Reforming Budget Systems in Countries of the Former Soviet Union.
It was no wonder that opening these countries to democratic choice, to economic freedoms, and to international markets brought economic collapse. However, markets will develop on their own--even in Soviet times, people survived by informal exchange and, while unaccustomed to many elements of modern marketing, finance, and business management, they did understand trade, as any visit to a FSU country today will quickly confirm. Given a modicum of economic freedom, even with ambiguous property rights and doubtful enforceability of contracts, markets emerged in response to individual demands for goods and services.
Government budget processes and procedures--the systems for making public sector resource allocation decisions --do not freely emerge like markets.(2) The fiscal system consistent with Soviet times was not compatible with markets, free choice, personal and economic freedom, and democratic decision making. The old system must stop and a new system--a framework of laws and human and physical resources appropriate to new tasks--must be created by conscious and concerted actions. Early successes with the development of markets in some FSU countries, notably the Russian Federation, mislead many observers into believing that fiscal reforms were working. Continuing evidence, however, including the great Russian crisis of 1998--a collapse emerging from unsustainable budget deficits and default on federation and regional debt issues--shows how wrong that observation would be.(3) Absence of restructuring and reform in the public sector are easily apparent throughout the FSU. Examples include contentious and confused fiscal relationships between tiers of government in the Russian Federation; budget processes unable to constrain fiscal deficits or to manage allocation of scarce resources among governmental programs; inability to pay wages, pensions, and other public obligations as they come due; unreformed tax systems that seem more designed to punish enterprise than to raise revenue; tax collectors and government agencies that work with barter instead of money; and tax administration that thinks in terms of collection by brute force rather than voluntary compliance.(4)
The following sections, based on several years of direct field work including extensive discussions with senior government officials in several FSU countries, explore critical problems in the budget systems and procedures of these countries that have complicated or prevented their contribution to the normal expectations of what these systems must do: (1) maintain aggregate fiscal discipline, (2) assure direction of public resources to areas of greatest national need, and (3) encourage efficient use of resources in public agencies (Campos and Pradhan 1996, 1997; Schiavo-Campo and Tommasi 1999). Even though public budgeting is a planning process in both centrally planned and market economies, the budget environment and the expectations for the budget are considerably different in these systems and the transition in process has been difficult. In spite of the years elapsed since the collapse of the Soviet Union and the independence of these states and considerable progress in developing effective and accountable systems in some countries, most of the budget systems have many changes ahead if they are to assure that government services will be within the actual available government revenues and will be appropriate to the needs of the public.
Economic Transition and Fiscal Turmoil
The economic transition after the fall of the Soviet Union brought huge economic decline in the countries of the FSU. Traditional markets for products of the Soviet state disappeared and production was not competitive in international markets; economic activity fell dramatically, as data in table 1 demonstrate. Real GDP in 1998 on average was only 56.2 percent of its level of 1990 (53.1 percent, if the somewhat more successful Baltics are excluded), on the eve of the collapse of the Soviet Union. However, much of the collapse occurred in the early days. From 1995 to 1998, the average is 106.9 percent higher (104.4 percent, excluding the Baltics), only a modest increase, but still far better than experienced in the early days.
Table 1 Gross Domestic Product and Consumer Prices in Countries of the FSU Average annual County Real gross domestic product consumer price change 1998 as 1998 as First year 1990/ 1995/ % 1990 % 1995 of growth 1998 1998 Armenia 46.2 115.1 1994 345.5 13.7 Azerbaijan 49.3 117.9 1996 260.4 7.2 Belarus 80.0 124.0 1996 361.1 63.2 Estonia 81.8 120.5 1995 93.1 14.0 Georgia 37.8 127.6 1995 395.4 15.6 Kazakhstan 61.2 99.7 1996 288.7 20.6 Kyrgyz Republic 60.9 120.1 1996 147.7 22.4 Latvia 56.5 114.2 1994 80.3 10.1 Lithuania 67.2 116.0 1995 124.1 12.6 Moldova 34.5 88.7 1997 167.8 13.9 Russia 57.6 92.6 1997 202.7 29.4 Tajikistan 41.3 102.4 1997 377.9 140.7 Turkmenistan 43.3 71.5 506.9 186.2 Ukraine 37.0 85.8 336.6 32.2 Uzbekistan 87.7 107.0 1996 249.3 50.3 Average: 56.2 106.9 262.5 42.1 Sources: IMF World Economic Outlook 1999; IMF Country Reports, various countries, various years.
The end to administered prices, combined with an effort to continue the old Soviet practice of financing budget deficits with creation of money, brought stunning rates of inflation: the average annual rate of change in consumer prices from 1990 to 1998 was 262.5 percent (303.3 percent without the Baltics). Again, the period after 1995 has not been so dramatic, as the effects of early price-liberalization work their way out and more governments realize that seignorage is not an all-purpose source of revenue in a market environment. The experience is good only in comparison with the earlier regime: an average annual increase of 42.1 percent (49.6 percent without the Baltics).
Table 2 shows that the collapse in economic production brought with it a collapse in general government expenditure and government revenue. Between 1992 and 1997 real budgeted government expenditures have been approximately halved across the FSU, with declines of approximately 84 percent in Tajikistan, 80 percent in Azerbaijan, 76 percent in Georgia, 63 percent in Ukraine, 62 percent in Armenia, 54 percent in Turkmenistan, 53 percent in Moldova, 52 percent in the Kyrgyz Republic, and 47 percent in Russia. Only in Estonia and Latvia have real expenditures increased, while in Lithuania they have remained constant.(5) Wide-spread currency devaluation and budget cutting brought continuing expenditure decline during 1998. For example, 1998 expenditures in Ukraine were cut by 5.6 percent of GDP compared to 1997 (IMF 1999/42). Further, these reductions underestimate the decline, as budgetary figures do not reflect the elimination of past off-budget social expenditures by state enterprises for health care, education and housing.(6) For example, estimates of social spending by enterprises in Russia for 1992 equaled nearly 4 percent of GDP (Freinkman and Starodubrovskaya 1995).
