programme). The accountability measure too is simple. The repetition in the
task gives incentives for people (recipients) to monitor the local officials.
The output of these programmes have been measured by the amount of expenditure incurred in these programmes. Thus
through a particular poverty eradication project a certain number of people would have been claimed to have been lifted above
the poverty line, since so much money was spent. This would obviously be a false measure. In many others expenditure itself
being the ‘target’ and the measure, the real purpose such as raising incomes, or improving nutrition would have been nearly
completely suppressed in evaluatory studies except in the assumed design sense.
Abhorring Output Measures
This is reflective of a more general tendency to use expenditure itself as a measure of government activities. At the accounting
level/information gathering level this is natural to the way the national income accounts treat purely public services, and the activities
of government. Here lack of excludability, which is total, and non-existence of markets make it infeasible to use any other method—
police and defence—make other measures that are also capable of being aggregated problematic. The GDP here is the expenditure
on wages and salaries and use of capital. The non-recognition of the performance-value axis of government expenditure means
that there is very little effort to find other measures including throughput measures (which may be possible when output measures
are not). It is also reflective of the lack of accountability in governments. Accountability itself is negated in terms of the primary
task when output measures do not exist. The only possible accountability then is to the rules and process of expenditure.
40 India Infrastructure Report 2003

True output measures would mean that the allocator /dispenser himself is being evaluated. The weak organizational separation—
‘after all it is all within government’—brings little or no pressures to search for true and possible measures of output of each
distinct department, activity or programme. Given such non-measurement of output, and when expenditures themselves are used
as output, there is little or no possibility of hardening of budgets. Practically then, only incremental budgeting is possible. The
search for output measures can be much resisted by bureaucracies, since they constitute an important element of change towards
greater functionality and task orientationa.
Thus output measures have come with great difficulty. And audit and accounting innovations like the value for money audit
of the British Government, which are functional rather than being purely procedural have had to face much resistance.
Discussions Ignore the ‘How To?’
Surprisingly, despite being studied for decades, this kind of state intervention through ‘targeted’ programmes, that are heavily
dependent on high levels of motivation and organizational capacity within government, have been routinely put forward as
‘solutions’ to all kinds of problems even by academics who clearly have no brief to restrict themselves to the governmental paradigm
in their thinking of alternatives. It is the ‘public policy orientation’ of the economists—the assumption that the state is an agent
of the common good of society,that is at the root of the problem. (For example, few non-right wing scholars have argued for
the abandonment of the PDS and sop programmes despite their vast leakages. They have typically asked for better efficiency
(how one brings that about was blissfully never their concern) and increased outlays. The public policy stance does not allow
us to recognize state failure that compounds market failure—and the latter that may well have been much smaller to start with.
The issue of ensuring incentive compatibility, organizational consistency and capacity with government are most remote from
the minds of such scholars, who unfortunately are well entrenched in the centre and left of centre (political) spaces. Similarly,
the right wing, in insisting on the removal of all subsidies and abandonment of all state extension, can be challenged to show
that alternative methods of subsidization and extension by the state can bring about major improvements to greatly reduce the
cost of subsidization. Today in many sectors the cost of one rupee of delivered subsidy can be as high as three or four rupees.
In the case of electricity the costs may be even higher, if the consumer side costs on paying customers, and the specific externalities
that improper use of electricity engenders on the firm are recognized. In some IRDP-type programmes, for instance, tribal
development programmes it may be as high as 50 times. In irrigation water if the externalities and cost of in-optimality in the
cropping pattern, wastage of water, and lack of use of water saving methods and technologies, etc. are included then they could
be very large and many times the value of the delivered subsidies.

a The experience of the DFID and the World Bank working with certain state governments to bring about expenditure accountability, reveals
the difficulties and the resistance in using output/more rational measures of performance, than expenditure itself. See Chapter 4.3 by Shantanu
Mitra and Vijay Pillai in this Report.

Multilateral agencies funding large infrastructure projects
have the comfort of the backing of the central government.
And these projects are accompanied by close monitoring
and conditionalities relating to performance. But there was
no attempt on the Planning Commission or Finance
Ministry’s part to impose conditionalities relating to
performance, pricing, etc. The lack of performance linkage
at the project levels were inter alia because:
• The Planning Commission in its expenditure
allocation and management aspect was entirely an extension
of the bureaucracy and, therefore, its evaluation and
examination was of particular tasks rather than a holistic
evaluation of the project.
• Unlike a developmental bank, there is no possible
aggregate measure of performance or activity for all the
projects under the Planning Commission.
• Follow up in government bureaucracies to moneys
given is difficult, and capital investment carried out by the
Planning Commission was no exception.
• Plan fund grants were not sought to be sharply
distinguished from devolved funds. So the idea of the
centre as a mere router of funds could have spilled over
on to the Planning Commission’s role.
• There was incentive incompatibility in bundling the
planning and project /programme roles together.
While the political reasons for the dismal state of public
expenditure accountability are recognized, the bureaucratic
and systemic forces that result in slack and waste are not
admitted as independent causal factors. Dilatory processes,
very low work loads on administrative staff and ill-defined
tasks give rise to poor implementation. The general belief
that political interference is the cause of administrative
failure and inefficiency is only part of the story.
Budgetary subsidies after 1992–3 have grown rapidly.
At current prices, there is a sharp decline in capital
expenditures from 1998–9 onwards. The attempt to
continue to reduce the fiscal deficit has resulted in a slow
down in capital expenditures. One would have assumed
that private expenditure would continue to rise, but after
1995–6 it has collapsed sharply remaining almost stagnant
from that point onwards. The lack of policy and regulatory
clarity in many investment-heavy sectors of infrastructure,
 Expenditure Accountability and the Society 41

Box 2.1.4
The Fiscal Deficit

The fiscal deficit has emerged as the one single shorthand way of assessing the progress of reform. The Fund/Banks’ approach
to reform which sees bringing down the fiscal deficit continuously till it reaches very low and sustainable levels, has swamped
the attention space so much that there is virtually no discussion on this objective of reducing the fiscal deficit by expenditure
cutting in policy forums and in the media. It is therefore important at the risk of being trite to go over the issue. The argument
about the fiscal deficit essentially emanates from two sets of considerations.
Sustainability and Identities
One is sustainability and the other the greater efficiency of private rather than public expenditures—at least at the margin.
Sustainability, that is, the consideration that the fiscal deficit as a proportion of GDP should not keep rising implies a ceiling
on the level of the primary fiscal deficit under assumptions of desirable or ceiling levels of inflation and monetization. The
sustainable level of the fiscal deficit is essentially arrived at by using accounting and definitional relationships and identities, and
certain ‘data’ covering structural ratios and assumptions about growth, which are considered as being unaffected by the level of
government expenditures, the fiscal deficit or money supply in the economy. Inter alia these are the money multiplier, the money
GDP ratio in the economy, the real interest rate, and most importantly the growth rate of the economy and its exports! Assumed
maximum ‘realistic’ levels of growth, constancy in the structural ratios then give the sustainable (maximum) level of the primary
fiscal deficit. Thus the maximum level of the primary fiscal deficit in 1991–2 would have been 2.3 per cent for the entire NFPS,
thus necessitating major change since the primary fiscal deficit was running at close to 5.5 per cent (Joshi and Little, 1996).
An Inadequate Framework
The same identities but with different assumptions about the growth rate of the economy and the growth rate of exports, the
tolerable inflation and the real interest rates, would give vastly different levels of the fiscal deficit that is sustainable. See Table
2.1.1. Thus the accounting framework can at best be a crude guide useful in situations of crisis (crowding out with high inflationa)
when the need to cut the deficit by reducing expenditure is obvious, but of little use at other times. This is because the
government expenditures and the money supply which go into the computation of the sustainable primary fiscal deficit can have
large effects on growth. In other words the real sectors cannot be assumed to be independent of the monetary sector, nor can
they be considered as truly exogenous in the conceivable long run. The matter is particularly relevant in the transforming
economies where monetary and fiscal policies can make a big and lasting difference to the growth rates. The keynesian percept
to use these when there are idle resources is a real opportunity in the transforming economies. This is because the correct
composition of expenditures that raises the investment rate, not only shifts the equilibrium (market clearing) level of output but
also raises the underlying productive capacity of the economy. The existence of disguised employment can in such economies
(when wages are un-indexed or only in part indexed) result in high growth without bringing about inflation b. In short, in
transforming economies as long as rising expenditures take the form of investment (or more correctly the investment rate does
not decline, but rises at least as fast as the incremental (design) capital output ratios), the products of which have markets at
home and abroad (exports), then the resulting growth need not invite the danger of inflation c.
Crowding Out and the Fiscal Deficit
The second reason for worrying about the fiscal deficit is the possible crowding out of private expenditures it can result in. This
is a more serious ground for reigning in the fiscal deficit even when they are sustainable. Unfortunately, even here there cannot be an
unambiguous answer. That there are two sides to any expenditure is not often recognized. All expenditures including the most
wasteful and abhorrent contribute to demand and, hence, towards determining the equilibrium level of output. Sudden falls in
the autonomous components of expenditure lower the equilibrium level of output, and can therefore keep up the fiscal deficit as
a proportion of GDP if the resulting output fall is large. The other aspect is its impact on the productive capacity. When the
expenditure is investmentd then there is a direct addition to the productive capacity. Private investment when it replaces government
investment that had involved much waste and cost overruns can result in faster growth of capacity output. See Box 2.1.1.
Economic Structure and the Fiscal Deficit
There would be situations when if the fiscal deficit is attempted to be cut beyond a certain level by contracting public
expenditures, it would not be matched by private expenditures. In that situation the fiscal deficit is necessary at least temporarily.
It cannot be brought down without the pursuit of structural reform, which enhances the space for private sector and private
a More correctly when the current equilibrium level of output is above the capacity output of the economy.
b Inflation can be expected to be directly related to the gap between capacity and equilibrium output.
c High and unsustainable inflation in the LDCs in Africa and Latin America have their primate cause in the capital flight that occurs from
these economies, and the causation is not the other way round as is generally assumed.
d In economies that are skill constrained (perhaps not India) expenditures that increase the availability of skills directly required in industries
can also act like ‘investment in capacity creation.’
42 India Infrastructure Report 2003

expenditures. Thus large-scale privatizations (sell off ) in such a situation would be required to reduce the fiscal deficit further.
These structural limits (that is, the constraints posed by lagging structural reforms to macroeconomic and fiscal reforms) to cutting
the fiscal deficit are not generally recognized. Attempts to reduce the deficit when the structural conditions still warrant large
public expenditures and deficits (as, for instance, when major investment-hungry sectors such as power, urban services and water,
roads and railways remain with the public sector) would be problematic. Orthodox pursuit of fiscal deficit reduction even in
such a situation would only mean the imposition of demand recession on the economye, and the fiscal deficit as proportion of
GDP may well remain at the original level despite the expenditure reduction.
It could even increase it, if certain revenues like taxes have high elasticities with respect to outputs that show much amplitudinal
variations.
Growth the Only Way Out
In India inflation fell only after a gap of about four to five years that is, in 1996–7, after the expenditure reduction and switching
had begun in 1991–2. This is as expected by keynesians, and is observed all over in market economies, even in the advanced capitalist
economies, where price stickiness is bound to be less.f During the period up to 1995–6 when inflation remained high, at around
8 per cent and nominal interest rates were 14+ per cent, the real rate was 7 per cent and was expected to decline as inflation declined
by the less orthodox who believed in the law of one price in a more long-term sense. But even then the expected fall in the nominal
rates did not happen. Today although the nominal rates have fallen to 10 per cent or below the inflation too has fallen to about
5 per cent or less leaving the real rate at around 6 per cent! For the manufacturing industry where the real rate based on its own
inflation at 3 per cent or less is as high as 7 per cent, the growth rate of output is barely 4 per cent. That virtually rules out any
fresh investments other than of the maintenance varietyg. At real rates in excess of the growth rates the budget deficit can only rise!
The answer, of course, is not further cutting of government expenditures as is the conventional belief in policy making circles, but
higher growth. Indeed, the sooner the government realizes this the better it is. The only way the fiscal deficit can come down today
is either through inflation that brings down the real cost of borrowing below the government’s growth rate of revenue or through
growth which increases this revenue, and reduces the base for the fiscal deficit.

e Which we had anticipated much before the present recession. (See Morris, 1997).
f Recent evidenceand thinking would indicate that wage price stickiness may not be the real basis for the delayed response of monetary policy
on prices but the very nature of the productive system of monopolistic competition.
g The fischer-open or the unarbitraged difference in interest rates too remaining positive, would also additionally imply a capital cost disadvantage
to domestic firms currently.

such as power and water, are no doubt responsible. In
other areas, especially in public infrastructural services,
there are opportunities to leverage private investments that
continue to be missed.
In the annual budget of the central government the tax
side of the budget receives more public attention and the
expenditure side is rarely discussed except in rather simplistic
ways that may amount to mere verbalization of the figures.
The diagnoses of expenditures are very rudimentary. The
reason for this disparity is that the budget comes bundled
along with major and minor policy announcements, as well
as announcements of programmes for spending, subsidy
and concessional credit, of funds to support an activity,
or the setting up of institutions and organizations. This
exercise of allocation of government expenditure is an end
in itself as this provides authority to respective ministries
to spend the public money. There are almost no discussions
on the deviations of the actuals from the budget proposals
of the previous years because efficiency of public expenditure
is generally accepted to be low. Fortunately, the data is now
presented in a little more detail to allow a rudimentary
discussion on this most vital aspect of fiscal performance,
and expenditure accountability.
The government had in 1997 set up the Expenditure
Reforms Commission (ERC), which has submitted its
various reports. The objective of the Commission was
expenditure reforms and the Commission adopted a simple
approach of identifying and recommending for abolition/
removing surplus manpower and expenditure. But, if one
reads between the lines of the ERC reports, it is possible
to gleam some of the crucial problems with governmental
processes today.
The range of possible changes that need to be initiated
is vast—active privatization, incentive compatible
contracting, income rather than production subsidies, direct
rather than price-based subsidies, independent and light
rather than government initiated and heavy (cost plus)
regulation, innovative ownership forms to bring about
incentive compatibility. All of these instead of requiring
all of the state to be well oiled, informed and motivated,
require only a part of the state to be so since, once decided,
the policies and frameworks would tend to carry the
process of reform and efficiency further. To have the right
policies and frameworks is to ensure implementation, and
there cannot be a hiatus between the two. This we had
argued at great length in the last IIR. Being too orthodox
 Expenditure Accountability and the Society 43

