Shorter Paper The Budget Deficit in Transition A Cautionary Note

This paper discusses major fiscal issues faced by the previously centrally planned economies in their transition to market economies. It focuses on the extent to which the budget deficit should guide policy and the interactions between fiscal policy and other aspects of the macroeconomy.


This paper discusses major fiscal issues faced by the previously centrally planned economies in their transition to market economies. It focuses on the extent to which the budget deficit should guide policy and the interactions between fiscal policy and other aspects of the macroeconomy.

During central planning, it was not really meaningful to speak of fiscal policy and public finance since these concepts imply the existence of private finance and thus a significant private sector.1 In that era, all was public in a way. However, many “fiscal” functions, as defined in market economies, were carried out not by government but by state enterprises. These enterprises often provided to their workers housing, hospital care, vocational training, kindergarten facilities, shops, pensions, various forms of welfare assistance, employment, and so forth. They were also responsible for much “public” investment. Because the state enterprises were required to hire workers even when they did not need them, official unemployment was nonexistent. As a consequence, no programs existed to protect unemployed workers. In economies in transition, state enterprises are still carrying out many of these fiscal functions and are still hoarding workers. For the transformation to market economies to be completed, most legitimate social functions must be shifted from the state enterprises to the government. As this shift takes place, spending by the enterprises will fall and spending by the government will rise.2

In spite of occasional divergent views, most economists agree that, in market economies, the fiscal deficit is a useful guide for assessing fiscal policy. A large deficit indicates that (unless a strong case to the contrary can be made) fiscal policy should become more restrictive. The concept of fiscal deficit implies that government activities can be sharply delineated from those of the private sector.3 Thus, expenditures and revenues are either public or private. The measure of the fiscal deficit requires that the difference between what the public sector spends and what it collects can be accurately calculated. With full employment, if the public sector spends more, the private sector must spend less, unless the economy is able to attract foreign saving. Thus, the existence of a fiscal deficit implies some (unwanted) crowding out of the private sector unless Ricardian equivalence or a Keynesian multiplier can be assumed to operate.

I. Budget Deficit Limits and Perverse Incentives

There are several reasons for caution when using this concept to guide policy action during the transition. Some of these reasons are conceptual. Others deal with measurement problems. Still others originate from the possibility that strict limits on an inadequate measure of the fiscal deficit may create perverse incentives, which may slow down the transformation.

Transferring Social Expenditure Responsibility

State enterprises in economies in transition continue to dominate economic activity and provide social services to workers and their families. Some of these services are provided in place of higher money wages. Others are provided in place of higher budgetary expenditure. In other words, these enterprises continue to perform some functions that in market economies are financed through the government budget.

The transition to a market economy requires that the budget assume the responsibility for those social functions that the country wishes to finance publicly.4 This transfer to the budget should occur regardless of whether the state enterprises are privatized or not. The faster it occurs, the easier it will be for the enterprises to be privatized or to become economically viable while remaining state owned.

The transfer of these functions to the budget will increase budgetary expenditure and, unless revenue is raised correspondingly, will also increase the budget deficit.5 Thus, the transfer should also reduce the credit needs of the state enterprises. For most of these countries, however, there is little solid knowledge about the potential impact on the budget of a total or partial transfer of these functions.6 In Russia, “the government appears not to have quantified the dimensions of this problem or planned a solution” (Wallich (1992, p. 7)). The more important are these functions and the faster and the larger is their transfer to the budget, the greater will be the growth of budgetary expenditure. Thus, ex ante limits on the size of the budget deficit should explicitly allow for the budgetary cost of transferring some of these functions from the state enterprises to the budget. 7 The possibility must be recognized that a limit on the budget deficit that ignores this transfer might be met by the government delaying the transfer of these social functions from the enterprises to the budget. This delay would slow down the process of transformation but would not reduce credit expansion if the banking system continues to react to the needs of the enterprises.

If a decision were made to subsidize some loss-making state enterprises, it would be better economic policy if they were subsidized through the budget rather than through soft and cheap loans from the banking system.8 However, subsidies financed through the budget increase the budget deficit, and cheap loans do not.