Table 2 General Government Activity in Countries of the FSU General General Real 1997 General government government govern- government Country revenue expenditure ment balance (percent (percent spending (percent GDP) GDP) relative GDP) to 1992 1997 1992 1997 1992 1992 1997 Armenia 29.1 17.4 66.7 24.1 37.8 -37.6 -6.7 Azerbaijan 61.5 17.4 57.9 20.2 20.4 3.5 -2.8 Belarus 46.0 40.9 48.8 42.1 70.8 -2.8 -1.2 Estonia 34.6 39.4 34.8 37.0 115.8 -0.3 2.4 Georgia 19.0 10.4 53.5 15.3 23.7 -34.5 -5.0 Kazakhstan 24.5 23.4 31.8 27.1 63.7 -7.3 -3.7 Kyrgyz Republic 17.5 17.6 35.1 23.3 47.9 -17.6 -5.7 Latvia 28.1 39.0 28.9 37.6 124.0 -0.8 1.4 Lithuania 31.6 33.5 31.1 35.4 99.8 0.5 -1.9 Moldova 30.3 34.3 54.2 41.1 47.5 -23.9 -6.8 Russia 38.3 33.0 56.7 40.4 53.3 -18.4 -7.5 Tajikistan 26.6 11.6 57.8 15.0 15.9 -31.2 -3.4 Turkmenistan 42.3 29.2 28.9 29.2 46.2 13.3 0.0 Ukraine 34.0 38.4 58.0 44.0 37.2 -24.0 -5.6 Uzbekistan 31.5 30.2 43.7 33.0 72.9 -12.2 -2.8 Average: 33.0 27.7 45.9 31.0 58.5 -12.9 -3.3 Sources: IMF World Economic Outlook 1999; IMF Country Reports, various countries, various years.
Some of the decline was warranted: the governments were involved in economic activities which perfectly well belonged in the private sector.(7) Business enterprise, left alone, could handle industrial and commercial ventures far better than could the governmental bureaucracy. The changed system ended the tie between state enterprise and provision of housing and other services to individuals working for those enterprises. Furthermore, lower government spending would be expected, in light of the reduced government revenues, as shown by table 2 data. During 1997, revenues averaged 27.7 percent of GDP (30.9 without the Baltics), compared with 33.0 percent in 1992 (33.4 without the Baltics).
The adjustment in expenditures did not match the reduced revenues, however, and the countries ran fiscal deficits. As a percentage of GDP, these deficits have steadily declined since 1992, but have yet to reach sustainable levels. While government deficits were generally huge in 1992 (averaging 13 percent of GDP), most governments had reduced them to levels well below 5 percent of GDP by 1997 (averaging 3.3 percent). This performance still left problems, however, and a new wave of fiscal instability occurred with devaluation in 1998 when, for example, the fiscal deficit quadrupled in Lithuania in the second half of the fiscal year from 2 percent to 9 percent of GDP, with a year-end total of 5 3/4 percent (IMF 1999/ 73). Problems linger. First, private markets--internal and external--seldom purchase all the new debt, so governments continue to resort to the printing press for finance. Second, smaller reported deficits often mask a larger true deficit--the governments simply do not pay their workers, pensioners, and suppliers.
This decline infected critical governmental services. Expenditures for functions such as education and health have experienced substantial real declines as table 3 demonstrates. Real education expenditures by 1997 were only 13 percent of their 1992 levels in Tajikistan, while they experienced a 70 percent decline in Georgia, a 60 percent decline in Ukraine, and a 58 percent decline in Azerbaijan. Health expenditures show a similar pattern, with a 71 percent decline in Armenia, a 69 percent decline in Tajikistan, a 65 percent decline in Georgia, and 60 percent decline in Azerbaijan. Some countries have, however, seen limited real increases, with Latvia increasing real expenditures in both service areas. Kazakhstan has also increased real education spending and Belarus, Lithuania, and Turkmenistan have done so for health services. Lithuania and Turkmenistan are within 10 percent of 1992 levels for education as are health expenditures in Estonia and Russia. Still, overall education and health expenditures are down by 29 and 24 percent across the FSU and expenditures for health and education services have declined as a portion of GDP.
Table 3 Real Government Spending on and Employment in Health, Education, and Public Infrastructure in FSU Countries (1997 Relative to 1992) Infras- Education Health employment tructure employment and and real expendi- Country real expenditures, expenditures, 1997 ture, real 1997 relative relative to 1992 1997 to 1992 relative to 1992 Expendi- Employ- Expendi- Employ- tures ment tures ment Armenia 58.2 101.4 28.7 97.4 52.84 Azerbaijan 42.3 93.5 39.8 93.5 24.49 Belarus 77.1 115.5 127.1 100.5 29.33 Estonia 58.9 107.2 93.8 103.0 151.41 Georgia 29.9 51.8 34.7 51.8 82.27 Kazakhstan 114.7 62.2 74.4 79.3 N/A Kyrgyz Republic 82.4 71.9 66.2 87.7 177.91 Latvia 137.7 103.6 118.3 102.4 70.61 Lithuania 92.7 63.4 116.0 63.4 60.73 Moldova 71.2 66.8 84.0 137.6 30.31 Russia 85.0 92.3 98.7 92.3 N/A Tajikistan 13.2 78.8 31.0 79.3 6.66 Turkmenistan 93.5 108.1 103.1 122.4 137.1 Ukraine 40.0 84.6 58.8 94.9 4.52 Uzbekistan 68.1 109.3 67.0 96.3 206.79 Average: 71.0 87.4 76.1 93.5 79.61 Note: Inconsistencies in data exist across sources. In all cases, anomalies were resolved in a manner favoring consistency in data over time for each country, with data from individual country reports generally prevailing over compendium sources. Rather than precise measures, figures presented should be viewed as estimates. Particularly for capital expenditures, definitional variations across time diminish the accuracy of comparisons. The occasional co-mingling of government and private components of employment in health and education reporting also diminishes precision. However, the overall dominance of public sector employment in these areas significantly limits any misinterpretation. Unavailable or anomalous data required the substitution of data as follows. * For education spending, unavailable 1992 data required the use of 1993 figures for the Kyrgyz Republic and 1994 figures for Armenia. Unavailable 1997 data required the use of 1996 data for Latvia and Lithuania. * For education employment, 1996 figures were substituted for 1997 for Estonia; figures for Latvia were calculated based on percent of labor force; unavailable 1997 figures required the use of 1996 data; 1991 and 1996 figures were substituted for Tajikistan. * For health employment, 1996 figures were substituted for 1997 for Estonia; anomalous 1992 data for the Kyrgyz Republic required the use of 1991 figures; figures for Latvia were calculated based on percent of labor force; unavailable 1997 figures required the use of 1996 data; 1991 and 1996 figures were substituted for Tajikistan. * For health expenditures, unavailable 1992 data for Maldova required the use of 1993 data. * Separate health and education employment were sometimes unavailable. Table entries are based on total government/state employment for Azerbaijan and Lithuania (for 1993 and 1998); combined health, education, and science employment for Georgia; and combined health, education, science, social, and cultural employment for Russia. * For capital spending comparisons, unavailable or anomalous 1992 data required a substitution of 1993 figures for Azerbaijan, Georgia, and Lithuania, 1994 data was used for Turkmenistan and Uzbekistan. Sources: Calculation based on World Bank, FSU *STARS* 1995, States of the Former Soviet Union 1995: data on diskette; World Bank, Development Database 1999; IMF Country Reports, relevant countries 1997, 1998, and 1999; IMF World Economic Outlook 1999.