Box 2.1.5
Employment and Administrative Expenses

Trends
Employment can hardly go down suddenly in government, and certainly not in India, where the government by past practice and
by the law cannot follow a policy of hire and firea. To a large extent, therefore, employment costs are due to the actions or non-
actions in the past. And current commitments to reduce expenditures and employment even when seriously pursued would have
only delayed effects. The employment growth in the 1980s in the public sector as a whole is worth recounting. In the private sector
employment growth was barely positive, despite the fact that the economy grew at rates close to 5.5 per cent per annum on the
whole, and the corporate sector at a couple of percentage points better. The reasons for this ‘jobless’ growth as it was then widely
called were rather simple. During the period of the late 1960s growth had been very slow, but labour militancy high, and managers
may have all but lost control over the shop floor. As the growth picked up, the impasse of the 1970s gave way to productivity
linked wage increases for wage labour, first in Mumbai and then elsewhere in western and southern India. This brought down
militancy and restored the control of the shop floor to the management. So the vast slack and over-employment of the past could
be gradually overcome in the 1980s by improving labour productivity through better control and by use of more capital. There
was therefore no real demand for fresh new workers. But when the growth continued into the 1990s after a short period of two
years of interruption during the stabilization, employment growth began again but on a very modest scaleb. In the public sector
though, the trends are revealing of the non-economic processes at work. Right through the end of the Mahalanobis Plan (the mid-
1960s) to the structural adjustment of the early 1990s, the employment growth rate has been at a steady practically unvarying
growth rate. The vast variations in output, has had little effect on the employment growth rate. The over-manning and over-
employment during the ‘Hindu Period’ 1965–79 had no impact at all on the 1980s when growth revived and shop-floor conditions
were better. Employment continued to grow and there was no taking in of the slack, that the private corporate sector witnessed.
Public Employment in the 1990s
After 1991–3, it was investment in the public sector that bore the brunt of the expenditure reduction that took place. As part
of the strategy and diktat there was a complete freeze on employment by PSUs. This was administratively hoisted on the PSUs
and, later extended to the general government. Employment growth rates turned negative, given zero new employment and the
natural attrition due to superannuation and resignations. So the benefits of slowdown and decline in employment growth at the
PSU and central government levels would only now be witnessed. Practically any TFP study of PSUs during the 1990s should
show strong productivity gains by the central government enterprises as a whole as their outputs rose in keeping with the growt h
of the economy but capital investments were either frozen or at very low levels and total employment declined. The huge gains
in productivity only tell us the enormous levels of slack that existed at the end of the 1980s in the PSU sector. The slack that
would have been there in the 1970s is mind boggling. By no stretch of imagination can the slack be said to have been roped
in even today! At the shop-floor level the slack today would be much less, but at white-collar levels it continues to be enormousc.
Effects Lag Actions
Let us consider the period of the 1980s when employment growth Eg was 2.5 per cent or more. An inflation level Ng of about
6 per cent, and a wage increase due to promotions /skill improvements, and a bracket increase Sg of 2 per cent is easily assumed.
If there is also a real wage increase (for the same bracket) Rg of 4.0 per cent then, take or leave a percentage point, the nominal
growth of salaries and wage expenditures Ng has to be of the order of 14.5 per cent.
Thus: Ng(14.5) = Pg(6.0) + Eg(2.5) + Sg(2.0) + Rg(4.0). If salaries and wage expenses of governments as a proportion of GDP
ought not to have risen, then the economy should have grown at rates in excess of 14.5 – 6 = 8.5 per cent in the 1980s. This
did not happen so the ‘adjustment’ was by rising inflation rates and delay in the implementation of partial indexation of
government employees wages to the general inflation. But the commitment to partial indexation while it could be shelved for
a while could not be indefinitely postponed and the implementation of the Fifth Pay Commission in the 1990s has suddenly
brought on to the government and PSU budgets major increase on account of salaries and wages.
Today the situation is characterized by a slower growth in the salary and wage expenses (excluding superannuation payments
and benefits) as follows: Ng(9.5) = Pg(5.0) + Eg(–0.5) + Sg(2.0) + Rg(3.0) which given an inflation of about 5 per cent implies
only a real rate of GDP growth of 5 per cent for the ratio to remain steady. As such the ratio can be expected to get steady
and would most certainly decline if the growth rate of the economy can be stepped up. The state governments not having as
yet reduced their employment growth rates to nil, would take much more time before the ratio stabilizes.

a It is not just the government proper but even enterprises and autonomous institutions that are so handicapped, because when the governments’
stake (or contribution) is 50 per cent or more, they have been ruled as state by the courts.
b In some new industries like the financial sector, IT and consulting and management services, all of much interest to professionals, the
employment growth rates were at record high levels.
c The wide spread use of IT and personal computers would have enhanced the slack even more.
44 India Infrastructure Report 2003

Box 2.1.6
Financial Sector Reform and the Fiscal Deficit

Early and simultaneous pursuit of financial sector reform raises the interest rates for the governments’ fresh borrowings. The
irrelevance of sequencing is based on the assumption of the law of one price, which actually is a mere conceptual abstraction. Indeed
the one thing that empirical investigations have revealed is that the law of one price is systematically violated. (Stein, Jerome, 1991)
‘Law of One Price’ is no Law
When financial sector reform in the key sense of market determined interest rate is pursued before or simultaneously with refor m
of the real sector, there is potential for the real side of the reform to be adversely affected, or to become more difficult. Firstly,
it is important to recognize that, unlike the widespread belief (essentially classical) of the interest serving as the price that brings
equilibrium between savings and investment, the nominal interest rate serves as an indicator of developments in the bond and
money markets. This essentially keynesian insight, would similarly tell us that there need not be any tendency, and certainly not
in the medium term, for the ‘law of one price’ to hold, that is for a constancy of the real interest rates. Constant real interest
rates is not something that emerges out of markets, but can be the result of a systematic and active pursuit of the same through
fiscal and monetary policy.
A period prior to reform when the public sector/government would have had access to savings funds at low interest rates (real-
nominal rates that are fixed and inflation being high) by bypassing the market. When it goes into the market for borrowings
the nominal rates would go up. At least till such time as there is downward pressure on money supply growth—given the need
to curtail expenditures—the interest rates would go up rather rapidly. Even when the price level falls or growth in the same
declines due to the delayed effect of restrictive monetary policy, the nominal rate need not fall since that would depend upon
the money supply growth. Only if investments in the rest of the economy decline and if the government is quickly able to shed
expenditures would it decline. But then the economy necessarily adjusts by deep contraction. If contraction beyond that required
for balance of payment (BoP) adjustment does not take place because private investments rise to keep the trend in investment
expenditures as a whole the same as before, interest rates would not decline. On sub-national governments, and parastatals and
local bodies with large borrowings from the centre and the markets, the fiscal deficit would widen significantly on this count,
and could even negate their efforts to reduce or slow down other expenditure growth.
Sequencing is Important
Many have, therefore, suggested the late pursuit of financial sector reform in adjustment. (Joshi and Little, 1996). The continuing
financial repression in such a strategy means that the burden of adjustment is shifted not only on the government (and on wage
and profit earners) but also on interest income earners. Since interest income accrues to accumulated financial savings, a small
adjustment therein can make room for much reduction in the adjustment required on the part of the productive enterprises
adding to real capital assets. This shift from interest earners (risk shirkers) to profit earners (risk takers) can speed up or maintain
the growth rate of the economy.
Indeed in such a delayed (or as many would call a correct sequencing of reforms) the fiscal deficit can actually be brought
down to very low levels as a proportion of GDP, since it is a strategy that builds on growth—the only ultimate basis for all
adjustment. The lower interest rates would mean that savers and investors in fixed income securities lose on their savings, but
they gain through the rise in opportunities for investment, and especially for such investments outside the government.
Unnecessary Burden on the Real Sector
Thus in India the early pursuit of financial sector reform resulted in vast and sudden income transfers via high interest rates
on to the banks (who in the repressed period held government securities) and to ‘small savers’ from whom government had
borrowed often at fixed tenures and rates. The key keynesian insight that the savings rate did not depend upon the interest
rates either nominal or real, but that savings is a residual aspect of the consumption function, implies that the possibility that
the real interest rates would affect savings is rather small.a Indeed low but positive real rates maintained through both price and
interest rate targets, the former in the short run and the latter in a more long run sense, even when repression is known to exist,
can result in vast increases in incomes and savings out of these, as the investment rate goes up. All successful East Asian
industrializers have shown this functional financial repression,b along with high and rising investment and savings rates. In short,
therefore, the fiscal situation and specifically the fiscal deficits of sub-national entities would have to be judged with reference
to the macroeconomic situation and the strategy of adjustment and reform pursued by the national government. To a large extent
the sudden and dramatic deterioration of the budgets of state governments, is due to this factor, compounded by the sharp rise
in salary and wage expenses due to the Fifth Pay Commission Awards.

a Except when they are permanently negative! This situation is better dealt with by the pursuit of policies that reduce the inflation rate below
the nominal rate of interest.
b The idea that ‘non-equilibrium’ low interest rates results in inefficient technology choice, in a growing economy is pure classical fib. But
such arguments without caveats have been made by the “financial repression” school led by Fry, Maxwell (1980).
 Expenditure Accountability and the Society 45

‘Small Savings’
The Y.V. Reddy Committee on Administered Interest Rates looked inter alia at ‘small’ savings policies, and made many important
suggestions. For a critical review of the same see Box 2.1.7 of this chapter by S.K. Barua. The idea of ‘small savings’ to raise
the savings rate through mobilization of savings among very humble and poor people makes eminent sense in a poor but
transforming economy. The small additional benefit in the form of a higher rate of interest is also worthwhile, provided there
are limits to investments in small savings instruments, so that small savings remain small in the sense of being used by people
with small incomes. But a transforming economy that seeks to step up growth rates through high and rising rates of investments
would need to have an income policy that allows government an additional instrument (including the rate of interest) on certain
designated instruments into which savings could be mobilized (and de-mobilized), depending on the situation with regard to
the price level. Thus, if through an income policy that can postpone or retain consumption, when inflation tends to rise, and
vice versa, high rates of investment can be pursued. This had been proved by the East Asian economies.
Additionally, it may be worthwhile to insulate a certain nominal level of financial savings per person from inflation by creating
bonds whose yields are indexed to inflation. This can make the system bear somewhat higher levels of inflation. They could
substantially raise the savings rate among small-income earners. A complete determination of all interest rates by the market
through dependence or indexation with particular markets or instruments like the bank rate as the Reddy Committee suggests,
may impose major uncertainty and volatility on not only small savings but also on entrepreneurs and firms taking risks to make
‘real’ investments.

Table 2.1.1
The Primary Fiscal Deficit: An Accounting Exercise
(Assuming that nominal interest rates change to adjust to inflation and real interest rates)

Primary fiscal deficit/GDP x 0.0236 0.0344 0.0519 0.0543 0.0675

Domestic debt GDP ratio b 0.54 0.54 0.54 0.54 0.54
Safe seniorage/GDP s 0.0132 0.0132 0.0132 0.0156 0.018
Safe govt primary borrowing abroad/GDP f 0.005 0.005 0.0225 0.0225 0.0225
Nominal interest rate domestic i 0.1 0.08 0.08 0.1 0.1
Domestic inflation rate Pg 0.05 0.05 0.05 0.05 0.07
Growth rate of real income Yg 0.06 0.06 0.06 0.08 0.08
Ratio of high powered money (H)
to GDP nominal H 0.12 0.12 0.12 0.12 0.12
Ratio of national foreign debt to exports D 2 2 2 2 2
Exports growth rate Eg 0.1 0.1 0.2 0.2 0.2
Foreign dollar real interest rate r$ 0.05 0.05 0.05 0.05 0.05
Exports to GDP nominal ratio E/yp 0.1 0.1 0.15 0.15 0.15
Govt share in national primary
borrowings abroad 0.5 0.5 0.5 0.5 0.5
Domestic real interest rate r 0.05 0.03 0.03 0.05 0.03
Alternatives Joshi Joshi Joshi Joshi Joshi

+r = 0.03 +Eg = 0.2 +Eg = 0.2 +Eg = 0.2

+E/yp = 0.15 +E/yp = 0.15 +E/yp = 0.15
+r = 0.03 +y~ = 0.08 +r = 0.03
yg = 0.08
pg = 0.07

about reform invites political resistance and does not lead
to a feasible path forward. This is rather silly when there
are solutions to the principal problems that are not only
feasible politically but actually bring forth additional political
support for the reform and improved working of the
government and public services.
The scope for privatization of significant parts of the
current government tasks is very large in India. In practically
every department there are significant tasks29—being
currently done in-house and marked by little accountability,
vast inefficiencies, and waste—that ought to be privatized,
29 To name a few: printing and publishing, construction and
maintenance of buildings, tree planting and gardens, cash handling,
kitchens and canteens, design and architecture, data analysis,
computer services, etc.
46 India Infrastructure Report 2003

Box 2.1.7
S ome O bservv ations on the ‘R
bser
Obser epor
eportt of the E
‘Repor xper
xpertt Committee to R
Exper evie
Revie
evieww the SSystem
ystem of A dminister
Administer ed
dministered
Interest Rates and Other Related Issues’