It would be unrealistic to assume that subsidies to enterprises, whether through the budget or through cheap and soft loans, could be completely and permanently abolished in the short run. Some of these enterprises (especially those in the defense industry) will be able to restructure and become economically viable only if they receive financial support for some time. It would also be unrealistic to assume that those who make or influence economic policy would look at cheap credit given directly through the banking system in the same way as they look at subsidies financed through the budget.9 Thus, the argument that has occasionally been made—that the provision of subsidies through the banking system or the budget is only an accounting issue and would disappear if one measured the fiscal deficit in a comprehensive and economically sound way 10—ignores the important point that economic policy is not always guided by the most economically correct concepts and that, in any case, the measurement of the all-embracing fiscal deficit is next to impossible under the circumstances prevailing in these countries. Thus, the measured (budget) deficit is likely to continue exerting more influence on policymakers and public opinion, both within and outside the countries, than the true but unknown (fiscal) deficit. For this reason, policymakers have an incentive to show a smaller budget deficit even when this action may conflict with other important economic and social objectives, such as the speed of transition to a market economy.

If the government raised taxes on state enterprises’ profits at the same time that the banking system extended cheap or soft loans to these enterprises, the budget deficit would fall without reducing monetary expansion. This implies that focusing on the budget deficit may lead observers and policymakers to miss the underlying cause of monetary expansion and could lead to the apparently anomalous situation (experienced by Poland and, especially, Yugoslavia) of high inflation with the budget in surplus.

Budget Deficit and Inflation

The relationship between inflation, the budget deficit, and the public debt is also relevant to this discussion. To the extent that there is a domestic public debt, an increase in the nominal interest rate paid on the public debt will be reflected in government expenditure and in the size of the budget deficit even if the rate of the increase matches exactly the increase in inflation—in other words, even if the real interest rate does not change. Given a significant public debt and a potentially high but unknown future rate of inflation, it is difficult to predict the effect on the deficit if the government pays a given real interest rate on its debt. Given a real rate of interest, the rise in the deficit will be a direct function of the size of the debt and the level of the nominal interest rate, which, in turn, will depend on the inflation rate. The potential effect of a change in the rate of inflation on the conventionally measured budget deficit could be very large.11

In the situation described above, a government that has an interest in showing a smaller budget deficit might be tempted to repress nominal interest rates, thus creating difficulties for the development of the financial sector and for the allocation of financial resources. Under inflationary conditions, concepts such as the operational or the primary deficit, which attempt to eliminate the impact of inflation on the deficit, may have to be used. But they will only correct the budget deficit rather than measure the true fiscal deficit. Furthermore, these concepts have their own limitations.

Budget Deficit and the Unemployment Rate

If state enterprises lay off redundant workers in order to improve their efficiency and, as a consequence, the government’s spending for unemployment compensation increases, the change will cause the budget deficit to rise. The rise will depend on the level of unemployment compensation per worker and on the number of unemployed workers. For a given level of unemployment compensation per worker, the larger the adjustment in the work force by the state enterprises, and thus the greater the number of unemployed workers, the larger will be the potential impact of this adjustment on the budget deficit. Should the government be excessively concerned about the size of its budget deficit, it may encourage the state enterprises to continue hoarding workers.

The potential impact of this factor on the budget deficits of countries in transition can be appreciated by the fact that falls in output that have, at times, been as large as, or larger than, 30 percent of GDP have generally not resulted in corresponding increases in unemployment.12 If these countries had behaved like market economies, they would have experienced unemployment rates not seen since the Great Depression of the 1930s.13 As long as the state enterprises continue to hoard workers, they will not be able to restructure and become efficient and competitive. It will also be more difficult to privatize them.14 However, when the enterprises fully adjust their employment to their greatly reduced production needs, and the governments take responsibility for financing the cost of unemployment, budgetary expenditures will go up. This increase must be taken into account in setting ex ante limits on the deficit, but it may be very difficult to forecast.

Budget Deficit and the Exchange Rate

If a country devalues its official exchange rate, its interest payments on foreign debt measured in domestic currency will rise, thus increasing public spending, and government expenditure on imports measured in domestic currency will also rise. Both of these factors would tend to raise the budget deficit.15 On the other hand, the devaluation will also increase the domestic value of exports by the public sector and the revenue from some taxes. However, if the proceeds from the public sector exports are outside the budget, even if the exporters are public institutions, the improvement will not be reflected directly in the budgetary accounts unless the exporters contribute significantly to tax revenue.16 In other words, the budget deficit will rise, and the rise will depend on the size of the devaluation.17 Rigid ex ante limits on the measured size of the fiscal deficit (that is, limits on the budget deficit) might encourage the authorities to delay the needed exchange rate adjustment especially when the external debt is large and the government’s imports are significant.