Expenditure declines are often not associated with corresponding declines in staffing. For example, with 42 and 71 percent declines in health and education expenditures in Armenia, employment has declined by less than 4 percent in education and less than 2 percent in health since 1994 (IMF 1998/22), with comparisons from 1992 showing even less decline in health and an increase in education employment. Table 3 shows that across the FSU education employment has been maintained at 87 percent of 1992 levels (compared to 71 percent for expenditures), while health has been maintained at 94 percent (compared to 76 percent for expenditures). Education employment has increased in five countries while real expenditures declined significantly. In only four countries have declines in relative education employment exceeded expenditure declines. On average, for the nine countries (excluding Tajikistan) that have seen more than a 20 percent real decline in education spending, employment levels continue at levels 45 percent greater than expenditures (90 percent greater if Tajikistan is included). For health, in none of the 11 countries (except Russia) experiencing real expenditure declines has employment equally declined. In the eight countries experiencing more than a 25 percent decline in real health expenditures, health employment levels have been maintained at levels averaging 73 percent greater. These employment levels are likely to be unsustainable at this level of funding.
Such dramatic reductions in revenues and public spending would stress even the most effective budgetary systems. Transition in budget expenditures appears to require an end to subsidies for consumer products; elimination of subsidies to state enterprises; reduced military spending; reliable, affordable, and better targeted social protection programs; direction of public investment programs toward higher yield endeavors; maintenance and enhancement of the social and economic infrastructure; reduction of surplus government employment; more attention to environmental protection; and government acceptance of social responsibilities formerly handled through state enterprises (Tanzi 1994, 157). These changes, though they seem reasonable enough, must be accomplished through the budget system, not through the mandate of a powerful dictator or ruling clique--that is a major part of what the transition from Soviet times really is about and these choices will not automatically emerge with transition to a free market.
Budget System Inadequacies for Transition
Under the Soviet scheme, "economic development ... [was] never allowed to be hampered by a lack of finance" (Wilczynski 1970, 146). Finance provided passive accommodation of the resource plan, and neither state enterprises nor ministries faced hard budget constraints. Within this philosophy, the financial budget was unimportant and ministries of finance were irrelevant except as bookkeepers. The result was a system incapable of fiscal discipline and reasonable budget allocation.
The budgeting structures of Soviet times were designed for the central physical plan. Gosplan (now subsumed into ministries of economy in most parts of the FSU) dominated decisions for government services and enterprise production. The plan included the entire economic sphere--government, industry, agriculture, etc.--without much differentiation between the parts. The conditions were those of fixed and administered prices, predefined production methods, and planned service distribution based on controlled demographic patterns and the public infrastructure. Features that made the budget systems proper for a stable and planned economic structure made it the worst of budgetary worlds for the transition environment. The budget structures emerging from the Soviet era carried traits from the old environment that hindered functionality in the new economic and political conditions.
The budget process outline has been in place in the FSU countries since the early days of transition--all countries adopted some sort of national budget code shortly after their independence.(8) The systems had the fundamentals: the codes require that the government prepare a budget proposal for a fiscal year, require legislative approval of the budget (some have two bodies, some have one, and not all are politically meaningful), and expect the approved budget law to be executed. (More recent revisions of these laws are much more comprehensive.) The annual budget law may not be approved by the time the fiscal year begins (not a surprise, in light of the difficulties that even seasoned democracies have with getting the budget done before the start of the fiscal year--but the FSU countries often have automatic continuing appropriation provisions to avoid government shutdowns), the budget law may be revised through the year and some countries have managed grave uncertainty and high inflation with several partial-year budgets approved through the year, and the approved budget may not be executed entirely. Furthermore, there may be quirks in the systems --for instance, the Milli Mejlis (parliament) in Azerbaijan may adopt or reject the budget submitted by the president, but may not change it(9) --but the procedural skeletons have been in place for some time.
The problems are primarily in the details, not the blue-print. The budget systems have weaknesses in (1) providing for transparent development and approval of a realistic, responsive, effective, and efficient plan for operation of the government, (2) attaining execution of the adopted plan, and (3) reporting and verifying the execution of that program. The adoption of unified treasury systems in many of the countries has greatly improved the capacity of the governments to mobilize, control, and monitor public funds, but the transformation of budget systems lags.
The budget coverage in FSU countries has typically been woefully inadequate, with much government money outside the normal budget process and documentation. Practices violate the basic rule of comprehensiveness that is so critical "to allow comparisons of the different ways a government uses financial resources" (Meyers 1996, 177). Indeed, treatment of resources can be completely nontransparent and outside general government budgetary scope. A comprehensive budget is critical for a full understanding of the policy intent and implications of government spending. Without including all funds, the budget fails to function for policy making, for deliberation, or for control.
In the early days of the break up of the Soviet Union, all levels of government created extrabudgetary funds. Many directed resources to state enterprises while other funds were controlled by government ministries. Special taxes or fees sometimes supported the funds, although some received general government revenues.
A review of FSU experience with integrating and controlling these funds in a comprehensive budget shows mixed accomplishments at best. In 1995-96 there were more than 1,100 extrabudgetary accounts or special funds controlled by ministries in the Kyrgyz Republic. Now, most have been eliminated or consolidated and folded into the treasury control system and budget. In Georgia, by year-end 1997, all special accounts of central government spending units were closed, except for accounts of pension funds and higher education institutions (IMF 1998/99). Remaining extrabudgetary funds, including the social security fund, employment fund, road fund, and privatization fund, equaled 19 percent of total revenue (1.9 percent of GDP) and 17 percent of total expenditures (2.5 percent of GDP), excluding the privatization fund. Belarus reorganized its budget and on January 1, 1998, incorporated extrabudgetary funds, with the exception of the social protection fund and special accounts of ministerial spending units. Newly incorporated funds include (in addition to defunct funds) the road fund, agricultural support fund, and employment fund. These funds were also incorporated into the central treasury, with local treasuries established during the year. The social fund and spending unit funds remain outside treasury control (IMF 1998/108). In 1997, monies excluded from the budget equaled 32 percent of revenue and 14.9 percent of GDP. Extrabudgetary funds incorporated in the budget for 1998 equaled 3.7 percent of GDP in 1997 and the excluded social fund, expected to have revenue of 10.8 percent of GPD, took in resources at a rate of 13.6 percent during the first quarter of the year.
In 1994, extrabudgetary funds in the Russian Federation included five social funds, the road fund, the ecological fund, the technological development fund, and 50 industrial funds (and all foreign exchange transactions and directed credits from the Central Bank). By 1998, the pension fund, employment fund, social insurance fund, and medical insurance fund accounted for expenditures equaling 9.3 percent of GDP, an amount equal to 63 percent of direct general expenditures of the Russian federal government or 30 percent of consolidated (federal and regional) general government expenditures of the federation (IMF 1999/100).(10) In Ukraine, "earmarked" funds represented 39 percent of total revenue and 14 percent of GDP in 1998 (IMF 1999/42). Estonian extrabudgetary social insurance fund expenditures equal 55 percent of on budget central and local general government spending in 1996 or 35 percent of all spending (IMF 1998/12). The 1999 budget was to be prepared on a general government basis: however, the consolidated framework was not implemented (IMF 1999/59). A new Lithuanian treasury law approved in March of 1999 provides for the establishment of a single treasury account and requires Seimas approval of any new extrabudgetary funds (IMF 1999/73).