Samir K. Barua
The Expert Committee for reviewing the system of administered interest rates for small savings schemes including the Public
Provident Fund (PPF) scheme, under the chairmanship of Y.V. Reddy, Former Deputy Governor, Reserve Bank of India, was
constituted in April 2001 by the Finance Ministry, Government of India. The Committee submitted its report in September 2001.
The Committee examined all the relevant issues connected with small savings schemes from the point of view of three
constituencies, namely, the small savers, the state governments, and the central government. To ensure acceptability of its
recommendations, the Committee has attempted to strike a balance between the needs of the three constituencies. As a result,
however, it has come out with a set of recommendations that lacks clarity and focus. We highlight some of the significant
deficiencies of the report and the recommendations.
The Committee stated that its objective was to ensure that the small savers ‘continue to get a safe avenue for investment at
reasonable rate of return’. The recommendations of the Committee, however, are implicitly coloured by a belief that the current
rates of return on small savings schemes are high, and that they are unduly high if the tax benefits available to savings under
these schemes are taken into account. It has therefore recommended withdrawal of tax benefits on short and medium maturity
(less than six years) instruments and fixation of rates on these schemes using benchmarks such as the rates on term deposits with
commercial banks and yields on government securities. While making these recommendations, the Committee does not make the
necessary distinction between ‘small savers’ and ‘small savings’. The distinction is important because of asymmetry of returns from
small savings schemes to small savers and large savers (those with high personal incomes).
Tax Concessions is Not the Way
Small savers are unlikely to be tax-payers. Therefore, they do not benefit from the tax concessions on small savings schemes. As
a result, while a 9 per cent tax-free interest rate remains 9 per cent for a small saver (non-tax paying), it effectively becomes 13.85
per cent for a person who is in the 35% marginal (highest) tax bracket. There is, therefore, little justification for the recommendation
that tax-free concessions should continue on long maturity (above six years) instruments, if the objective was to benefit small
savers! The benchmarking of rates on small savings schemes to other rates prevailing in the economy would in fact reduce the
interest rates available to small savers significantly. The decline would hurt the small savers. If the objective were indeed to benefit
‘small savers’ then the correct approach would have been to offer higher taxable interest on investments upto specified limits,
while withdrawing tax concessions completely. Such an approach would have completely eliminated the distortions introduced
by tax concessions as well as ensured that the extra return being paid on savings up to a ceiling truly benefited small savers.
The targeting of the schemes would have been much better. Going back to the earlier example, taxable interest rate of 13.85
per cent is far better for small savers than 9 per cent tax-free interest rate.
To ensure acceptance of the recommendations by the state governments, the Committee has recommended that the entire
collection through the small savings schemes should be transferred to the states. However, recognizing that ‘complete decentralization’
would result in rise in the risks associated with investments in these schemes that ‘could have an adverse effect on the overall
mobilization of such savings’, the Committee recommended that the current arrangement of mobilization should continue. This
essentially implies that the central government would continue to guarantee payment on these instruments (to the investors) in
case of default by the state governments. Given the financial indiscipline of the state governments, such guarantee by the central
government could land the central government into a difficult situation. The guarantee also raises the issue of moral hazard—
the states might be tempted to raise resources through this route with abandon and mis-spend the resources raised. Decentralization
should, therefore, be accompanied with shifting of responsibility on the states to raise resources through small savings schemes
without guarantee from the central government. In such a case, it is quite possible that different states would have to offer different
rates on the investments, depending on the risk perception of investors about the financial viability and credibility of plans of
the states. Competition among the states would benefit the small savers and improve the state of public finances in the country.
Financial Discipline of State Governments is a Must
Since 100 per cent transfer of funds collected through the small savings scheme implies considerable loss of capital receipts t o
the central government, the Committee took care to highlight the benefits to the central government from the proposal in the
form of higher tax collections and higher contributions to General Provident Fund (GPF). Recognizing that despite these
additional inflows, there is likely to be a decline in the total resource available to the central government, the Committee suggested
that the attendant shortfall could be made up through additional market borrowing. Such a substitution would in fact reduce
the cost of funds for the central government. This suggestion essentially implies that the aggregate deficit of the central
government would rise, since higher transfer of funds collected from small savings schemes is not to be offset through lower
transfer of funds from the central to the state government under other heads. It, therefore, appears that the recommendation
was directed at increasing the transfer of resources to the states. Recognizing that alignment (and hence reduction in the current
 Expenditure Accountability and the Society 47

scenario) of interest rates (to some benchmark rates in the economy) on small savings schemes could result in decline in the resource
mobilization through small savings schemes, the Committee suggested special market borrowing programmes for the states. This
would imply that the indebtedness of the states would rise. The recommendation, therefore brushes aside the need to bring about
greater financial discipline among states. The recommendation also implies that the states would be burdened with higher interest
costs (since the cost of funds from small savings scheme would be higher than market borrowing costs). As a result, the
recommendation potentially worsens the financial situation of the states. In the light of the above discussion on the fiscal situation
of both the states and the central government, the logic of the proposal of 100 per cent transfer to states is quite unclear.
The Committee also visualizes that in the foreseeable future, most small savers would join privately managed social security
schemes. It assumes that the change in management of savings would automatically bring about higher returns for the savers.
This may not be true for a variety of reasons. First, the government will have to permit far greater flexibility in the composition
of portfolios of the social security schemes that are to be managed by private agencies. Second, there is a paucity of investment
vehicles in the Indian economy for suitable diversification of investments for risk reduction. This is unlikely to disappear unless
the government permits investment of savings outside India (essentially permitting capital account convertibility). Third, going
by the performance of privately managed mutual funds, there is no reason to believe that private management of social security
schemes would necessarily result in better returns to subscribers.
Little Clarity
In summary, the report of the Committee appears to be a jumble of ideas that lack clear reasoning and recommendations that
conflict with each other and that sometimes go against the basic assumptions made for the analysis. In an attempt to please all
the constituencies, in particular, the state and the central governments, it has framed recommendations that are likely to prove
detrimental to fiscal management. The one concrete, unambiguous recommendation of aligning rates on small savings to certain
benchmark rates in the economy goes against the interests of the small savers, making a mockery thereby of the stated objective
of encouraging savings and providing adequate returns to the small savers.

recast as PFI contracts. Passing such sections/units around
from one ministry to another in the name of rationalization,
as the ERC has recommended is hardly the best way out.
The singular lack of exploration of alternative ways of
subsidization and government procurement, and carrying
out extension that can potentially be much more efficient
than the processes that exist today is a major failure in the
discussion. In so far as supply/procurement of infrastructure
services are concerned new designs for procurement are just
emerging in road construction and maintenance. But in
nearly all other areas like irrigation and drinking water,
sewerage services, subsidy administration where simple
changes in design could have plugged the vast leakages,
raised resources and increased the benefits to those intended
there is still no improvement.
With such a system in place, internal officers with deep
knowledge of their departments would be willing to
contribute to the task of identification and elimination of
the slack and waste, even as external consultants study
government departments to make their own assessments.
There is, of course, the third way which needs to be
pursued along with more expansionary monetary policy
and that is to enhance revenues, especially non-tax, and
reduction in the expenditure on subsidies. Our principal
focus is on this aspect of government budgets since the
scope here is enormous, and it directly bears on public
services provisioning, that includes infrastructure. The
NIPFP of 1993 concluded that the total subsidies, both
budgeted and implicit, are as high as 15 per cent of GDP!
Since explicit subsidies then were about 3 per cent, the
implicit subsidies were in the range of 12 per cent. Since
then there is little reason that this would have reduced in
any substantial manner. These estimates were on the basis
of recovery of operating costs/operational losses of
enterprises, providing services such as electricity and
irrigation. If the more reasonable assumption of recovery
at long run marginal cost/willingness to pay, whichever is
less, is made, then the extent of the implicit subsidies that
could be reduced is considerable and even a modicum of
change here can bring the fiscal deficit down in a functional
manner. If this is not attempted because it is difficult and
instead the investment or vital social services expenditures
are cut any further, then not only would they impose a
demand depressionary effect, but in the medium to long
term, by slowing down the growth in the quantum of
productive assets in the economy, could revive the
inflationary pressures if in future demand picks up due to
exogenous reasons.
48 India Infrastructure Report 2003
APPENDIX 2.1.1
Key Programmes and Schemes initiated by the Government of India during the Five Year Plans for the Rural,
Urban and Poverty Alleviation Areas (Programme; Plan Period)a
POVERTY ALLEVIATION
Accelerated Rural Water Supply Programme (ARWSP); Fifth/
1976
Integrated Rural Development Programme (IRDP); Sixth / 1980
Integrated Wastelands Development Programme (IWDP);
Seventh/1988
Employment Assurance Scheme (EAS); Eight/ 1993
National Social Assistance Programme (NSAP); Eight/ 1995
Indira Awaas Yojana (IAY); Eight/ 1996
Swarn Jayanti Gram Swarozgar Yojana (SJGSY); Ninth/ 1999
Jawahar Gram Samridhi Yojana (JGSY); Ninth/ 1999
Credit cum Subsidy Scheme (CSS); Ninth/ 1999
Samagra Awaas Yojana (SAY); Ninth/ 1999
Rural Sanitation Scheme; Ninth/ 1999
Pradhan Mantri’s Gramodaya Yojana (PMGY); Ninth/ 2000
Food for Work Programme (FWP); Ninth/ 2001
Economically Weaker Sections (EWS) Housing; First/ 1952
Urban Community Development (UCD); Third/ 1966
Environmental Improvement of Urban Slums (EIUS); Fourth/
1972
Sites and Services (S&S); Fifth/ 1974
Urban Basic Services for the Poor (UBSP); Eight/ 1992
Prime Minister’s Integrated Urban Poverty Eradication
Programme (PMIUPEP)
National Slum Development Programme (NSDP); Eight/ 1996
National Housing and Habitat Policy (NSHP); Ninth/ 1998
a Prepared by Mukesh P. Mathur
2.2 BUDGETARY PROCESSES: A POLITICAL ECONOMY PERSPECTIVE
The perspective taken in this Section 2.2 is that political
structure influences incentives, choices, including those of
politicians, and their outcomes in a country. Historical
shocks and past administrative choices are responsible for
the current structure. Budgetary processes and policy are
generally endogenous. There exist leverage points for change
(Morris, 2001), and change can be rapid once it reaches
a critical mass. Well-coordinated macro policy, including
The paper was presented at a workshop on the India Infrastructure
Report (IIR), 2003. Comments from and discussions with the
participants were very useful. The author thanks Arindam Das-
Gupta, and Sebastian Morris for generous inputs, and Niraj Goyal
for helping her understand the cost and responsibility accounting
successfully introduced in a multi-national company.
EMPLOYMENT RELATED
Self- Employment Programme for the Urban Poor (SEPUP)
Nehru Rojgar Yojana (NRY); Seventh / 1989
Swarna Jayanti Shahari Rojgar Yojana (SJSRY); Ninth/ 1997
PUBLIC DISTRIBUTION SYSTEM (PDS)
Antoyodaya Anna Yojana (AAY); Ninth/ 2000
Annapurna ; Ninth/ 2000
SELECTED SPECIAL PROGRAMMES
Integrated Development of Small and Medium Towns (IDSMT);
Fifth/ 1979
Low-Cost Sanitation (LCS); Sixth/ 1980Mega City Project
(MCP); Eight/ 1993
Accelerated Urban Water Supply Programme (AUWSP); Eight/
1993
Providing Solid Waste Management and Drainage in Ten Selected
Towns having Airfields of India Air Force; Ninth/ 1999
Sources:
Government of India, Annual Plan, 2001–2, Planning
Commission, New Delhi.
Government of India, Annual Report, 2001–2, Planning
Commission, New Delhi.
Government of India, Approach paper to the Tenth Five Year
Plan- 2002–7, Planning Commission, New Delhi, 2001.
Government of India, Ninth Five Year Plan,1997–2002,
Thematic Issues and Sectoral Programmes, Planning Commission,
Vol. II, New Delhi.
Ashima Goyal
infrastructure spending, with some restraints on political-
bureaucratic choices, could create incentives for rapid
growth. Growth eases political adjustments. It makes longer-
term sustainable redistribution feasible, and raises returns
to choices that improve human capital. The popular
awareness of the long-term consequences of past choices
that were focused on the short-term, and the bankruptcy
of governments, may make some changes incentive-
compatible for politicians. Others would require public
awareness and pressure.
Section 2.2 surveys theory, and the experiences of India
and other countries to bring out the systemic incentives
that affect political behaviour, government budgets, and
expenditure. The surveys are all partial. For example, only