Shifting Expenditures out of the Budget

Another perverse incentive that may arise from the attempt to show a smaller budget deficit or to stay below a too rigid limit is that of pushing expenditures out of the budget and into other parts of the public sector, whether extrabudgetary accounts or lower levels of government. Indeed, extrabudgetary accounts have proliferated in many economies in transition, though information on them has been elusive. Also, sizable expenditures have been shifted onto local government. In this context, it may be worthwhile to cite from a study on fiscal decentralization in Russia by Christine Wallich (1992, p. 3):

In the fiscal program for 1992, most of the major cuts were made in central government expenditures—centrally financed enterprise investment, producer and consumer subsidies, and defense. These cuts were followed by a decision to delegate an important part of social expenditures (early in 1992), and investment outlays (later on) to the subnational level. On the tax side, the budget envisages a marked increase in taxes, primarily on petroleum products and foreign trade. Thus, virtually all additional revenue will accrue to the federal government, while most of the additional social expenditures will emerge at the subnational level (italics in original).

Wallich’s conclusion is also worth citing:

The basic strategy has been to “push the deficit downward” by shifting unfunded expenditure responsibilities down in the hope that the subnational level will do the cost cutting (pp. 3-4).

These are just a few examples of the possibilities that call for caution in the use of the budget deficit in guiding economic policy during the transition. They do not exhaust the possibilities.18 They all point to the need to ensure that perverse incentives are not created by reliance on narrow concepts of the deficit. These incentives will exist when the deficit covers only a part of the public sector and when the limits imposed on it are too rigid. In this case, there will be an incentive to “park” the deficit where it cannot be measured or to postpone structural adjustments that may tend to increase the size of the measured deficit.19

II. Need to Separate Fiscal from Monetary Policy

A total and sharp separation between fiscal and monetary policy is next to impossible in any economy. There are always overlapping areas. For example, when central banks change the discount rate or engage in open market operations, these actions have an impact on the budget deficit. In general, these two policies are less clearly separated in developing countries than in industrial countries. In Latin America, for example, the central banks have, in several cases, engaged in activities that have caused them to experience large quasi-fiscal deficits.20 In Chile, for example, a quasi-fiscal deficit arose during the 1980s as a result of the central bank assuming the bad debts of the commercial banks. This assumption was clearly a fiscal operation that could have been carried out by the budget if budgetary conditions at the time had been more sound. In other cases, the central banks assumed the foreign debt of enterprises.

In general, the less developed the institutions of a country, the more difficult it is to separate monetary and fiscal policies and the more difficult it is to measure the true size of the fiscal deficit. More often than not it is the weakness of the fiscal institutions (for example, the inability to raise needed revenue) that forces the monetary institutions to assume a fiscal role. However, making the budget deficit look better than it is also plays a role.

Normally, economic policy is improved when monetary and fiscal policies are pursued according to their own specific roles. In this case, if an enterprise needs to be subsidized, the subsidy should be given through the budget. If this results in a budget deficit, the deficit should be financed by the country’s treasury through the sale of bonds carrying market interest rates.21

The budget deficit is a highly watched economic barometer. A larger deficit always sends warning signals, both domestically and internationally. Therefore, countries have an incentive to limit its size either by genuine fiscal adjustment or by pursuing second-best policies, the result of which may be even more damaging than if the budget deficit had been allowed to rise. In other words, the bias is almost always for showing a budget deficit that appears smaller than it would be if the true fiscal deficit were correctly measured. At times, this reduction is achieved through policies that do not reduce the real or underlying fiscal deficit but only the measured budget deficit. In other words, gimmicks are used to lower the budget deficit artificially.22

During the transition, to the extent possible, it is important to contain the inflation rate without resorting to direct price controls.23 To achieve this objective, the most important macroeconomic instrument for both market economies and economies in transition is credit expansion, not the budget deficit. Once total credit expansion is determined, the countries should be encouraged to continue improving the efficiency of the economy as well as the fiscal accounts. They should be encouraged to reduce particular categories of public expenditure, to make others more productive, and to raise government revenue in an efficient way to contain the public sector fiscal deficit. They should also be encouraged to speed up the various adjustments mentioned above, even when these adjustments may result in increases in (measured) budget deficits. Important among these adjustments is the transfer to the government of the essential social functions shouldered by the state enterprises.