At best, these funds impede the allocation of public resources to areas in greatest need of public attention--the same problem of "budgetary constipation" associated with earmarking in western fiscal systems.(11) At worst, the funds offer an easy avenue for corruption and an escape from transparent functioning of government, especially when they flow through bank accounts completely outside general government control or supervision.
Inadequate Disclosure of Special Funds
Statements of aggregate transfers from the state budget to major special funds such as social funds are often included in budgets but with minimal detail on the expected operation of these funds. The scope of these funds is not insignificant. For example, in Azerbaijan's 1998 budget, 340 billion manat were allocated from the budget to the social fund (an amount equal to 11.1 percent of total budget revenue and 9 percent of expenditures). Allocations to lesser funds, such as road, environmental, forest, and business assistance funds are often not individually identified in legislatively adopted budgets.
The fact that special funds have dedicated revenue sources or are part of a social insurance scheme is often used to justify reduced central oversight. This promotes both a lack of accountability for the usage of these resources and an artificial barrier to the consideration of these expenditures within the context of overall public budgetary policy.
Off-Budget Discretionary Accounts
In addition to extrabudgetary and special funds, discretionary funds controlled by ministries have been common. These off-budget accounts render budgets an incomplete statement of public resource allocation and can create an easy avenue for corruption and diversion of public funds. These funds, accumulated from the proceeds of fees charged by individual spending units of ministries, often contribute significantly to the financing of social sectors, particularly health and education. In the past, these funds often existed as separate spending unit bank accounts. While significant progress has been made by eliminating such funds (in the Kyrgyz Republic), by incorporating and controlling them through treasury accounts (in Kazakhstan and Azerbaijan), and by requiring ministries to submit proposals for their expenditure to the ministry of finance for approval (in Azerbaijan and Russia), these funds are generally not subject to appropriation and are expended outside the budget. The closing out of these funds has been difficult. With the introduction of an interim treasury system in the Ukraine in 1997, budgetary payments have been progressively incorporated and agency bank accounts closed. However, through the end of 1998, spending agency accounts were estimated to have remained at approximately 2 1/2 percent of GDP (IMF 1999/42).
This fee and charge revenue can be substantial and in at least one country it has been used to quadruple the official wage and salary paid to teachers and doctors.(12) Ministries also acknowledge the existence of additional informal fees paid directly to education and health staff. The Lukman health program has documented substantial unofficial payments for health services in Turkmenistan (Holland and Ensor 1996). Ministerial health officials in one nation have estimated that the formal and informal fee systems have accounted for as much as 65 percent of total annual institutional resources, relegating budgeted amounts to a minor contribution of 35 percent. A more common figure is 10 to 20 percent.
Government-capital infrastructure investment is generally necessary for economic growth and successful transition.(13) While the data are sketchy, table 3 suggests that real capital expenditures have declined by an average of more than 20 percent between 1992 and 1997. This average masks dramatic declines of 95 percent for Ukraine, 93 percent for Tajikistan, 75 percent for Azerbaijan, and 70 percent for Belarus and Moldova, with increases of 107 percent for Uzbekistan, a donor-financed 78 percent increase in the Kyrgyz Republic, a 51 percent increase for Estonia, and a 37 percent increase for Turkmenistan.
Capital expenditure decision-making processes in the FSU include mechanisms and incentives which are unlikely to produce efficient public capital investment decisions. In Soviet times, central planners located infrastructure according to the plan, not according to criteria that would make sense in terms of market conditions or preferences of the population. This centralization of infrastructure control continues in many countries of the FSU. Budgets often exclude externally financed capital expenditures but do not include budgetary implications of capital spending occurring under government guaranteed credits. Where the spending is included in budgets, it may be presented as a lump sum, not divided among ministries ultimately responsible for the infrastructure. This structure erodes budget comprehensiveness and prevents rational trade-offs regarding the operating and capital components of the budget.
There are also problems regarding governmental responsibility for capital expenditures. The FSU countries commonly assign financial responsibility and decision making for capital projects to central governments and maintenance and operating responsibilities to subnational governments. For example, local administrations are generally responsible for the operation and maintenance of schools and health facilities, but generally not for their construction (or even the decision as to whether or not to construct a facility). Capital projects are often financed based on annual negotiations between local or regional governments and ministries: for example, the ministry of finance and possibly a ministry of economy. This separation of capital decisions from operating decisions is potentially inefficient. Subnational administrations may attempt to shift direct costs away from maintenance activities toward other spending categories under the assumption that higher level budgets will cover the long-term costs of such decisions by granting future capital and major reconstruction funding. Thus, the split tends to lead to maintenance disincentives, as local administrations may shift these costs to higher levels through capital construction/reconstruction.
The Blur between Public and Private
The Soviet state made little distinction between governmental or social functions and functions that would become private with the transition to a market economy. Combining the total economy within a single resource allocation process, while necessary for a central economic resource plan, is not appropriate for an economy in which many choices are to be made by individuals and private enterprises. The problem for FSU countries has two dimensions: enterprises producing goods and services for the private market must provide some normal government services, and state ministries operate market enterprises within their organizational structure. Both create difficulties for the efficient and effective functioning of the budget system (and of the economy).
Social Responsibilities of Commercial Enterprises
The Soviet state assumed that many governmental services--early, primary, and secondary education, recreation, many municipal services, health--would be provided through the enterprise for which the individual worked. It was part of the concept of the individual and the state as a holistic production machine. When open markets started to emerge, enterprises needed to shed governmental functions if their production costs were to be competitive with private producers. Many of these functions are best handled by local governments, but these governments frequently receive only limited transfers from the national governments and typically have almost no authority to raise revenue on their own. What independent resources they have are usually limited to charges and some miscellaneous licenses of modest productivity.
The FSU countries have faced a double budgetary challenge: (1) to determine what assets under the control of state enterprises in Soviet times should be divested from those enterprises to the balance sheets of local government and then (2) to assign revenue resources that will be appropriate to these new responsibilities. Neither is a simple task for decision makers brought up in the Soviet system, but both are critical for defining and distinguishing private and public responsibilities. Competitive forces from global markets have forced most FSU countries into divestment of these assets, but few have established revenue flows to localities adequate to the responsibilities--with the result being dramatic decline in many government services now provided by subnational governments.