the incentive features of choices made by fast-growing
countries, such as the US and China, are explored. It is
not possible to do justice to the enormous debate on India’s
development vis-à-vis the political economy. Although the
rapidly expanding debate on governance30 addresses some
of these issues, the latters’ role in India’s developmental
history has been somewhat neglected. It is necessary to try
to understand why policies that are known to be desirable
are not implemented, and what changes would make them
more feasible. The new feature is the emphasis on
disincentives built in due to responses to macro shocks,
given the structure and history.
There is a large body of recent theoretical and empirical
work in the new political economics literature on the
incentive effects of budgetary procedures and political
processes. Lessons have been extracted both from theory
and country experience relevant to Indian conditions.
SOURCES OF SURPLUS
Laffont (1999) remarks that economists can serve as
technicians and give politicians information which will help
them favour their constituency at the expense of other
groups. This leads to a worse average outcome. But they
have an alternative. ‘By suggesting constitutional rules which
decrease the discretion of politicians even at the cost of
some efficiency losses, economists can enhance expected
social welfare.’ This approach differs from the earlier Chicago
school political analysis, since the latter took the political
structure as given and uncontrollable. It focused on interest
groups and the deadweight losses of rent seeking activities
but ignored asymmetric information which gives the source
or supply side of influence activities. Transaction costs and
information imperfections are one reason for the inefficient
political equilibrium, and potential distributive benefits.
Simple rules that take these factors into account reduce
inefficiencies and yield extra output which is then available
to implement the rules. If political economy is a game of
redistribution of rents, the economics of incentives provides
the necessary tools for modelling interest-group rent capture
as a function of deeper economic parameters. Hoff and
Stiglitz (2001) argue that for deep changes to be possible,
it is necessary to understand political interests. The role
of the politician is then in between, neither like the predatory
Government of the Public Choice School, nor like the
optimal planner of neo-classical theory. The surplus out of
change can be used in ways that further enhance it, which
in turn makes cumulative improvement possible.
30 In particular the IIR, (3iNetwork, 2001 and 2002) focus on
regulation and governance in the context of infrastructure. See the
overview chapters by Morris (2001, 2002), and others by Anant and
Singh, Morris, Pandey, and Varma.
Expenditure Accountability and the Society 49
One definition of politics is ‘a mechanism for making
collective choices, for gaining power and using it in a
conflict of interest among economic actors.’ Political
economics recognizes the existence of political constraints
that cause actual policies to differ from optimal policies.
It leads to a normative research programme for discovering
which policy-making institutions produce better policy
outcomes. Since policies are endogenous, change can come
only by changing underlying institutions. Ideology and
even ideals have a role in political decisions, but once
perverse incentives affect a critical mass, it becomes difficult
for the few to resist the trend.
A planner’s optimal solution frequently differs from one
that is influenced by political forces. A planner’s solution
for aggregating individual preferences to maximize social
welfare is Pareto efficient so that no one individual can
be made better off without making another worse off. The
outcome is always on the Pareto frontier or set of such
Pareto efficient points. In political economics the focus is
on the political process by which the policy choice emerges.
Therefore the outcome may not maximize social welfare,
and may not even be Pareto efficient. If a mechanism can
be designed that moves society to the efficient point, a
surplus becomes available that may help to incentivize the
mechanism. The electoral process throws up political power,
and the political process in its quest for appeasing
heterogeneous interests, determines expenditure allocations.
Are changes that can lower waste and improve outcomes
possible in these deep-seated institutions and processes?
Public goods are non-excludable so there is a free rider
problem in their provision, although some infrastructure
can be priced. If consumers cannot be excluded from
consumption of a good there is no direct incentive for
them to pay for the good, leading to the collective action
problem. In the latter, the net benefit to a participant in
any group activity, after subtracting costs from gross benefits
arising from contributions to the group’s activity may be
negative. It is analogous to the Prisoner’s Dilemma Game
(PDG), in that, the rational outcome where everyone
maximizes their own welfare makes everyone worse off
since now public goods are under-provided (Hardin, 1982).
There is a natural role for the state in coordinating to the
better equilibrium31. Thus although politics is about power,
conflicts, and re-distribution, it is also about solving the
31 Basu (2000) argues for this view of the state and of law, as
an equilibrium selection device, that solves a coordination failure.
Since both are endogenous, they can change only the selection of
equilibrium, not the economy game. He points out that it was the
19th century British philosopher, Hume, who first understood the
role of the state in this way. While our collective action problem
gives rise to a PDG, there can also be a simpler coordination failure
with multiple Nash equilibria.
50 India Infrastructure Report 2003
collective action problem. It can do this by changing zero
or negative sum conflicts into interactions that yield a
potential surplus.
To sum up, externalities can arise from higher growth,
asymmetric information rents, and coordination problems.
These are all potential sources of surplus, which if realized,
can incentivize change. But the collective action failure that
gives rise to coordination problems influences political and
budget processes. Asymmetric information adds ineffi-
ciencies to the latter two. All these influence growth, which
is itself a source of surplus. Therefore deep and sustainable
change needs to act on these foundations.
Dominant international ideas have influenced Indian
policy-makers. Independent India inherited a strong state,
and wartime controls, from the British. The dominant
development idea was of state-led industrialization. But as
growth slowed down after about ten years, political pressures
resulted in direct short-term transfers that had indirect and
non-obvious long-term costs. The incentive properties of
these interventions only harmed growth. Current dominant
ideas that suggest a more balanced role of government give
an opportunity to set right some of these distortions.
BUDGETARY INSTITUTIONS
Theoretical models allow sharp derivations of the
consequences of structure and behaviour. If the aspects
focused on are dominant in real economies, such models
aid in understanding actual tendencies at work. In the next
two sub-sections we present such model-dependent results32,
where the models capture some relevant aspects of Indian
reality, related to budgetary institutions and political
processes. We briefly indicate in what respect the model
or result is relevant to India. In the sub-section ‘History’,
these results are used to explain aspects of a more detailed
Indian developmental history.
The major rules and institutions that determine
government spending are quantitative limits or budget
laws, and procedural rules. Finally, implementation methods
and tax-transfer policies also affect incentives.
Laws and Rules
Limits can be legislated on different government deficits
and debt or targets set, and ceilings put on expenditure.
But it is always possible to evade a budget law. Empirical
work, with data for the US states, has shown that laws were
complied with but only in letter. The laws had very little
32
The survey in this and the next sub-section is largely based
on Blanchard (2000), Drazen (2000), Goyal (1999, 2001), Laffont
(1999), and Persson and Tabellini (1999, 2000, 2002).
effect because creative accounting was often able to violate
their spirit33. For example, fiscal restrictions affected the
debt mix, but not the total. There was a shift to non-
guaranteed debt. Indian national and sub-national
governments are also able to hide deficits by juggling
extraordinary expenditure categories and so on. The problem
is compounded by difficulties in measuring government
debt or deficits. Internal as well as external incentives for
budget control are therefore required in addition to laws.
Examples of the former include norms, ethics, and accepted
practices. Better-designed budgets should make it easier
to plan and run programmes, and together with more
transparency, work to increase voter pressure to limit
budget sizes. It would remedy the frequent complaint of
Indian local governments about the imposition of rigid
budgetary practices from above, without a consideration
of local needs and capabilities. Flexibility would moderate
the effect of rigid budget laws in inducing a larger initial
size to take care of unforeseen contingencies.
Budgetary Procedures
One basic reason a democratic government may end up
overspending is the common pool problem. This arises
whenever spending decisions are made in a decentralized
way, in favour of different groups, and the tax rate is then
residually determined. The concentrated benefits are fully
internalized, but each group internalizes only a fraction of
the social marginal cost of the tax increase required for
funding, since the pool of tax revenue available for funding
is common. There exist specific procedures to help
overcome the problem, and limit the size of the deficit.
Some that have been examined analytically and tested
empirically are discussed.
Consider the following stylized budgetary process. Any
one representative, chosen to be the agenda setter, makes
a policy proposal. The legislature votes on the proposal,
if a simple majority approves, it is implemented, otherwise
the status quo remains. This is known as a closed rule since
proposals cannot be amended. It turns out that spending
will be lower if ministers without constituencies, such as
the Prime Minister or the finance minister, are given
special powers. These include agenda setting rights, right
of veto, and setting binding limits on allocations. It should
be clear who is responsible for an observed abuse of power.
33 See Drazen (2000), pp. 701. The Enron, Anderson and
WorldCom debacles have highlighted a similar problem in American
corporate governance. The letter of the law is often followed, but
many innovations including off balance sheet items violate its spirit.
Proposed reforms include stricter CEO accountability, control of
conflict of interest in the board of governors, transparency, and
market discipline.

This hierarchical process strengthens collective interests,
against the selfish interests of ministers who, under a more
democratic decision process, would push the interests of
their specific constituencies. The drawback is that minority
interests can be neglected. In addition, if the status quo
entails zero-base budgeting, this acts as a further disciplining
factor to control excess spending, since in that case the
default alternative involves minimal spending. The basic
reason separation of powers reduces expenditure is because
it removes conflict of interest; whoever cuts slices from
the cake should not be responsible for the distribution of
the slices. The agent who has proposal rights over expenditure
should not be the one who stands to directly gain from
it. The countervailing pressures generated induce, although
they may not ensure, a more fair and efficient division.
Under an open rule other legislators have amendment
rights, and there can be separation of proposal powers. For
example, there may be a vote on an amendment offered
on an initial proposal, with further rounds of amendments.
Although the possibility of amendments in the budgetary
process increases the size of the deficit, there is a theoretical
result that wasteful expenditure can be higher under a
closed rule compared to an open rule (Baron, 1991). The
reason is, if legislators are forward looking and sophisticated,
they internalize the stages of the budget process. Therefore
in order for them to support a budget proposal, when no
amendments are allowed, the net benefits they gain must
exceed those that could be expected if the proposal was
defeated and the budgetary process continued. To gain
majority support such a proposal needs to be large enough
to please everyone. There may be non-budgetary or office
related payoffs.
A similar problem as occurs in the closed rule can defeat
two-stage budgeting (2SB). Under 2SB, aggregate spending
is decided before allocations of expenditure in order to
insulate the total from interest group pressures. But if
sophisticated legislators realize the implications of their
decisions for subsequent appropriations, 2SB need not
lower budget size in equilibrium. A larger size would be
set at the first stage. The value of 2SB in lowering budgetary
waste has empirical support, however. Particularly when
there are agency problems the separation of powers in 2SB
helps voters to discipline their agents, the legislators. For
example, if the agenda setter is not the ‘residual claimant’
at the second stage, he will have no interest in expanding
the budget.
In Indian central government budget practices, primary
responsibility for the budget-size does lie with the finance
minister and the Prime Minister, although budgets tend
to be incremental so that this is not really exercised. And
forward-looking beneficiaries are able to get what they
want by, for example, padding salary bills (Das-Gupta,
Expenditure Accountability and the Society 51
1999). But still deficits were in control until the period
of unstable coalition governments and competitive populism
that set in from the 1980s. Although there are pre-budget
meetings with the different interest groups, these tend to
be routine. Widespread media debate after the budget
speech can lead to substantive changes34. A vote is taken
in the legislature before the budget bill is passed.
Implementation
Budgetary implementation through control and monitoring
make incentives low powered. The use of self-motivation
is minimal; agents are supposed to obey instructions. This
works better for simple processes or in the early stages of
development or planning when effective and low-cost
monitoring is feasible. In more complex processes monitoring
needs to be complemented by delegation to residual
claimants. The latter have high-powered incentives, since
the returns from greater efficiency stay with them. Such
changes characterize the movement from planning to more
market-led development. If politicians plan to stay in power
for long, campaign contributions are important, and there
is little competition so that reputation does not matter, they
will still prefer simple command-and-controls mechanisms
(Laffont, 1999). In India political fragmentation occurred
rather than competition. Reputations were uniformly bad
and a system was not devised for campaign funding.
Therefore politicians clung to discretionary power to be
able to implement short-term transfers and secure votes.
Following the control philosophy, budgetary
implementation in India, although quite detailed, tends to
focus on monitoring inputs. This makes participants strategic
and forward looking in trying to maximize their allocations,
but with no incentives for delivery. The horizon is very
short-term regarding outcomes. Modern management
information systems, in contrast, clearly allocate res-ponsibility
for outcomes, measured as variance from budgeted amounts,
right down the hierarchy, together with continuous feedback.
Mid-term correction is possible and delivery is ensured.
The system is transparent and incentives are very different.
Daily expenditure control and tax-transfer mechanisms
also affect budgetary implementation. One such mechanism
is user charges for public services. In India, as a populist
measure in the context of low growth and major cost
shocks of the early 1970s, many such user charges were
frozen. Apart from expanding budget deficits, this had
34 An example of this is the rollbacks announced in 2002. The
budget initially attempted to burden the middle classes. It was
perceived to be unfair and there was widespread protest in the media.
Finance Minister Sinha’s response was that he was not out to win
a popularity contest, but the Prime Minister intervened after the BJP
lost some by-elections, a new, more politically savvy, finance minister
was brought in who announced concessions for the middle class.
52 India Infrastructure Report 2003
major incentive effects on the provision of public services.
Exploring these will help to understand the disincentives
built into the Indian system.
Two extremes in pricing rules have opposite effects on
the incentives of a provider of a public good or service,
to lower cost or to improve quality. With a price cap,
incentives are high-powered since the residual profit share
lies with the producer. The latter is motivated to decrease
costs. Since rate of return regulation reimburses the cost
of the service, incentives are low-powered. Profits from
improvements do not stay with the producer. Thus there
is a trade-off between rent extraction and providing
incentives for additional effort. With a low enough price
cap the regulator extracts all rent, but incentives to invest
and improve quality may fall. Since costs rise with quality
low-powered incentives are required for the provision of
quality. There is a disincentive to invest in the presence
of low price caps also because sunk costs made for
investment may be expropriated. If even costs are not
covered rent is over-extracted35.
Optimal cross-subsidization is derived if social welfare
is maximized, subject to a budget constraint. This gives
the Ramsey–Boiteux rule that the price marginal cost gap
should be inversely proportional to elasticities of demand.
Thus the poor should be charged higher prices if they have
a lower elasticity of demand, but they can be cross-subsidized
as long as the budget is balanced. The rule has been
generalized into the theory of optimal non-linear tariffs.
These pricing principles are violated if, in the aggregate,
charges do not even cover costs. Cross-subsidization is
then excessive and generates inefficiencies that raise cost
further. Moreover, competitive entry occurs if, with cross-
subsidization, the price in any sub-market exceeds cost,
or if new technology lowers cost or breaks a natural
monopoly. Then government revenues from these sub-
sectors fall further. This happened in the Indian
telecommunication sector where expensive international
calls were used to subsidize local calls.
The underlying political structure influences the budget
processes adopted. This is discussed in the next sub-section.
POLITICAL PROCESSES
Since the focus is on systemic incentives, the role of
individual politicians, their leadership or the lack of it, is
abstracted from. It is possible for individuals to rise above
a perverse system, but the majority gets affected adversely.
Consider voters (the principals) who delegate control rights
and collective decision-making to elected officers (their
agents) because of costly and asymmetric information.
35
Many chapters in IIR 2001 cover these issues in detail. See
the overview in 3iNetwork (2001).
Since the preferences of voters and their elected
representatives are not perfectly aligned there is an agency
problem. Elected politicians know more than the voters do,
so that there will be information rents associated with the
provision of any public service. A stronger media and other
measures to improve transparency can lower such rents, but
not eliminate them altogether. Revelation mechanisms, such
as supervision or separation of powers, cannot fully extract
politicians’ information (rent) or prevent the politician from
favouring interest groups, unless there is perfect enforcement
by courts, perfect foresight, and rationality.
There are three degrees of freedom—the electoral
structure which selects the politician with residual control
rights, the discretionary decisions left with him, and the
non-contracted states of nature. There is a trade-off between
rents and efficiency. If the decisions of the politician are
restricted, his discretion and rent or political capital-taking
falls, but ex-post efficiency may also be destroyed. If
politicians chosen as residual decision- makers for ex-ante
non-contractible social decisions have more instruments
available, the outcome is more efficient and, the under-
utilized surplus from asymmetric information is less, but
they have more discretion to favour their own constituencies.
Separation of Powers
The Constitution uses separation of powers, checks and
balances, and elections to align incentives, but it is an
incomplete contract. Under non-Contractability about ex-
post states of nature there are cases where the constitution
provides no guidance and the politician has residual control
rights. Features such as a specific allocation of proposal
and veto rights, that create a conflict of interest among
politicians, can reduce the rent they capture.
One conflict of interest that the Constitution builds in
is that between the legislature which makes laws, and the
judiciary which interprets them. The judge-based system
imposed by the British government, was based on legislative
statutes meant to keep a subject-nation in control, unlike
their own common law tradition based on past judgements.
There was a strong contempt of court clause to give more
power to the judges. Even after independence the heavy use
of law continued in order to enforce development, to enable
confiscation of property, etc. So the legal system functioned
as an instrument of the government. Judges were also
expected to interpret the Constitution to maintain individual
liberty and social justice, and defend the weak against the
strong. Conflicts often arose with the government. In the
1980s public interest litigation (PIL) became a useful way
to put pressure on the executive to deliver, and to cover
lapses in legislating, as governance failed. Conflicts
sharpened; some landmark judgements were made. Since
legislators will not pass laws against themselves they resist