Another important point to recognize is that which makes the economies in transition different from normal market economies: for given total credit, an increase of the (measured) budget deficit does not necessarily crowd out what, in market economies, is the private sector. Much of the credit expansion goes to the state enterprises and the government, and only a small residual goes to the emerging private sector. A large reallocation of credit between the budget and the state enterprises can take place without necessarily affecting the private sector. This reallocation could be fully consistent with the objectives of the transformation. If state enterprises (a) shed redundant workers, (b) transfer to the government the responsibility for essential social functions, (c) no longer finance nonessential social functions (such as vacation for workers), and (d) compensate the workers for a fraction of these reductions through some adjustment in the level of nominal wages, then the net result of these changes could very well be a larger budget deficit with a more efficient economy. This change could also be achieved without additional credit expansion if the additional credit to the budget to finance its new responsibilities were compensated by lower credit to the state enterprises to reflect their reduced social responsibilities.

III. Concluding Remarks

The paper has shown that in situations where the role of the government has not yet been determined, where the budget must assume some responsibilities now carried by state enterprises, and where “property rights” within the public sector are ill-defined, the budget deficit, which is calculated by looking at the behavior of budgetary revenue and expenditure, has less informational value than generally assumed. This deficit may often widely differ from the true fiscal deficit for the whole public sector and may even move in a different direction. The true fiscal deficit would include the fiscal activities of the state enterprises and the central bank.

Under particular circumstances, containing the budget deficit will not necessarily make more resources available to the private sector and may even slow down the structural reforms necessary to make the transition successful. The above conclusion should not. however, be understood to mean that fiscal policy is irrelevant. If there is a message that follows from the discussion, it is (a) that the most comprehensive and economically sound measure of the fiscal (not budget) deficit should be used to guide policy and (b) that the transfer of functions from state enterprises to the government will need to be taken into account in setting limits to the size of the deficit. If this is not possible, then the budget deficit should be deemphasized. In any case, structural reforms and total credit expansion must receive the full attention they deserve.


  • Begg, David, and Richard Portes, “Enterprise Debt and Economic Transformation: Financial Restructuring of the State Sector in Central and Eastern Europe,” CEPR Working Paper No. 695 (London: Centre for Economic Policy Research, 1992).

    • Search Google Scholar
    • Export Citation
  • Blejer, Mario, and Adrienne Cheasty, eds. How to Measure the Fiscal Deficit: Analytical and Methodological Issues (Washington: International Monetary Fund, 1993).

    • Search Google Scholar
    • Export Citation
  • Blejer, Mario, and Ke-young Chu, eds. Measurement of Fiscal Impact: Methodological Issues, IMF Occasional Paper No. 59 (Washington: International Monetary Fund, 1988).

    • Search Google Scholar
    • Export Citation
  • Fry, Maxwell, “The Fiscal Abuse of Central Banks,” IMF Working Paper No. 93/58 (Washington: International Monetary Fund, 1993).

  • International Monetary Fund (1992a), World Economic Outlook (Washington: International Monetary Fund, October 1992).

  • International Monetary Fund (1992b), World Economic Outlook, Interim Assessment (Washington: International Monetary Fund, December 1992).

    • Search Google Scholar
    • Export Citation
  • Musgrave, Richard, The Theory of Public Finance (New York: McGraw-Hill, 1959).

  • Stiglitz, Joseph, The Economic Role of the State (Oxford: Basil Blackwell, 1989).

  • Tanzi, Vito. “Fiscal Policy and Economic Reconstruction in Latin America,” World Development, Vol. 20 (May 1992), pp. 64157.

  • Tanzi, Vito., “Fiscal Policy, Growth, and the Design of Stabilization Programs,” in Fiscal Policy, Stabilization, and Growth, ed. by Mario Blejer and Ke-young Chu (Washington: International Monetary Fund, 1989).

    • Search Google Scholar
    • Export Citation
  • Tanzi, Vito, Mario Blejer, and Mario O. Teijeiro, “Inflation and the Measure of Fiscal Deficits,” Staff Papers, International Monetary Fund, Vol. 34 (December 1987), pp. 71138.

    • Search Google Scholar
    • Export Citation
  • Wallich, Christine I., Fiscal Decentralization: Intergovernmental Relations in Russia (Washington: World Bank, 1992).