Commercial Operations of Government Ministries
The lure of market revenue and stressed budgets have induced government institutions in several countries to establish commercial operations to supplement budgetary resources. Sometimes the product or service sold is closely associated with the governmental function, but sometimes there is little relationship. There are problems in both instances. It is not unusual for schools to have farming operations, for ministries of defense or justice to operate hotels and restaurants, or for natural resources (particularly oil and gas) to be developed by governmental agencies. In contrast to the revenues entering the accounts and funds noted in an earlier section, these commercial operations have no discernible relationship to a core government function. Given that the revenues are often entirely off-budget, incentives exist to divert resources away from normal governmental functions of the ministry to these activities in an attempt to generate ministerially (and even personally) appropriable and discretionary revenues. The result is likely to be neglect of the basic operations of the institution (as staff follow individual incentives) and a further reduction in the level and quality of services. The problem is especially acute in Turkmenistan, where ministries routinely operate hotels and other commercial establishments. In addition, self-financing ministries and agencies may be entirely outside both the budget and treasury accounts. These include activities involving agricultural development, oil and gas, transportation, and communication. Of course, some of these are purely commercial and the reasonable remedy would be to shed them from government ownership. In some countries, for example, Azerbaijan, many of these entities receive abundant government subsidies and are self-financing in name only. The problem is greatest when the single government agency mixes commercial development, regulation, and policy functions--and probably also serves as revenue collector for the government.
Even when the commercial operation is related to the core governmental function, problems remain. Limited funds from the budget have induced governmental institutions, notably schools and hospitals, to charge for services that are, as a matter of government policy, supposed to be available without charge to any person--formal and informal fee systems have transformed government policies of free and open health and education systems into charge based systems. While establishing a fee-for-service financing structure can make sense for certain specialized services (so long as the revenues are controlled by the budget system), fundamental services of these institutions have a strong public good component. Failure to make them freely available to all will have long-term impact on the nation. For instance, charging fees for primary education--not unheard of in some FSU countries--and denying the service from those who do not pay will have devastating social consequences. The lure of revenue from certain parts of financially-stressed organizations, especially when that revenue is outside budget control and financial accountability, will distort operations away from the public interest.
Problems in Budget Formulation
"[I]t is not possible to implement well a badly formulated budget" (Schiavo-Campo and Tommasi 1999, ch. 1, 1). In most of the FSU countries, budget arithmetic continues to displace concerns with public priority and the quest for efficient delivery of government services. Much of this is a direct outgrowth of the general unimportance of finance and the vital position of the resource plan in Soviet times. The systems continue to pay little heed to financial limitations (resources were driven according to the physical plan and finance was passive under the old regime) or to responding to the preferences of the citizenry. That conflicts with the development of a public sector consistent with the environment of a market-oriented economy. In the old system, finance was passive: whatever result emerged from the plan would be financed by adjustments in markups in state-controlled prices, by adjusting payments taken out of wages before the employee was paid, and, if necessary, by printing more money or by issuing new credits for enterprises or agencies. Finance by money creation was not inflationary because prices were rigidly controlled; shortages might occur and quality of product might suffer, but prices would not rise. Ministries of finance were as passive as finances. Ministries of finance served no role in budget choice or priority setting, but spent their time checking the arithmetic in calculations based on the plan and the norms. There were no meaningful budget (financial) constraints on spending units. Institutional structures were woefully inexperienced and inadequate to deal with the uncertainty of a market based economic structure.
Problems with current budget processes begin with weak instructions issued to ministries to guide preparation of their budget estimates and proposals. The instructions emphasize mechanical details, reminders of rules and regulations under which the ministries must operate, lists of forms to be used, and timetables. These instructions lack policy direction, are issued relatively late in the budget-development process, and they include no indication of the general level of resources that might be available either to the ministry or to the government as a whole.(14) The formulation of regulations and arithmetic is fine under the old system of central plan, but completely inappropriate for transition. As it stands, budget submissions of ministries are largely irrelevant in determining the final budget adopted, and the expertise of line ministries for responding to national policy is lost.
Some countries, notably the Kyrgyz Republic and Kazakhstan, have recently made preparation more meaningful by issuing budget instructions earlier in the budget cycle and by giving ministries an expenditure envelope within which they prepare their budget proposals. Such expenditure envelopes, determined with the backing of the government and issued by the cabinet, are critical to developing budget priorities (Campos and Pradhan 1997). On the whole, however, the FSU countries still have much to do to make ministry budget proposals an effective and meaningful part of public expenditure planning and management.
Revenue forecasts in FSU countries have been almost hopelessly inaccurate. Revenue collections would often miss the forecast by 25 to 50 percent, with catastrophic effects for budget execution. Political manipulations can be involved here as they are in countries with far greater experience with forecasting in a market environment, but in the FSU an endemic problem has been the use of revenues flowing from achievement of the economic plan instead of a realistic forecast of what likely economic conditions would yield. In many cases, the forecast has been produced by asking enterprises what their economic plan would yield for the year, usually without giving guidance as to what sort of economic-market conditions were anticipated. The accumulated results gave the revenue forecast. In the transition environment, the plan stood no chance of being realized and the forecast revenue stood no chance of being produced---even if there were perfect collection of all revenue owed. Basing the forecast on state enterprise production plans in the face of disappearing markets for old Soviet products, not on a realistic expectation of what could be expected in that difficult economic environment, was a prescription for fiscal disaster. The failure to execute the revenue forecast then requires a higher-than-anticipated deficit, cuts in government spending, or failure to pay bills when they come due. Many problems in budget execution stem from these unrealistic revenue forecasts.
Making the shift from planned to forecast has been difficult. Budget building via revenue forecasts from a probabilistic future rather than as levels predetermined and controlled through a planned economy has required a rethinking of some foundations of budgeting in these countries. Progress made in economic stabilization and International Monetary Fund (IMF) assistance with forecasting have greatly improved the accuracy of more recent revenue forecasts. These forecasts have begun to be integrated with budget development. However, this integration often occurs late in the process, and revenue and expenditure forecasting tends to not have extended beyond the single budget year. Baltic countries appear to be somewhat more advanced in this area, however. For example, Latvia began a two-year rolling government budget submission in October 1997 linked to a medium-term macro-economic framework (IMF 1999/77). In Central Asia, the Kyrgyz Republic 1998 budget procedures law also requires the linkage of the budget to a medium-term financial planning horizon. Realistic short- and medium-term forecasts are critical to establishing fiscal discipline (Campos and Pradhan 1997).
Adequate spending unit flexibility and accountability for performance are critical in establishing budget formulation and execution which effectively carry out government priorities and meet programmatic needs of populations (Campos and Pradhan 1997). Budget construction throughout most of the FSU is driven from bottom-up processes, often based on out-of-date personnel and facility norms (regulations) which were once centrally sanctioned as a budget calculation basis, but have no relationship to finances available during the transition. Priority setting is largely ad hoc and often occurs as a residual decision after initial budgetary requests are made by sectoral ministries and spending units. In many cases, little prioritization occurs in the initial submission of proposed budgetary requirements from sectoral ministries.