attempts to enforce greater probity and transparency on
them, leading some commentators to suggest an independent
law board to pass laws that affect legislators (Haldea, 2002).
But there are limits to judge-based activism. Too much
power to any one arm tends to corrupt it. It also goes
against the native Indian jury system of wise men who did
not directly confront the ruler, but reminded him indirectly
of the Dharmashastras and local norms. Judicial instincts
retain enough of the old value-based system to moderate
conflict (Dhawan in Frankel et al., 2000). Moreover, the
rule of law is expected to aid development of markets by
upholding certain basic principles, and ensuring
predictability in transactions across agents, over space, and
time (Trubeck, 1972). Excessive legislation, regulation, or
judgements, can impede these developments and restrict
individual liberties, leading to corruption and delays, which
make a mockery of the law. India has an archaic,
cumbersome system of laws whose basic underlying principle
of control is no longer relevant. There are huge backlogs
and delays in justice. If the law is ineffective, then more
of it, or imposing stricter punishment, may not work.
Instead, simplifying it, and bringing it closer to the way
people actually live, may work. Judges also do not have
the technical knowledge to pass judgements on areas where
expertise lies with the executive and should therefore just
enforce contracts (Anant and Singh, 2002); other means
of ensuring better governance are still required.
Hardin (1999) argues that a Constitution is not only
a contract but also a coordination device that can serve
to moderate group conflicts and push society to a better
outcome. Constitutions differ across countries. Hardin
suggests that the US Constitution, while protecting individual
liberty, property rights, and social justice, was able to
coordinate the society around commerce and prosperity.
Universal suffrage did not lead to expropriation as was
initially feared; opportunities expanded for the less well-
off also. Elster (1995) has argued that since it requires a
super majority to change the Constitution, it is a means
of credibly binding the government against future reneging,
much as Ulysses bound himself against the sirens. This
reassures against future imposts and expands social surplus
available. The Indian Constitution has many of these
desirable features, and can serve such a purpose, but the
electoral structure built in has had adverse effects.
Electoral Structure
Given the Constitution, elections impose ex-post discipline,
and, together with lobbying and legislative bargaining,
influence outcomes. One general principle is that forward-
looking private maximization can lead to extreme outcomes,
reversing institutional safeguards. Elections limit rent
extraction only if electoral promises are binding and if
Expenditure Accountability and the Society 53
alternative candidates are good substitutes for voters. But
often neither condition holds. In conditions where pleasing
the voters is very costly, or the politician has more discretion,
he may choose to grab as much rent as possible. Voters
can punish a candidate if he has not performed, but the
punishment is ex-post, not ex-ante. Reputation may not
be sufficient to enforce ex-ante discipline if alternatives
available are poor, as in India, where all politicians are
tarred with the same brush.
The structure of elections, written in a country’s
Constitution, also affects politicians’ incentives. A
representative democracy may have proportional or
majoritarian (‘first-past-the-post’) voting. India has the latter
system. In the pure form of the first, parties are given seats
proportional to their overall vote total. A minimum percentage
vote is required for a party to enter a legislature. In the
second, each district elects a single member by plurality
vote. Varying mixtures of the two systems are possible. The
way voter preferences reflect in electoral outcomes will
differ in the two systems, especially when there are more
than two parties. In general, under proportional elections
in a single national district, the incumbent needs the support
of a broader coalition of voters. So he provides more public
goods, which serve a larger population. Under single member
districts and plurality rule, it is necessary to win only in a
few pivotal districts. Therefore incumbents make transfers
to smaller geographic voter-groups. This tendency has
exacerbated India’s regional and caste-based electoral
polarization. A downside of proportional elections is since
voters vote on party lists rather than for single individuals,
corruption and equilibrium rent rise for politicians. Their
incentives to work for voters fall.
Both voters and parties can behave strategically so
conclusions have to be drawn with care. Even so, when there
is sharp polarization among voters and multiple parties, the
representatives of smaller parties in the legislature will normally
be below their strength in the electorate. But their power
in the government may be disproportionate, because if no
party has a majority they may be critical in a coalition.
Therefore legislative influence may not be monotonic in
vote share. For example, in a model with three parties it
can be shown that first and third ranked parties form the
governing coalition (Austen-Smith and Banks, 1988). Voters
will vote strategically, rather than sincerely. Sincere voting
is according to actual preferences, and strategic takes account
of other’s choices. Opportunistic candidates will allow for
this in taking up pre-election positions. Under proportional
voting the share of minorities in the legislature can increase,
but their importance in winning and losing falls. It is not
true that a coalition government will necessarily be unstable36,
36 While, in Italy, coalition governments have been highly
unstable, Germany, Luxemburg, and Austria have been governed by
coalition governments, over much of the post-war period.
54 India Infrastructure Report 2003
but if there is a pivotal party this will dictate policy preferences.
Thus those with a small block of votes may have large power
if they are required to help form the government. In India,
since the 1990s, unstable minority-dependent coalition
governments have become the rule. Shifting to some form
of proportional voting may increase the share of minorities
but decrease their power, and make for more stable
governments.
An interesting question is, will parties pamper swing
voters, whose loyalties are fluid, or will they target core
supporters? The latter could either be groups they can
reach effectively, or those who are ideologically fully
committed to the party. Opportunistic politicians will
favour core supporters if they believe that is the cheapest
way to buy votes, otherwise they will take them for granted
and court swing voters who may switch votes in exchange
for favours. Core supporters may even be taxed since the
party knows how to effectively extract resources from
them, without any change in their loyalty. Indian voters
have understood this and are beginning to insist on
performance, rather than being satisfied with ideology.
There are dramatic reversals in election results and
incumbents often lose elections.
If an electoral system favours many small parties,
governments tend to have short horizons and avoid tackling
hard choices. Grilli, Masciandaro and Tabellini (1991) find
that in Organization for Economic Cooperation and
Development (OECD) countries with such systems debts
and deficits rise. The greater the government
fractionalization, that is, the greater the number of policy-
making veto actors, the more difficult it is to devise plans
that distribute costs neutrally among their various
constituencies (Franzese, 2002). Therefore, reducing
fractionalization in Indian politics is a priority.
A presidential-congressional regime is an alternative to
a parliamentary regime. Equilibrium rents are lower in the
first as separation of powers and conflict among politicians
help voters control the agency problem. But the presidential
system supplies less public goods and more targeted
redistribution compared to the parliamentary system since,
under the latter, incentives to form broad stable coalitions
are higher. Thus, if public goods are under-provided, the
parliamentary system is better, but if rent extraction is high
or punishment from losing the next election is small because
entry barriers are high, the presidential system would work
better for voters. In India public goods are under-provided
and pressures to increase their supply must be maintained,
so that a parliamentary system should be better suited.
The democratic voice is valuable and must be preserved.
It allows robust institutions to develop that reconcile
different interests and make it possible to implement the
common interest. In a careful empirical study of the
outcomes associated with democracy, Tavares and Wacziarg
(2001) find that democracy reduces growth by reducing
physical capital and raising government consumption, but
stimulates growth by improving human capital and reducing
inequality. Redistribution is responsible for the second set
of factors, but excessive redistribution for the first set. Our
argument is that the second set can dominate if interest
groups are reconciled in a way that helps long-term growth,
for example, by developing human capital rather than
making short-term consumption transfers (Goyal, 2001).
In the sub-section, ‘History’, we find these conceptual
categories to be useful in interpreting the Indian experience.
HISTORY
If the various aspects of the structure of an economy are
captured by a theory, it is useful as an aid in bringing out
underlying causal factors, and explaining stylized historical
facts. Although one must always be aware that in a complex
system there are multiple interacting tendencies, the focus
here is on the consequences of political structure and
macroeconomic shocks. These partly explain the perverse
administrative choices and incentives that were built in.
India has a parliamentary democracy, which is expected
to lead to broader redistribution, larger government, and
more public goods. It has majoritarian, as opposed to
proportional, elections, which lead to less public goods,
more targeted transfers and larger government. They also
lead to less representation, but more power for minorities
under multiple parties and coalition governments. India’s
poverty, caste, religious, and regional heterogeneity added
to the effects of the electoral structure as multiple parties
appeared. Provision of public goods has been falling and
transfers targeted to well-defined electoral groups have
been rising after the 1970s. We argue that these ill effects
and pressures became apparent after growth faltered in the
mid-1960s.
Macroeconomic Events
The post-war control regime was continued in order to
target accelerated planned development with equity,
according to the development ideas of the time (Patel,
2001). This worked reasonably until the late 1960s when
there was a growth slowdown, partly due to a series of
shocks. There were, of course, inefficiencies due to the
import substitution regime, and the failure to carry out
land reforms and develop a broad demand base. The
country had not been able to make much of a dent on
poverty. This was also the period of the oil shocks and
sharp peaks in inflation, which hurt the poor. Realizing
 Expenditure Accountability and the Society 55

that the latter were a large vote block37, Indira Gandhi were derived from the Congress in various ways. The third
coined the slogan of ‘garibi hatao’38, and used it aggressively phase, starting with the state assembly elections of 1993–
in the early 1970s, in her political battle with the old 5, was of a competitive multi-party system. In the 1998
Congress Guard. Disregarding the rise in costs, several elections, the Election Commission registered 176 parties;
user charges were kept fixed, which led to more and more 41 won representation to the Lok Sabha, but of these 7
distortions. In the years that followed, subsidies and transfers had two or three members and 14 had one member each
increased while current and future provision of public (Frankel, 2000). The current National Democratic Alliance
goods suffered. As these effects cumulated, the revenue (NDA) coalition of 1998, led by the Bhartiya Janta Party
deficit became positive, or government consumption (BJP), is the first genuinely non-Congress government.
exceeded its income. Figure 2.2.1, shows that the first year As the Congress lost dominance and intense multi-party
the revenue deficit became positive was in 1980–1 and it competition set in, populist schemes multiplied. By the
grew every year after that. The government was now 1980s populist Central Sponsored Schemes (CSS) became
borrowing for consumption, and accumulating debt. These a way for the central government to directly reach the
policies, once implemented, set in vicious or virtuous self- masses. New schemes were announced every year, although
sustaining dynamics partly by creating interest groups or targeting was poor and waste and corruption proliferated.
constituencies they favoured39. The majoritarian democratic With multiple competing parties, swing votes became very
regime was emphasizing targeted transfers at the expense important for winning in our first-past-the-post system.
of public goods. Identity politics, based on caste and religion, was a way
of cultivating these swing voters. The Congress cultivated
40000 the Muslim vote, encouraging a conservative backlash40.
Yadav provides evidence that electoral participation by
30000 minorities and backward castes has risen since the 1970s,
compared to other groups. He calls this the ‘second
20000
democratic upsurge’ (Frankel, 2000). Sometimes minority
1000 groups had enough electoral power to swing outcomes in
their favour, for example, when parties representing them
0 were pivotal in a coalition. Defections and frequent
elections, as was the norm with unstable coalition
–10000 governments, meant that politicians were concerned only
1980–81
1972–3
1974–5
1976–7
1978–9