Vito Tanzi is Director of the Fiscal Affairs Department. He holds a doctorate from Harvard University. Very useful comments were received from S.J. Anjaria, Gerard Belanger, Guillermo Calvo, Carlo Cottarelli, Manuel Guitian, George I den, John Odung-Smee, and several colleagues within the Fiscal Affairs Department. The author is solely responsible for the views expressed.


Recall that, in market economies, public sector activity is largely justified by (private) market failure. See Stiglitz (1989) and Musgrave (1959).


This does not imply that the government needs to assume all the functions now carried out by the enterprises.


It also implies that the fiscal activities of the government can be separated from the monetary activities. For example, the fiscal activities of central banks must be assessed as part of fiscal policy.


The functions not taken over by the budget should no longer be the responsibility of the public sector and the state enterprises. By adjusting money wages, workers could be given greater discretion over the use of the total compensation they receive while the state enterprises could stop having to provide social services,


Throughout the paper, the term “budget deficit” is used for that part of the fiscal deficit that is actually accounted for by the budget. This is the measure of deficit normally reported in newspapers and in government reports. The theoretical concept of the fiscal deficit is much more comprehensive and much more difficult to measure (see various papers in Blejer and Cheasty (1993)). It extends to the whole public sector and includes quasi-fiscal deficits.


There is no available estimate of the magnitude of these social functions. The World Bank is attempting to quantify them for Russia. The assumption is that they are quite important.


These limits may be self-imposed or negotiated with international institutions or other creditors.


The current situation in Russia, where the central bank continues to subsidize the state enterprises in particular sectors through subsidized credit and, in the process, creates inflationary pressures, shows the importance of this issue.


For one thing, it is difficult to calculate the subsidy element of subsidized credit. See, for example, the paper by Michael Wattleworth, “Credit Subsidies in Budgetary Lending: Computation, Effects, and Fiscal Implications,” in Blejer and Chu (1988). Also, subsidies, given through the budget, would most likely be associated with more careful, or at least more politically debated, decisions about which enterprises should be subsidized.


This measure would count the social expenditure by state enterprises in the same way as budgetary expenditures. It would also count the subsidies given by the banking system as public spending. See, for example, Begg and Portes (1992).


The relationship between the fiscal deficit, public debt, and the inflation rate was discussed in Tanzi, Blejer, and Teijeiro (1987). For references to the impact of this factor on the fiscal deficit of Latin American countries, see Tanzi (1992).


The situation is changing. For example, the unemployment rate in Hungary has risen from less than 1 percent in the first half of 1990 to more than 11 percent in September 1992. In Bulgaria, it has risen from 1.6 percent in 1990 to 14 percent in 1992; in Poland, from 6.3 percent in 1990 to 13.5 percent in October 1992; and in Romania, from zero in 1990 to 10 percent in 1992. For the fall in output up to 1991, see IMF (1992a. p. 46). For the estimated fall in 1992, see IMF (1992b, p. 19).


Rates of 25 percent or higher were recorded.


In fact, privatization may in some cases raise the size of the fiscal deficit if the social expenditures that the government has to take over from the state enterprise exceed the revenue from the sale of the enterprise.


If the government is unable to pay the interest on the foreign debt, there will be a difference between accrued and cash measures of the budget deficit. Domestic arrears to and from the government also create differences between cash and accrued measures of the budget deficit.


The difficulties of the central governments of many of these countries, in controlling the accounts and the actions of the whole public sector, indicate that the budget may not be the beneficiary of devaluation.


This discussion points to the importance of bringing the foreign trade area within the tax net. For example, it is important for imports to be subjected to the value-added tax.


For example, a solution to the bad debt problem of the commercial banks and the arrears among enterprises might be delayed by the fact that the solution would raise the budget deficit. Begg and Portes (1992) have argued that if the deficit were properly measured, these policies would not change its size. However, the assumption of the bad debt of the banks by the government, by replacing bad assets with good assets, would allow them to give more loans that might increase the liquidity of the economy.


There will also be an incentive to reduce the cash deficit by postponing payments, thus increasing arrears.


See the recent paper by Fry (1993).


When the financial market is not well developed, or when the credibility of the government is not good, this source of financing is not available. However, if the deficit is to be financed through inflationary finance, it may still be better if this finance is directly allocated to the budget and the budget finances the various activities including subsidies to state enterprises.


This is not only a problem of economies in transition. In fact, it has been a frequent problem in adjustment programs. See Tanzi (1989).


However, because of the changes in relative prices during the transition, it may not be wise to aim for an inflation rate that is too low since this would require a fall in prices in some sectors.