Norms for the calculation of expenditure requests lock ministries and spending units into a specific production process which is often outdated and inapplicable to the current situation. Rigid production relationships imposed from the top are entirely wrong when resources have fallen, when prices (and input price ratios) are changing, and when customers are moving. For example, suppose a ministry of health wished to alter the assignment of doctors and support staff. Because of regulations regarding the number of staff per hospital bed, the ministry of health must first reduce the number of beds supported. If the intended innovation was to alter this distribution, then the regulation is self-defeating. Similarly, ministries can shift resources between institutions, but oftentimes not without first shifting physical resources (again, reallocating support resources by reallocating beds). This perpetuates existing (unattainable and inefficient) service-delivery structures, denies the ability of ministries to exercise reasonable discretion and expertise in service delivery, and eliminates any ability to reasonably hold ministries and spending units accountable for delivering services appropriate to population needs. The norms are usually doubly damaging. Not only do they produce inefficient and ineffective expenditure proposals, they are usually designed to fund and support government physical infrastructure (for instance, hospitals and schools), not to deliver services to the citizenry.
The result of this estimating exercise is an entirely unrealistic total for ministries and for the government. In a typical example from one FSU country, the ministry of education operating budget request for 1996 was 282 percent of the amount finally allocated by the ministry of finance. In the same country, total initial requests by budget units for 1997 were nearly five times the final budget total, and the ministry of finance estimated that only 15 percent of requested requirements were provided for in the 1998 budget. When proposals are for multiples of the amount reasonably attainable, the request loses value as a prioritizing or policy vehicle. Calculating displaces priority setting.
Sectoral departments of ministries of finance verify calculations and review and reduce these requests which have been made oblivious to the existing budget constraint and without substantive policy guidance. The reductions usually appear arbitrary and focus on specific line-items (for example, economic articles) across an entire sectoral ministry for categories such as utilities, supplies, and capital, with the ministries to determine how these reductions will be allocated across spending units within the constraints of norms. This process is a remnant of the structural and norm based allocation mechanisms under the Soviet system, but the finance ministry lacks information from line ministries and executive policy guidance that would allow it to do anything else.
Budget submissions omit details of operating plans, substantive program proposals and policies, and comparisons between current and previous budget years. The historical role of the ministry of finance--a verifier of calculated coefficients--dominates and makes the ministry hesitant to insert itself in sectoral decisions other than line-item allocations. In the dynamic fiscal and budgetary environment of transitional economies (and market economies, in general), this ministry of finance reluctance is inappropriate. Relevant information must be required in spending unit budget submissions to allow the ministry of finance to perform this function. Arbitrary spending reductions may, for instance, cause allocations for electricity for schools and hospitals to be only a fraction of expected consumption or even zero in regional budgets in Russia and elsewhere, with full understanding that normal consumption will be continued. The result is an abrogation of meaningful decision making and programming of arrears into the budget in the form of "forward commitments" (Schiavo-Campo and Tommasi 1999, ch. 7).
Problems in Budget Presentation and Adoption
Budgets presented to parliament lack information necessary for reasoned decisions about government proposals, and appropriations emerging from parliament do not establish clear accountability for the use of public funds. With only broad totals by functional classification and economic article, the budget as adopted also provides insufficient detail to provide legal accountability for budget execution.
Materials Sent to Parliament
Parliaments in most FSU nations receive only rudimentary requests, minimal supporting material, and little background information for assessment of government proposals. The typical single-year presentation, a common practice in early days for all FSU countries that continues in some, lacks the longitudinal perspective provided by the classic three years: most recent closed year, the current year, and the budget year. Statements of priorities of services or service levels are not presented, nor is any information regarding the details of proposed programs included. The presentations are normally classified by function and by expenditure article and not by administrative unit--a carryover from the era of the central plan. Under the plan, the only issues about the budget were whether the regulations were followed and whether the arithmetic was right--and that eliminated a need for all those accompanying materials.
Single-year data presentation provides no context and conflicts with the generally recognized need for longitudinal information for informed budgetary decision making. Without comparative information it is virtually impossible to assess the longer-term nature of priorities or for the document to serve the needs of planning or accountability. Without longitudinal detail on the financing of particular programs or units, it is not possible to hold administrative units accountable within the context of past budgetary allocations. A presentation for only the budget year makes analysis almost impossible.
Accountability for Resources
FSU countries use a functional expenditure classification in the adopted budget, organized by broad function and controlled across economic expenditure article (line-item). Exhibit 1 provides a typical example of functional divisions along with the expenditure article breakdown. Budget submissions to parliament commonly provide little more than summary expenditure totals by function with supporting detail generally limited to government-wide and functional economic article breakdowns. The sectoral classification structure used prior to independence--the sectors were a critical element of the central plan--was easily converted to the functional classification recommended by the IMF and translatable into the government finance statistics (GFS) classification (Schiavo-Campo and Tommasi 1999, ch. 3). However, governments have been reluctant to track appropriated spending to ministries and spending units, even when tracking is available for internal administration. Ministries of finance translated budgets adopted by sector into budgets for each spending unit, but the translation occurred after the parliament acted and it was not part of the appropriation law.(15) Functional spending unit allocations made below the level of ministries have been largely at the discretion of the ministry, in accordance with expenditure article requirements and norms.
Exhibit 1 Broad Functional Budget Classification and Economic Article Expenditure Structure of FSU Countries Broad functional divisions Broad economic budget articles (line-items) General executive and legislative Wages and salaries (P) state services Deductions for social Defense contributions (P) Court, public order and security Current expenditures bodies General operating supplies Medicines (P) Social--Cultural funding Soft inventory and clothes Education Food (P) Health Business travel Social protection and insurance Transport Culture and physical training/ Communications recreation Utilities (communal services) Repair/rehabilitation Economy funding Research Household/housing Inventory Fuel and energy Rehabilitation of structures (P) Agriculture, forestry and Other current expenditures fisheries Mining and minerals Interest payments Industry and construction (internal/foreign credits and Transport and communication debt) Other economy Subsidies, payments and transfers Nonassigned expenditures Subsidies Intergovernmental payments State debt/crediting Nonprofit organization payments Pensions Reserve funds Grants/stipends (P) Transfers from extrabudgetary Capital investments funds Purchasing capital assets Capital construction Credits to other institutions Capital repair Intergovernmental transfers Creation of state resources Land and nonmaterial assets Principal payments (internal/ foreign) Lending State participation in internal investment State investment in foreign investment "(P)" Indicates articles which are often protected (that is, resources can be used for no other purpose) in several FSU countries. Sources: Functional Budget Classification code, Republic of Kazakhstan 1997; and State Budget Expenditures Functional Classification code, Ministry of Finance, Azerbaijan Republic 1996. State Budget Expenditures Functional Classification code, Ministry of Finance, Azerbaijan Republic 1996. Kazakhstan includes 1,800 levels in its functional classification (generally such detail is reserved for administrative purposes in the FSU); Azerbaijan's classification manual identifies 83 economic expenditure articles.