1982–3
1984–5
1986–7
1988–9
1990–1
1992–3
1994–5

with the votes that would get them to power. Once in

power they were concerned with ‘loot’ in order to buy votes
and legislators in the future. Since state elections were
Fig. 2.2.1 Revenue Deficit (Rs Crores) separated from those at the centre in 1971 (Jha, 2002),
frequent elections kept this pressure up continually and
Political Changes
harmed longer-term development.
Political fragmentation made matters worse. In the first The democratic empowerment of the poor, with the
twenty years after independence the Congress Party provided lower castes and regional parties acquiring voice and vote,
a stable government. There was no serious opposition. A was a positive and necessary development in the path to
period of bipolar consolidation followed, with the gradual a stable, mature, and inclusive democracy. But it further
development of an opposition system to the Congress encouraged identity politics. The first reaction of new parties
(Frankel, 2000). The Congress first lost the elections after to the acquisition of power was consumption transfers to
the Emergency of the mid-1970s. In 1979, 1989 and the their support groups, especially as belief in a vibrant future
mid-1990s different governments were formed, but all was missing since the development policies of the past had
not delivered growth. Institutions of governance were
37 For example, 60 percent of Mumbai’s population lives in
undermined. Pai (2002) documents how, in Uttar Pradesh
slums, providing a concentrated target for politicians. This is not to (UP), identity and caste politics were responsible for poor
rule out that some politicians may be sincerely concerned to better
the lot of the people. But they still did not use the most effective 40 Vote bank politics created bitterness and set the communities
means of doing so. against each other. In Gujarat, Narendra Modi’s inaction and the
38 She first used the term while intervening in a debate in the NDA’s support to his government despite accumulating evidence of
Lok Sabha on the Constitution (24th Amendments Bill, 4 August the state government’s failure and even complicity in the anti-Muslim
1971, see GOI, 1975). riots of March–April 2002, are said to be motivated by the BJP’s
39 Stiglitz and Hoff (2001) develop the idea of past choices recent election losses, and the assembly elections due in Gujarat.
creating interest groups under the concept of dynamic political economy. Playing the Hindu card became an electoral strategy for the BJP.
56 India Infrastructure Report 2003
governance and low development expenditure. Each short-
term coalition government reversed jobs given to other
caste groups and put its own group in positions of power.
Frequent transfers of officials politicized and demotivated
the bureaucracy. It is true, as Pandian (2002) points out,
that arguments about the common good and efficiency can
be used to keep the oppressed down. A creative ferment
is necessary for them to come up, but the policies followed
do not have to be regressive if a longer-term perspective
is taken. In the states of south India where the movement
is older, it has achieved progressive reform, emphasizing
education, and going beyond symbolic gestures such as
building memorials to Ambedkar.
Administrative Choices
Administrative choices led by the populist stance, made
matters worse. After the oil shocks, and in the presence
of an inflation sensitive electorate, many prices of public
services were kept fixed; incentives and efficiencies were
sacrificed for consumption re-distribution. The choices
made amounted to protecting the poor through current
transfers, rather than building their assets and human
capital, when it was the latter that was the sustainable
option. This was a rational social outcome because the rich
could often escape imposts in the long-term, and the poor
had high discount rates and pessimistic growth projections.
The objective of providing government services at
affordable prices led to cross-subsidization both in the
provision of specific products and across government
functions. Low price-caps for many public goods led to
systematic incentives to lower quality and investment. Thus
falling efficiency and rising costs compounded the problem
of low user charges, and prevented a natural fall in prices
from improvements in technology and organization. But
where the government had monopoly power and was servicing
the rich, prices were raised much above costs of production,
or indirect charges, such as the prices of intermediate
goods, which were not obvious to voters, were raised (Morris,
2001). As the rich turned to private providers, revenue
losses contributed to the inability to service the poor
adequately. The cross-subsidization was not sufficient to
cover costs. General revenues did not even cover government
consumption expenditures, as the widening revenue deficit
shows. Long-term human and physical capital formation
(including infrastructure) were harmed. Again we see public
good provision suffering in order to raise targeted transfers.
Power, telecommunications, railways, roads, education,
all illustrate the effects of these incentives. Quality and
availability of services fell. Poor quality served as a cost
that lowered consumption demand. For example, freight
rates were raised to subsidize passenger travel. The railways
steadily lost customers to diesel trucks. Since diesel was
also subsidized this further reduced government revenues,
apart from raising pollution levels. The average rate per
passenger kilometre was raised from 4 paise in 1980–1
to 15 paise in 1991–2, but the rise in the average rate per
tonne kilometre was more over the same period (from 10.5
to 35.1 paise), although the demand for freight is more
elastic. International call charges were kept high, to subsidize
domestic calls; again the aim was to tax the rich and
subsidize the poor. In 1998, the cost of calling the USA
using Internet telephony was Rs 4 per minute but the
public sector monopoly charged Rs 75 per minute.
Companies formed in the USA made it possible to call
from India and pay abroad. Thus price distortions invited
the entry of companies and the government could not
retain its domestic monopoly and lost revenue. As new
technology and companies enter and cut international call
rates, the public sector giants are being forced to follow
suit41. This is the history against which reform is taking
place in the provision of public sector services.
The special powers of the finance minister and the
Prime Minister were not able to protect the Central Budget
once the era of competitive populism set in. In addition,
perverse incentives adopted in the accounting procedures,
led to problems on both the revenue and expenditure side.
Budgeting is only incremental, with a mechanical increase
in Plan expenditure42. The division between Plan and non-
Plan expenditures, together with the focus on starting new
schemes, has harmed maintenance. After the shocks of the
1970s, as fund constraints appeared, it was easier to
postpone investment plans. Moreover, decision-making
and monitoring procedures became inadequate as
complexity increased and contributed to delays.
The centre has constitutional powers to restrict borrowing
by states, and states do not have the ability to monetize
their debt43. This works to limit their deficits but a
number of avenues are available to soften their budget
constraint. Disbursements from the Finance Commission,
Central Schemes and the Planning Commission are multi-
41 See Goyal (1999) for more details and examples. IIR (2001)
documents many such examples. The point made (see Morris, 2001)
that cross-subsidization often raised the prices of intermediate goods,
again provides evidence of the politician’s tendency to give short-
term or direct benefits at the expense of rising long-term or indirect
costs.
42 These points are well accepted in the Indian debate. Amaresh
Bagchi and Govinda Rao reiterated them at a meeting in IGIDR
and at a Mumbai University seminar respectively in 2001. See also
Rao (2002).
43 Armijo (2002) argues that since inflation hurt the poor, and
there was universal franchise, macro policies were conservative in
India. One aspect of this was limits on the states’ ability to
undertake deficit financing. In Brazil, where universal franchise was
granted only in 1986, credit limits on states were lower and inflation
much higher.

purpose, poorly designed, and, therefore, fungible. The
presence of multi-lateral lending institutions may worsen
the problem. States can float special purpose vehicles to
borrow funds, and do other smart accounting to soften
budget constraints (see Rao, 2002). Good performance has
not been rewarded, so even though more equality in
revenue sharing was built in across the states, equality in
per capita incomes was not achieved. The Eleventh Finance
Commission put in some performance incentives, and
improvements in the states’ revenue collection have been
reported. On the other side, delays in specific funding
further delay projects and expand their costs. Poor
management practices and bankruptcy is making these
problems more acute. Low recovery of user charges and
the absence of a service tax are constraining the states’
revenue. The direction of reform towards simpli-fication,
rationalization, and harmonization is well known, but is
not being speedily implemented, even after being initiated
in 1993. There are similar problems at the third tier, the
Urban Local Bodies (ULBs) and the Panchayats. The centre
has not performed its coordinating function well. States
were able to hold it to ransom especially as regional parties
became critical. Adversarial positions taken in negotiations
on issues that affected a number of parties contributed to
delays (Basu, 1995). The common pool problem was
aggravated and free riding became pervasive. The centre
used its power over states only to impose distorting policies
that fragmented markets.
The choices made led to the conflict between rich and
poor, labour and capital, the present and the future, being
handled in a way that lowered incentives for expansion of
the cake. The plan emphasis was on long-term capital goods
but implementation led to a short-term focus. Redistribution
concentrated on short-term consumption subsidies rather
than on assets, education, health, and infrastructure. The
poor were needed for votes, the rich for buying votes. Apart
from short-term consumption redistribution to woo voters,
the rich were given opportunities for rent seeking in a
system of controls and distortions44. Labour laws secured
jobs for a small labour elite, but restricted the expansion
of such jobs. Indians became used to direct subsidies and
indirect costs, or short-term benefits and disguised long-
term costs. Although these measures did not deliver long-
term benefits voters valued them because in a low-growth
44 Bardhan has developed the idea of warring interest groups in
a number of articles. Bardhan (2002) sums up his position as that
there were multiple interest groups with the power of veto, but no
one was strong enough to take over the government. Therefore the
collective action problem could not be solved, and the country settled
for short-run transfers at the cost of long-term growth. The position
in this Section 2.2 differs in that democracy should be able to achieve
common interests, but macro shocks, the electoral structure, and
administrative choices created perverse incentives in the Indian case.
Expenditure Accountability and the Society 57
environment indirect benefits are often discounted because
of their uncertainty and greater chance of going to others.
Moreover, in the absence of a viable two-party system voters
had little choice. If status quo is lower the agenda setter
can extract more of the surplus from voters. In the more
recent period, however, a strong anti-incumbency vote has
developed, suggesting that delivery has begun to matter.
Citizens are seeing the long-term consequences of short-
term populism. Political parties are careful to listen to the
people. There are signs that they have realized this. Governance
is the new slogan; for example, the Congress has moved
from ‘garibi hatao’ to governance at its 2002 meeting.
Even after reforms and the move towards more market
oriented high-powered incentives, government procedures
often continue to be based on the philosophy of monitoring
and control. Sometimes rules and procedures appropriate
in one stage persist in another where they become harmful.
Many improvements are possible, but deep and sustainable
changes require addressing the underlying causes. Politicians
have to be motivated to push the changes. It may be useful
here to see how better-performing countries have ensured
a stimulating macro-environment and used high-powered
incentives.
OTHER COUNTRIES
Two countries that have grown rapidly in the past decade
and used macro- policy and incentives to do so are China
and the US45. Non-transparency moderated international
pressures and allowed China to use massive fiscal-monetary
stimuli according to domestic requirements. One-party
dictatorship made it possible to reverse direction quickly
when inflation rose too high. Thus stop-go cycles led to
some waste but facilitated required structural changes, and
induced high savings. Moreover, China successfully used
economic incentives, even without full property rights. The
US succeeded when it shifted from an overly strict budget
law to a more realistic one that allowed counter-cyclical
fiscal stimulus and monetary accommodation, in an interest
rate-targeting regime.
The United States of America
In the 1990s, the American budget deficit, which had
mushroomed in the 1980s, was converted into a surplus
while the economy prospered. Among the factors that
made this achievement possible, were rules that limited
wasteful government expenditure, and allowed monetary
stimulus to which the economy responded. Supply-side tax
45 The discussion of the US experience follows that in Goyal
(2002), and that of China follows Goyal and Jha (2002). See also
Morris (2002).
58 India Infrastructure Report 2003
cuts in the Reagan era were expected to stimulate output.
But they could not deliver, and since there was no
compensating cut in government expenditure, the deficit
grew rapidly. It fluctuated around a peak ratio to Gross
Domestic Product (GDP) of 6.1 per cent in 1983. The
Gramm–Rudman–Hollings Act passed in 1985, set yearly
ceilings for the deficit with a target of a zero deficit by
1991. It had some effect, but creative accounting—because
of design flaws (including overstrictness)—prevented the
realization of the target. The Bill also lost credibility because
a very large deficit was required to fund the 1990 crisis
in savings and loan institutions. Therefore, it was replaced
by the Budget Enforcement Act of 1990, which corrected
the design flaws shown up by experience. First, rather than
restrict the deficit itself, constraints were placed only on
spending. Caps that enforced small reductions in
discretionary spending were set, but escape clauses were
provided for emergencies. A ‘pay-as-you-go’ rule meant that
new transfer payments to individuals could be made only
if these transfers were demonstrated to have assured funding,
so that they would not increase deficits in the future. As
revenues fell in a recession, the deficit could increase, since
restraints only covered spending. This macro-stabilization
provided another escape clause. Such flexibility lowered
pressure to break rules, gave the Act more credibility, and
contributed to its success. There was a systematic attention
to incentives. For example, there was a shift from matching
to block Federal grants for state welfare payments; the
former had motivated the states to overspend. After twenty-
nine years America had a balanced budget.
China
China is a one-party dictatorship, but before its high growth
phase, it shared with India the pressures of a large number
of poor. Both had a rigid bureaucracy and controls that were
meant to ensure redistribution and to reduce inequalities.
Even without private property rights and functioning markets,
China was quicker than India to decentralize, create intense
local competition, and incentivize economic decision-making.
In India checks on economic decisions created delays and
discouraged productive activity. China focused more clearly
on high growth, and was thus able to overcome other
obstacles and allow the regime to adapt.
Partly because of the socialist ideas of equality, without
populist pressures of ensuring votes, protecting the poor
largely took the form of human capital development—
improvements in health, education, and endowments.
During the period when food remained a major part of
the average household budget, adequate supplies and low
prices of essential foods were maintained. As long as these
minimum requirements were met, the government was
able to extract obedience.
After 1978, economic decision-making was decentralized
to farmers, firms, and lower-level governments and incentives
were given. The people by now had the skills to respond
rapidly. The government policy of ‘crossing the river while
feeling the stones’, prevented too large and disruptive changes.
Mistakes were made, but were as quickly reversed.
Decentralization and a conscious policy of taking intellectuals
abroad, preparing thoroughly, and examining alternate future
scenarios, overcame some of the information inadequacies
of a dictatorship. More drastic reforms such as privatization,
closing down bankrupt banks, restructuring the bureaucracy,
and introducing the Voluntary Retirement Scheme (VRS)
were only undertaken in the late 1990s after twenty years
of rapid growth. China had always controlled its labour
markets tightly. As the boom in exports absorbed labour and
migration restrictions were relaxed, urban decay was
minimized through policies like the one that new firms had
to provide housing for workers. The idea was to make those
who stood to benefit most from it, to contribute to
infrastructure development.
The official debt GDP ratio remained at around 10 per
cent, but official data on government debt are suspect.
Lower-level borrowing is not well captured. There are
many contingent and hidden liabilities in the banking and
pension systems. Recent reports suggest that, properly
measured, deficit ratios are as high as Indian ones. But
the lack of transparency and capital controls deflected
international pressure for macro-stabilization, and allowed
them to implement large macro-stimuli and create a good
infrastructure for growth. In the 1998 slowdown, for
example, the fiscal stimulus from large-scale public sector
infrastructure projects was about 2.5 per cent of GDP.
Rules to restrain credit expansion were enacted later. After
a new Budget Law in 1995, the central government could
not borrow from the central bank on current account,
although deficit financing was permitted on capital account.
The deficit had to be financed with government bonds.
More stringent budget balance restrictions were imposed
on local governments. The law strictly controlled their
bond issuance and restricted their borrowing in the financial
market. An independent auditing system was also
introduced for enforcing the budget law.
Till 1993, 70 per cent of the central bank’s loan to state
banks were made by the central bank’s local branches,
which were heavily influenced by the local governments.
In 1993, the central bank centralized its operation. In
1995, China passed the Central Bank Law to give the
central bank the mandate for the monetary policy
independent of the local governments. These reforms
reduced the influence of the local governments on the
monetary policy and credit allocation decisions. In 1998,
the central bank replaced its 30 provincial branches with