The functional classification serves policy planning well, but lacks the structure to identify the purposes to be achieved by the allocation of budget funds and does not afford clear accountability for expenditure of funds. Appropriating only to the functional level blurs legislative control, reduces transparency in use of funds and authority for specific programs and policies, and prevents clear legal responsibility for appropriated money. In general, a ministry or other spending unit can be held accountable for activities, operations, finances, and results; a function cannot. Without the link to administrative unit, there is no transparent and unambiguous accountability. Without administrative classification in appropriations, parliaments lack mechanisms for holding agencies accountable for proper expenditure of funds and program outcomes. Fortunately, some countries have now begun crosswalks to administrative unit classification.
Problems in Budget Execution
In light of the way budgets are formulated and approved, it is not surprising that there are problems in budget execution. The most important of these problems are extreme variances between adopted and executed budgets, development of budgetary arrears, execution of the budget through barter and offsets, and excessive control in execution. Proper budget execution requires mechanisms to assure: (1) conformity to financial and policy grants of authority; (2) adaptability to changes in the macroeconomic environment; (3) problem correction during program implementation; and (4) efficient and effective resource usage (Schiavo-Campo and Tommasi 1999, ch. 6). FSU systems fail across all of these dimensions.
Variance with Adopted Budgets
Actual spending often diverges significantly from the adopted budget. In the early days of independence, during near complete economic collapse and hyperinflation, budgeting was usually done quarterly or even monthly. Annual budgets were adopted, frequently well into the budget year, but they were transformed dramatically in execution so that a comparison between adopted and executed was often not very meaningful. For example, in 1994 Uzbekistan made quarterly adjustments to its budget, based on actual revenue and execution figures. While fiscal variances were rather small, these were the result of comparisons of actuals to the adjusted plan rather than comparisons to the initially adopted budget. A similar situation existed in Turkmenistan when the 1996 budget was adopted quarterly, but variations appeared to remain. Under the economic conditions the countries faced, it is difficult to fault them for the adjustments; forecasts were as horrible as the economic conditions, data were not available, and the countries had little experience with managing in the new environment. In general, as economic systems have become more stable, comparisons of adopted to actual expenditures have become more meaningful but the record continues to show problems.
Table 4 presents spotty data on budget execution for Ukraine, Russia, Kazakhstan, Azerbaijan, Estonia, and Latvia. Evident in this table are the substantial variations between what is budgeted and what is spent. Aggregate expenditure variations of more that 25 percent have not been uncommon. Further, these variations are underestimated to the degree that they are based on a comparison to adjusted plans. Also masked within this are even more dramatic variation in expenditures across function. For example, variation in the 1994 state budget of Kazakhstan ranged from -76.5 percent for social security and welfare to more than a 3,000 percent excess expenditure for communication and energy (World Bank 1996b). Likewise, the variations in social sectors such as health and education have been more pronounced than total budgetary variation in Azerbaijan (with health variations of 42, -37, 0, and -27 percent and education variations of 63, -26, -5, and -13 for 1994, 1995, 1996, and 1997). As would be expected, variations across individual articles of expenditure are significantly greater.
Table 4 Variation (as percentage) between Budgeted and Executed Expenditures (Selected FSU Countries, Selected Years) Country 1992 1993 1994 1995 Ukraine(1) -26.63 -35.63 -9.84 Russia(2) -5.10 -24.90 -24.94 Kazakhstan(3) -29.80 Azerbaijan 22.00 -36.00 Estonia(4) Latvia Country 1996 1997 1998 1999 Ukraine(1) -21.20 4.45 Russia(2) Kazakhstan(3) -10.8 -18.0 Azerbaijan 5.00 -6.00 Estonia(4) 0.39 Latvia -5.10 (1) Figures are calculated based on percent of GDP budgeted for public expenditures, versus those actually expended. Figures for 1997 reflect only first 9 months of budget execution. (2) These figures reflect departures for the budget as adjusted in December of each year. As such, it is likely that they seriously understate the level of annual variation. (3) These figures represent execution of the total state budget. For 1994, the republican component was executed at -42.5 percent, while territorial budgets were executed at -2.6 percent, figures for 1997 were -12.0 and -7.5 percent and for 1998 were -17.1 and -15.4. There were inconsistencies between figures presented by the IMF. Alternative calculations show a 1998 underexecution halved to -9.38 percent. Sources: Hinton 1998; World Bank 1996b, 72.; World Bank 1996a; Azerbaijan Ministry of Finance; and IMF 1999/59, 1999/77, 1999/95, 1999/96, 1999/99.
Flexibility in budgeting is desirable, particularly where economic conditions are volatile; because several FSU countries are heavily dependent on global natural resource markets, the volatility is likely to continue for many years. However, flexibility should be based on reasonable policy choices, not driven by chance. At the same time, to effectively deliver services, line agencies need predictability and some level of certainty regarding the flow of budgetary resources (Campos and Pradhan 1997).
A number of countries identify in their budgets (or in their budget system law) the priority for payment in case of revenue shortfall; the priority is normally for object class (such as, wages of government employees, pensions, medicines, and subsidies to individuals,) not for services that might be most significant for national strategic priority (see exhibit 1). Ministries often have no freedom to manage the budget reductions, but instead must work through the input-based priority, regardless of any acute service delivery needs of the population they serve. The prevalent systems afford neither adequate flexibility nor maintenance of control. Control is directed toward only economic article and critical sanctions against overspending (Campos and Pradhan 1997) are often nonexistent. The result is a breakdown in accountability of government managers for meeting the acute service delivery needs of the population.(16) A fictitious aggregate fiscal discipline is imposed in a manner which nullifies the implementation of programmatic priorities and fosters tremendous inefficiencies and ineffectiveness in service delivery.
Arrears, Barter, and Budget Offsets
Nonpayment and payment substitutes plague the countries of the FSU, complicate budget execution, and cloud many aspects of the budget process. The nonpayment disease and its transmission are deceptively straightforward. Once begun, these forms of transactions have been described as having "chain letter" effects, as one barter recipient needs to entice others to barter to dispose of goods received in the previous transaction (IMF 1998/58). In Russia, over half of industrial output is exchanged in barter. Promissory notes (veksels) are also used in inter-enterprise trade and their value is estimated to exceed the size of ruble broad money (IMF 1999/100). Total payable from trade credits reached over 70 percent of GDP in 1998 (while receivable were only 45 percent), with overdue debts reaching 40 percent and growing.(17)
Enterprises lack money to pay their taxes, so governments do not collect the revenue needed to execute their approved budgets. For example, in Russia total enterprise tax arrears average approximately 5 percent of GDP per year; the total stock of tax arrears is about 9 percent of GDP (IMF 1999/100). Tax arrears in Georgia reached 28 percent of collections (2.5 percent of GDP) by year-end 1997 (IMF 1998/99). In Ukraine, the stock of outstanding tax arrears at year-end 1998 was 13.2 percent of GDP, more than double the 1997 total with a flow of new arrears equal to 8.5 percent of GDP (IMF 1999/42). In Moldova, tax arrears equaled 10.9 percent of GDP in 1995, 13.7 percent in 1996, 11.2 percent in 1997, and 14.7 percent in 1998 (IMF 1998/58 and 1999/110). Tax arrears in Armenia were equal to 2 percent of GDP in 1996 and rose considerably through September of 1997 to between 2.3 and 2.6 percent (IMF 1998/22). Tax arrears in Belarus are a bit clouded due to a shift to reporting deferred taxes; however, they appeared to run at 2.2 percent of GDP for both 1996 (3.0 percent with deferred taxes) and 1997, with a substantial 42 percent increase in the stock of deferred taxes during the first quarter of 1998 (IMF1998/108). These tax arrears have been reinforced by a collapse of traditional institutional and administrative capacity to collect taxes and a lag in restructuring tax administration to match a restructuring economic system (Deppler and Odling-Smee 1997).