9 cross-province regional branches. This reform further
minimized the influence of the local governments on
monetary policy, and closed inflationary loopholes.
These changes were part of the new feature of the 1990s
worldwide—the adoption of fiscal and monetary rules to
improve incentives. The rules have worked best when they
have been used to facilitate macro-policy coordination and
stimulus. The appropriate structure varies with levels of
development. A democracy cannot afford Chinese-style
stop-go cycles, or make the quick reversals. Therefore, in
the former, self-enforcing general rules, which allow sufficient
discretion for flexibility, should work better.
POLICY AND STRUCTURAL CHANGE
We have demonstrated the importance of macroeconomic
shocks and the disincentives that can be built in by
inadequate responses to them. Therefore discretion needs
to be constrained by rules that would ensure adequate
responses to future shocks, without long-term damages.
Even in a complex system it is possible to identify marginal
feasible changes that can trigger others.
• A Fiscal Responsibility Act would restrain fiscal
populism. It should have escape clauses to allow counter-
cyclical stimulation of the economy, and protect investment
in human and physical infrastructure. Monetary policy
then can be supportive and stimulate growth. Such a set
of policies is self-enforcing, given democracy and existing
economic structure. Monetary policy can be bound less
strictly than fiscal policy because democratic pressures
work to force it to keep inflation low46. Higher growth
will facilitate the removal of political hurdles, and changing
incentives in administration.
Electoral Reform
Deep structural changes would also minimize perverse
responses to shocks, and have better incentive features.
Electoral reform may be possible in the longer-term. Even
a few of the structural reforms suggested below have the
potential to reduce populist pressures on Indian politicians.
The long history of backwardness, in some sections of
society, makes measures to empower these sections
necessary. Bhargava points out that many such measures
are provided in the Constitution itself (Frankel, 2000). The
reforms suggested below would actually increase the
representation of minorities, but reduce their power over
political outcomes, so that more long-term measures to
benefit them could be taken.
46 This policy coordination is discussed in detail in Goyal
(2002). Morris (2001, 2002) also argues for interest rate targeting
and a high growth target.
Expenditure Accountability and the Society 59
• A presidential system would probably be too
authoritarian to balance India’s diversity. Moreover, since
the supply of public goods needs to expand, a parliamentary
system is better. But shifting to a system with some features
of the proportional voting, with some lower limits on the
minimum size of parties, would raise the supply of public
goods, reduce transfers, decrease swing votes, vote blocks,
fractionalization, and instability in politics.
• Hung assemblies that give importance to identity-
based splinter groups can also be reduced if a variant of
the French system is followed whereby, in the absence of
a clear majority, there is a re-poll among the two parties
with the highest totals. This would lower instability in
governments, frequent elections, and their ill effects. Some
change is required because of extreme fragmentation in
Indian politics. If state and central elections are coupled
once more, governments will not have to continually
implement populist schemes to appease voters (Jha, 2002).
Unfortunately, the recently concluded Venkataswami
Commission on constitutional review has not suggested
any serious reform.
• The anti-defection law has effectively stopped single
defections, but since it does not rule out splits or group
defections, we have the spectacle of groups of MLAs
being hidden away to keep them from temptation in
toppling episodes as was observed in Maharashtra in
2002. Splits were allowed because it was thought that they
might be based on principle rather than on the lure of
office or lucre. But that does not seem to be the case,
leading to suggestions that the anti-defection law be
extended to splinter groups (Parikh 2002). This would
make coalition governments less hostage to vested
interests, while democratic freedoms of choice could still
be exercised before elections.
Political Changes
Some political changes are feasible even without changes
in the electoral structure.
• There are signs that the political system is moving
towards single party-led broad coalitions, one towards the
right and the other towards the left of the political spectrum.
This could lower electoral competition, but this would
require the large parties to become more inclusive and
sensitive to regional aspirations. If two big parties, one on
the moderate right and the other on the moderate left,
accommodate local satraps, and each sits regularly in
opposition, there could be more genuine consultation
across parties and more stability (Chidambaram, 2002).
• Even without such an outcome, it may be possible
to achieve consensus across parties on major economic
policies. All parties more or less agree to the necessity of
reform today; bankrupt governments have no other choice.
60 India Infrastructure Report 2003
The broad coalition governments of the 1990s have
hammered out large areas of consensus (see Box 2.2.1).
If an all-party consensus can be reached banning populist
policies, such as free electricity, no government would fear
an electoral reversal from implementing the policy, and it
would become politically feasible. It is the fear of the other
party using populism to gain electoral mileage that is
preventing deeper reform today. Such self-binding would
improve governance and basics. Voters have begun not to
reward governments that give populist sops but do not
deliver the basics. They have voted out incumbents on non-
performance. Both rich and poor are converging on
governance. As controls dissolve, the ability of the politicians
to extract rent from the rich is falling, and the poor are
now in a position where they see the decay in vital economic
services that affect them in the long-term. Political parties
have realized this, and therefore may be willing to agree
to a consensus that would allow them to deliver on
governance. Then short-term transfers would fall and allow
a rise in public goods provision.
Using Democracy
Democracy is not just about voting. There are other ways
in which democratic pressures can enforce governance.
• The media and the people, bolstered by special
institutions, can function as watchdogs in areas where
politicians use power for personal gains or do not deliver
(Paul and Gopakumar, 2002) New communication
technologies and norms favouring transparency make this
easier. India has a long tradition of open discussion and
respect for different views. But the peoples’ right to knowledge
needs to be better implemented. It is time the Official
Secrets Act, inherited from British times, was modified.
Citizens’ groups and non-government organizations (NGOs)
have become active in enforcing civic rights and improving
urban infrastructure (Raval, 2002). If they publicize the
failures of politicians and indirect costs of populist policies,
it will increase the accountability of politicians and motivate
them in the right directions. For example, in cross-
subsidization, the government keeps ‘poor’ class consumer
prices low, raises ‘luxury’ goods prices, but increases
intermediate prices (Morris, 2002a). This eventually raises
a wider range of prices, besides bringing about severe
distortions. Such effects should be made transparent.
• More feedback is a strength of democracy, and
today’s technologies allow faster feedback, so that the speed
of decision-making can be raised. Inequality and low
growth heighten conflicts that make a democracy slow
moving. Therefore conflict-resolution strategies that focus
on the common good, and not on specific groups will help.
Mechanisms that resolve conflicts in ways that increase the
surplus or growth are required. For example, giving free
school meals or dry rations to girl students, illustrate how
policy can improve returns to choices that raise human
capital. The latter carries externalities for society that
benefit all. Another example is if mandated contributions
to a worker’s welfare fund across industries47 accompany
laws that make it easier to fire workers. This illustrates
another general way of solving the collective action problem:
gainers should partially compensate short-term losers. If
such laws increase employment both industry and workers
would gain in the longer-term. Democracy is one of the
best ways to integrate heterogeneous groups. Moreover, its
balance and variety allow it to survive shocks better. As
India’s polity and economy matures, it is showing resilience
in the face of shocks that would have precipitated a crisis
earlier48. Some parts benefit from adverse shocks and can
have positive feedback for others.
• It is said that the Indian democracy has too many
checks that paralyse decision-making. But because of the
past emphasis on monitoring and the use of law to enforce
development, these checks and balances apply more on
citizens. If some of these are designed for politicians and
bureaucrats to create conflict of interest among them, it
would restrain the legislature and the executive from using
power to benefit themselves, and would align officials
interests with the common good. Separation of power
concentrates information rents in a single office. If the
other office is not a residual claimant on rent, its interests
would be aligned with voters, and it would share information,
resulting in more transparent policy. If one office has the
power of initiative and a veto but not an amendment right
(and therefore is not a residual claimant), and the other
has allocation rights, the two have a strong conflict of
interest and no incentive to collude, which helps the voter
control their rents. If combined with other design features
to speed decision- making, this need not increase delays.
Administrative Changes
Administrative changes, both at the design and procedural
level, can improve efficiency. Changed circumstances
sometimes make it possible to change set ways of doing
things.
• Independent professionals should be appointed as
regulators, not retired bureaucrats. The present generalist
bureaucracy could be modified, giving it a core with more
47 Kavita Khanna, as a representative of the confederation of
Indian Industry (CII), made the suggestion during an episode on
labour laws in the programme, The Big Fight, Star News.
48 Bimal Jalan, Governor of the Reserve Bank of India (RBI),
made the point that although shocks had regularly led to crises in
the past, India faced a number of adverse shocks in the late 1990s
without a crisis. Speech at the release function for the India
Development Report, May 2002.
 Expenditure Accountability and the Society 61

Box 2.2.1
Is Consensus Possible on Policy?

The United Front and the National Democratic Alliance (NDA) were both broad coalition governments, one to the left and
the other to the right of the Indian political spectrum. The first came up with the Common Minimum Programme (United
Front, 1996), and the second with a Joint Manifesto. Studying the two statements gives an indication of the extent to which
a consensus across parties is possible in economic policies. Each involves a consensus among the parties comprising the coalition;
the two together give an idea of feasible consensus across the Indian right and the left.
Being political statements, both emphasize food security, health, water, housing, education, and governance. The United Front
puts more emphasis on price stabilization, provision of basic needs, and technology, while the NDA places emphazis on agriculture
and infrastructure for employment, swadeshi with openness, privatization, commercial principles in general but protecting farmer
subsidies.
Thus both are concerned with higher growth, and there is a general agreement on the reforms and other measures required
to achieve it. The left is more concerned with affecting transfers to the poor and preserving public property. But it is willing
to accept that more of these transfers should come in the form of building human capabilities, and to make allowances for the
bankruptcy of government.
There are sensitive electoral issues like subsidies and user charges that neither document touches upon. These are precisely
the areas in which all parties would gain, if there were an all-party agreement that none could gain political mileage by reneging.
They would be able to deliver better governance, and compete on that basis.
The NDA document for the first time talks of the necessity of consensus across the political spectrum. ‘Governance must
become unifying, not divisive, in its practice. It is this mindless manner of the domination of the majority that has led to bitterness,
hostility, and confrontation—and has even made the opposition and ruling parties like permanent adversaries. This destructive
trend of politics of negativism and political untouchability which has distorted our body politic in the last few decades calls for
an immediate corrective. We will, therefore, strive to develop national consensus on all major issues confronting the nation by
involving the opposition parties and all section of society in dialogue. We will also try for a consensual mode of governance as
far as practicable.’ Words need not imply actions but they are a first step.
There are other initiatives to achieve this. The CII has set up a public policy council to facilitate policy dialogue and consensus
across political parties. Sanjiv Goenka is the current head (Soota, 2002). Chidambaram (2002) is also of the view that such
consensus is feasible. Prime Minister Vajpayee went to brief Sonia Gandhi, the leader of the Opposition, after the Jammu terrorist
attack in May 2002, in order to involve all political parties in India’s response, and in his Independence Day speech again
emphasized the necessity to build consensus on economic policies.

independence from politicians (limiting their rights to
transfer officers) plus a contract system where an incoming
government appoints its own officers, but based on strict
professional criteria and skills.
• Like China India should give civil servants some
benefits from growth, so that they are motivated to encourage
it instead of using their discretion to create delays to make
money. The rewards need not be only monetary but could
be in terms of professional development or reputation,
thus creating a strong work ethic. Since bureaucrats have
multiple tasks monetary incentives alone are insufficient
and can take attention away from essential tasks whose
output cannot be measured. Even so, better management
systems can ensure accountability and transparency, with
emphasis on delivery (Morris, 2002b). Better pay and
equipment must accompany incentive payments linked to
performance. Monitoring on outcomes should be designed
to encourage risk taking in bureaucracy. For example,
bankers should be judged on their entire loan portfolio,
not on individual loans, and the emphasis should be on
the quality of the loans, not quantity. They should be given
an official incentive payment on loans repaid, instead of
the illegal cut in loans sanctioned, that is common at
present.
• Well-sequenced, contextual policies can be designed
like those in China where construction firms were asked
to provide housing for workers, thus making direct
beneficiaries contribute to infrastructure development, or
regional bank branches were closed down to contain the
inflation of credit. New transfer schemes must be asked
to demonstrate the source of future funding. The systematic
removal of discretionary controls will reduce sources of
corruption and delay. As complexity and asymmetry of
information increases, outcome oriented, market-based
delegation works better.
• India has a three-tier budgeting process—at the
central, state and local levels. ULBs49 give an example of
problems at lower levels. Services are generally poor. ULBs
start with a balanced budget, since the law requires it, but
49 The workshop on the IIR 2003, gave the author an
understanding of ULB financing. In particular she would like to
thank K.P. Krishnan and Ravikant Joshi for detailed discussions on
this topic and refers the reader to their chapters in this Report for
more details.
62 India Infrastructure Report 2003
end the year with deficits. These are adjusted in creative
ways, contributing to the general opacity of the budget.
Data are either not current or are difficult to interpret.
Schemes are delayed for lack of funds, while sometimes
money is available but authorities do not know of it. The
required reforms are known but are not implemented.
Opacity continues in spite of many training sessions,
capacity building, and attempts at standardization. Perhaps
it facilitates the siphoning off of money, so political bosses
do not want transparency. Politicians must be motivated
to actively push for these changes or at least stop resisting
them. This can happen if deeper changes in the political
structure make the provision of public goods to a broad
set of the population rather than to narrowly targeted
transfers in the politicians’ interest.
• Alternatively, conflicts of interest could be used or
created to decrease political power in administration, or
voter pressure could be used to monitor and ensure
performance. It may be possible to introduce management
accounting if the utility of the capacity created is demonstrated
at the local level. Transparency and standardization improve
implementation. Information technology gives fast useful
feedback, makes decentralization and faster decision-making
possible. Coordination of central transfers and their quick
disbursement according to simple non-discretionary rules
will also be motivating. On the expenditure side, performance
clauses must emphasize maintenance. Programme budgeting,
with a clear idea of cost-effective expenditures targeted to
programmes demarcated on functional grounds, such as
administrative, or social headings, rather than the present
classification into development and non-development
expenditure, would be much more effective. Simple forward-
looking rules can prevent bottlenecks, even in complex
situations (Basu, 1995).
• Increasing the rewards to self-policing works better
than a strict law, which is kept in letter but violated in
spirit. This often happens with the PILs where citizens’
groups bring court injunctions against the executive to
enforce service delivery. Direct voter pressure on politicians
and changes in the incentives of politicians and bureaucrats
will deliver more genuine changes. Tough penalties often
require tougher evidence, so that conviction becomes
more difficult. Rewards motivate relatively more, compared
to punishment. Under asymmetric information a carrot
relaxes both the participation and incentive compatibility
constraint, since it improves returns to the better alternative.
A stick just relaxes the incentive compatibility constraint
since it reduces returns to the worse alternative. And
growth makes more carrots available.
• Awards to states have been based on filling trend
non-plan revenue gaps. This rewards fiscally profligate
states, but has not succeeded in closing gaps in per capita
incomes (Rao, 2002). Transfers from the rich to the poor
states can continue by having higher block grants for the
latter; but an incentive component linked to fiscal prudence
and efficiency may motivate them to take on more
responsibility. This, together with reforms such as allowing
them to levy service tax, making user charges more feasible,
tighter budget constraints and better accounting,
competition among states, openness to the world, may help
spread better practices among the states. Tax mobilization
in the states has been better than that at the centre in 2002,
because of the competition and the incentive features that
the Eleventh Finance Commission built in.
• Public investment in human and physical infrastructure
continues to be essential given the severe shortages in this
area and the large externalities it carries. The public sector
also has the advantage of being able to borrow more cheaply
to finance such activities. But some commercial principles
are required in allocation and pricing. Otherwise the
asymmetric information and rent-seeking activities that
have been examined so far imply that politicians may follow
self-serving policies such as building huge highways to their
small villages. Innovative public–private joint initiatives
and independent agencies for finance, allocation, and
regulation, need to be established in this area. But attention
should be paid to the dynamic sequencing of reform. For
example, more competition should accompany privatization,
to prevent the creation of private monopolies (Hoff and
Stiglitz, 2001), and interest groups that impede further
reform. Political forces are actually working to make such
changes more likely. More and more bankrupt public bodies
are welcoming private participation as a means of protecting
them from voter and worker pressure. It is a way of tying
their hands; they can say it is the private partner who is
insisting on payment, performance, and raising prices.
Market procedures can short-cut delays due to adversarial
negotiations. For example, auctions can be used as a way
to determine values available to settle claims50. Although
these changes would reduce their discretion, it is possible
politicians will be forced to adopt them.
CONCLUSION
In India politicians get too much rent and too many resources
are dissipated in short-term transfers to build vote banks.
In this Section 2.2 some systematic changes have been
discussed that can release resources for activities that generate
externalities and stimulate growth. Politics has two basic
functions. It can solve the collective action problem and
coordinate for a better outcome, but it can also use conflict
among groups to maintain power. India has seen too much
of the latter function. Conflicts make a democracy slow
50 Raghuram Rajan pointed out, in a lecture at IGIDR, that this
is used in Sweden to determine the value of bankrupt firms.

moving under inequality and low growth, therefore changes
that improve outcomes for all, and not only for specific
groups, are required. Higher growth reduces group conflict.
The general principle is that policies should improve returns
to choices that raise human capital. The latter benefits more
than just the recipient and improves future prospects.
When the underlying political structure has major effects,
the question arises, will politicians allow the changes
required? If they are aware of the consequences they will
only allow changes that are in their own interest. If policy
is endogenous so are institutions. Ideas do have an effect
but take time to acquire a critical mass. Deep political
changes may be some time in coming, but the bankruptcy
REFERENCES
Anant, T.C.A. and J. Singh (2002), ‘The Framework of
Regulation’, in S. Morris, (ed.), India Infrastructure Report:
Governance Issues for Commercialization, New Delhi:
3iNetwork and Oxford University Press.
Armijo, L.E. (2002), ‘Democratic Inclusion and Macroeconomic
Moderation: Counterintuitive Observations on India and
Brazil’, paper prepared for the workshop ‘Creating and
Revitalizing Democratic Institutions: Context and Challenge
in Latin America and South Asia,’ Institute of International
Relations, University of British Columbia.
Austen-Smith, D. and J. Banks (1988), ‘Elections, Coalitions,
and Legislative Outcomes,’ American Political Science Review,
82, 405–22.
Bardhan, P. (2002), ‘The Political Economy of Reform in India’,
in R. Mohan (ed.), Facets of the Indian Economy, New
Delhi: Oxford University Press.
Baron, D. (1991), ‘Majoritarian Incentives, Pork Barrel Programs,
and Procedural Control,’ American Journal of Political Science,
pp. 35, 57–90.
Basu, A. (1995), Public Expenditure Decision Making: the Indian
experience, New Delhi: Sage.
Basu, K. (2000), Prelude to Political Economy, New York: Oxford
University Press.
Chidambaram, P. (2002), ‘On Economic Reforms and Political
Compulsions’, at the Maharashtra Economic Development
Council meet on 10 April, Taj Rooftop,
Das Gupta, Arindam (1999), ‘Expenditure Management and
Budgeting in India’, background paper for chapter 4,
‘Governance, the Business of Government’, in World Bank,
India: Reducing Poverty and Accelerating Development, New
Delhi: Oxford University Press, 2000.
Das, K. (1999a), ‘Politics of Industrial Location: Indian Federalism
and Development Decisions’, Economic and Political Weekly,
Vol.32, No.51, pp. 3268–74.
Das, K. (1999b)‚ ‘Underdevelopment by Design? National Planning
and Vital Infrastructure in Orissa’, Working Paper No. 107,
Gujarat Institute of Development Research, Ahmedabad.
Dobb, Maurice (1948), Soviet Economic Development Since 1917,
New York: International Publishers.
Expenditure Accountability and the Society 63
of governments is forcing reforms, reforms are changing
incentive structures, and higher growth rates are reducing
the conflicts and short-term perspectives. Indians had
become used to direct subsidies and indirect costs, or
short-term benefits and disguised long-term costs. But they
are beginning to see the long-term consequences of short-
term populism. Political parties have understood this
perception; for example, the Congress’ slogan, has changed
from ‘garibi hatao’ to governance.
Changes will occur, but with conscious design they can
be surer and faster. It is often said that a few good-
principled people can make a difference; but good systems
can induce better behaviour from most people.
Drazen, A. (2000), Political Economy in Macroeconomics, New
Jersey: Princeton University Press.
Elster, Jon (1995), ‘The Impact of Constitutions on Economic
Performance’, Proceedings of the World Bank Annual
Conference on Development Economics, 1994, Washington
D.C.: World Bank.
Engerman, Stanley L. and Robert E. Gallman (eds) (1986),
Long-Term Factors in American Economic Growth’, Studies in
Income and Wealth Series, Vol. 51, Chicago and London,
University of Chicago Press.
Evans, Peter (1984), Dependent Development: The Triple Alliance
of MNC, State and Local Capital in Brazil, Princeton, New
Jersey: Princeton University Press.
Frank, Andre Gunder (1967), Capitalism and Underdevelopment
in Latin America. Historical Studies of Chile and Brazil,
New York and London: Monthly Review Press.
Frankel, F.R. (2000), ‘Introduction/Contextual Democracy:
intersections of society, culture and politics in India’, in
F.R.Frankel, Z. Hasan, R. Bhargava, and B. Arora
(eds),Transforming India: social and political dynamics of
democracy, New Delhi: Oxford University Press.
Franzese, R.J. (2002), Macroeconomic Policies of Developed
Democracies, U.K.: Cambridge University Press.
Fry, Maxwell J. (1980) ‘Saving, Investment, Growth, and the
Cost of Financial Repression’, World Development Report,
8(4), 317–27.
GOI (1979–80), ‘Performance Budget of Ministry of Energy
(Dept. of Power)’, Government of India.
GOI (1987–88a), ‘Performance Budget of Ministry of Energy
(Dept. of Power)’, Government of India.
GOI (1987–88b), ‘Performance Budget of Ministry of Water
Resources’, Government of India.
Government of India (1975), ‘The Call of Change’, in The Years
of Endeavour: the Selected Speeches of Indira Gandhi, August
1969–August 1972, New Delhi: Director, Publication
division, Ministry of Information and Broadcasting.
Goyal, A. (2002), ‘Coordinating Monetary and Fiscal Policies:
a role for rules?’, chapter 11, in Kirit S. Parikh and R.
Radhakrishna (eds.), India Development Report, 2002, New
Delhi: IGIDR and Oxford University Press.
64 India Infrastructure Report 2003
——— (2001),’Governing a Democracy: Constrained Discretion
for Development’, presented at the GDN 2001 Conference,
Rio de Janeiro, December 10–13, 2001. Available at http:/
/ctool.gdnet.org/conf_docs/CP_(Ashima_Goyal).pdf.
——— (1999), ‘The Political Economy of the Revenue Deficit’,
in Kirit S. Parikh (ed.), India Development Report, 1999,
New Delhi: IGIDR and Oxford University Press.
Goyal, A. and A.K. Jha (2002), ‘Dictatorship, Democracy and
Decentralisation: Macro Policy in China and India’, presented
at seminar on Foreign Investment and Economic
Development in India and China’, IIM Calcutta.
Grilli, V., D. Masciandaro, and G. Tabellini (1991), ‘Political and
Monetary Institutions and Public Financial Policies in the
Industrial Countries’, Economic Policy, 13 October 341–92.
Haldea, G. 2002, ‘Rescue Reforms: Players Cannot Also Be
Umpires’, Times of India, 24 July.
Hardin, R. (1999), Liberalism, Constitutionalism, and Democracy,
New York: Oxford University Press.
——— (1982), Collective Action, Baltimore: Resources for the
Future and John Hopkins Press.
Hoff, K., and J. E. Stiglitz (2001), ‘Modern Economic Theory
and Development’, in G. M. Meier and J. E. Stiglitz (eds.),
Frontiers of Development Economics, New York: Oxford
University Press and World Bank.
IIR (2002), India Infrastructure Report: Governance Issues for
Commercialization, (eds.) Morris, S. and Rajiv Shekhar,
3iNetwork, New Delhi, Oxford University Press.
Jha, P.S. (2002), ‘Centre cannot hold’, Hindustan Times, 29
March, Delhi.
Joshi and Little (1996), India’s Economic Reforms: 1991–2000’,
Oxford University Press, New Delhi.
Krugman, Paul (1996a), Pop Internationalism, Cambridge and
London, MIT Press.
Krugman, Paul (1996b), ‘The Myth of the East Asian Miracle’,
in Krugman, Paul (1996)
Laffont, J.J. (1999), Incentives and Political Economy, 1997
Clarendon Lectures, Oxford: Oxford University Press.
Lewis, Arthur W. (1954), ‘Economic Development with
Unlimited Supplies of Labour’, Manchester School, Vol.22,
May, pp. 139–91.
Mankiw, George (2000),’The Inexorable and Mysterious Trade
off Between Inflation and Unemployment’, Series: Harvard
Institute of Economic Research Working Papers, Harvard
Institute of Economic Research, No.1905. (http://post.
economics.harvard.edu/hier/2000papers/HIER1905.pdf)
Morris, S. (2002a), ‘Overview’, in 3iNetwork, India Infrastructure
Report: Governance Issues for Commercialization, New Delhi:
3iNetwork and Oxford University Press.
———2002b, ‘The Challenge to Governance in India’, in
3iNetwork, India Infrastructure Report: Governance Issues for
Commercialization, New Delhi: 3iNetwork and Oxford
University Press.
——— (2001), ‘Overview’, in 3iNetwork, India Infrastructure
Report: Issues in Regulation and Market Structure, New Delhi:
3iNetwork and Oxford University Press.
Morris, Sebastian (1987), ‘The Process of Investment Decisions:
A Study of Delays and Cost Overruns in Public Sector
Projects’, mimeo, Institute of Public Enterprise, Hyderabad.
Morris, Sebastian (1990), ‘Cost and Time Overruns in Public
Sector Projects’, Economic and Political Weekly, Vol.25, No.47,
pp.154–68.
Morris, Sebastian (1994), ‘Some Issues in the Debate in Policy’,
Economic and Political Weekly, July 2, 1994, pp.1669–73.
Morris, Sebastian (1997), ‘Why not Push for a 9 Per Cent
Growth Rate?’, Economic and Political Weekly, May 17–24,
pp. 1153–65.
Morris, Sebastian; Rakesh Basant, Keshab Das, K. Ramachandran
and Abraham Koshy, (2001), The Growth and Transformation
of Small Firms in India, New Delhi, Oxford University Press.
Nurkse, R. (1953), Problems of Capital Formation in
Underdeveloped Economie’, Oxford: Blackwell; New York:
Oxford University Press.
Pai, S. (2002), ‘Electoral Identity Politics in Uttar Pradesh:
Hung Assembly Again’, Economic and Political Weekly, 6
April, pp. 1334–41.
Pandian, M.S.S. (2002), ‘One Step Outside Modernity: caste,
identity politics and public sphere’, Economic and Political
Weekly, 4 May Vol. 37, No. 18, pp. 1735–41.
Parikh, K.P. (2002), ‘Overview: Ten Years of Reforms, What
Next?’, in K.S. Parikh and R. Radhakrishna (eds.) India
Development Report, 2002, New Delhi: IGIDR and Oxford
University Press.
Patel, I.G. (2002), Glimpses of Indian Economic Policy: An Insider’s
View, New Delhi: Oxford University Press.
Paul, S. and K. Gopakumar (2002), ‘How Public Feedback Aids
Public Accountability’, in S., Morris (ed.), India Infrastructure
Report: Governance Issues for Commercialization, New Delhi:
3iNetwork and Oxford University Press.
Persson, T. and G. Tabellini (2002), ‘Political Economics and
Public Finance’, in A.J. Auerbach and M. Feldstein (eds),
Handbook of Public Economics, Vol. 3, chapter 24, North-
Holland: Elsevier Science.
——— (2000), Political Economics: Explaining Economic Policy,
Massachusetts: MIT Press.
——— (1999), ‘Political Economics and Macroeconomic Policy’,
in J.B. Taylor and M. Woodford (eds), Handbook of
Macroeconomics, Vol. 1, chapter 22, North-Holland: Elsevier
Science.
Rao, M.G., (2002), ‘State Finances in India: Issues and Challenges’,
Economic and Political Weekly, Vol. 37, No. 31, 3 August.
Raval, S. (2002), ‘Urban Heroes’, India Today, 13 May.
RBI( 2001), ‘Report of the Standing Committee on International
Financial Standards and Codes’, Chairman : Y.V. Reddy,
Reserve Bank of India, Mumbai.
Rosenstein-Rodan, P.N. (1943), ‘Problems of Industrialisation
of Eastern and South Eastern Europe’, Economic Journal,
Vol. 53, June–Sept., pp. 202–11.
Soota, A. (2002), ‘CII’s New Mindset’, The Times of India, 6 May.
Stein, Jerome L. (1991), International Financial Markets:
Integration, Efficiency, and Expectations, Cambridge: Blackwell.
Tavares, J. and R. Wacziarg (2001), ‘How Democracy Affects
Growth’, European Economic Review, 45(8), pp. 1341–78.
Trubek, D.M. (1972), ‘Towards a Social Theory of Law: An
Essay on the Study of Law and Development’, The Yale
Law Journal, Vol. 82, No. 1, pp. 1–50.
United Front (1996), Common Minimum Programme, text reported
in the Economic Times, Mumbai edition, 6 June, pp. 22.
Wallis, John and Douglass North (1986), ‘Measuring the
Transaction Sector in the American Economy’, in Engerman,
L. Stanley and Robert E. Gallman (eds) (1986).