As a result of these tax arrears, government workers do not receive their wages, pensions are not paid, and suppliers do not get paid for the services or goods they have delivered. That means government suppliers do not have money to pay their taxes, and so it goes. Cash shortages recur, not because the money stock is too small (some have argued that the solution is to put more money into the economy, a standard Soviet-era response) but because too many enterprises are value-subtracting, rather than value-adding. The enterprise simply does not receive enough revenue--there is an inadequate market for what it produces--to cover its costs. Without effective bankruptcy processes, zombie-enterprises haunt the landscape, and arrears and noncash settlement of fiscal obligations are the results (see panel 1).
Panel 1 Orchestrated Non-Monetary Budgetary Transactions Reinforce Inefficiency in Both "Public" and "Private" Sector Activities Governments provide a budgetary accommodation to inefficient and nonprofitable ("zombie") enterprises by doing business with them in barter and budget offsets, thereby allowing debts to be settled in concessionary nonmonetary transactions. The offset transactions needed to clear accounts may involve the government and a number of different enterprises, each trading amounts owed among each other. In Russia, local and regional "balancing committees" have been established comprising representatives from the tax inspectorate, the finance and economics committees, the statistical committee, the oblast administration, and major local enterprises that meet weekly to try to work out clearing exchanges. The transaction costs are substantial in time and record keeping. Similar committees have been established in Kazakhstan, Turkmenistan, and other FSU countries. These transactions are then accounted for as if cash had changed hands, although the values can be at any level that seems convenient. On the revenue side of the budget, taxes are sometimes collected in kind. For instance, in the Kyrgyz Republic, regional governments established warehouses to hold and distribute taxes collected in agricultural products, mostly grain for use as in-kind payment of pension benefits, and subnational governments in Russia have collected cars instead of cash from automobile factories. Taxes on utility enterprises have been forgiven in exchange for nonpayment of utility bills by government. Agricultural products have also been used to "pay" salaries of teachers and health workers, as well as enterprise employees. Government employees must then engage in a secondary occupation as marketer of bartered products to convert these goods to items needed for living. This has created enormous inefficiencies in the operation of governmental agencies. Government services are distorted so that they are delivered where there are barter opportunities, not where the priority is highest. These exchanges also make data on revenue and expenditure suspect. Forecasts of revenue will seem extremely accurate, as trade goods can be assigned whatever value the officials wish.
This situation is further reinforced by an elaborate system of in-kind tax payment and offsets. Because enterprises lack cash, both enterprises and governments have established systems for receiving payments (including taxes) in-kind. The result is that employees and pensioners often receive wages and social support in the form of a mix of manufactured goods or agricultural products acquired in the offset process.(18) This process has evolved from an exchange system in which the government offsets an enterprise's tax liability based on the outstanding accounts receivable from the government to a system where enterprises negotiate with the government to accept their product (for which government may have little use) in lieu of cash tax payments (IMF 1998/58 and IMF 1999/100).
Offsets and barter have become preferred means for tax payment by the state agriculture sector in Moldova. The ministry of finance has directed (commanded) participation in complex offset transactions between accounts of apple farmers, canning enterprises, and natural gas, electricity, and transport enterprises.(19) The ministry has even used processed apples to pay for its own liabilities (IMF 1998/58). In Moldova, noncash in-kind and offset payments to the budget and social fund reached 37 percent and 50 percent of total resources during the second half of 1997 and have hovered at approximately 35 percent of the consolidated budget from 1996 through 1998 (IMF 1999/110). While the ministry of finance vowed to cease accepting offsets, the social fund remains an active participant in the process. In Ukraine, noncash revenues (from offsets) constituted 30 percent of total revenue in 1997 (up from 20 percent in 1996). For 1998, they ended the year at 28 percent (IMF 1999/42).(20) In Armenia, offsets amounted to 16 percent of tax revenue in 1997, down from 21 percent in 1996 (IMF 1998/22).(21) At the federal level in Russia, these tax offsets constituted 15 percent of noninterest expenditures and are estimated to average 50 percent of tax revenue at local levels (IMF 1999/100). Not only are these offsets provided for goods for which the government has no direct use, there is also a tendency to overvalue products to assist specific industries. Estimates suggest that in Russia offsets purchase only 50-75 percent of their face value, resulting in a real fiscal loss of 2.5-5 percent of GDP, or 8-15 percent of total government expenditures. As a result, government payment arrears (budget arrears) proliferate.(22)
Some substantial governmental budgetary arrears result from spending units obligating more money than their appropriation. The more frequent case in the FSU occurs when spending units do not receive budgeted funds, usually because there has been a revenue shortfall without an accommodating adjustment of spending authority.(23) That makes reported expenditures only partially accurate as a measure of resources consumed by the government, just as it leaves workers, suppliers, and pensioners in a difficult financial condition. The problem for workers and pensioners is especially acute because, as the free market expands into consumer goods and services, they need money to survive. Prices are no longer controlled, fewer goods are provided at subsidized prices, and more services (for example, utilities) must be purchased instead of being distributed free of charge, so waiting for payment becomes increasingly difficult. The arrears, which logically are forced, noninterest-bearing loans to the government, create major difficulties throughout the economies of the FSU countries. They accommodate government operations, but in the process cause budget execution to break down.
Table 6 presents central government budget arrears across selected FSU countries.(24) Clearing arrears has been a continual process. However, once a stock is cleared they tend to be replaced by new debts.(25) Kazakhstan's central government cleared wage and utility arrears by year-end 1996, however, local arrears for wages continued at approximately 1 percent of GDP. Nonwage local arrears equaled 1.5 percent and enterprise arrears equaled 6.4 percent of GDP in 1997 (Kaser 1998). Pension arrears were cleared in 1997 at an amount equal to approximately 2 percent of GDP (IMF 1998/84, 17). By year-end 1997, political pressure in Russia resulted in an effort to virtually clear government wage and pension arrears; however, arrears for regional nonwage spending was approximately 1 percent of GDP. By 1998, the combined total of arrears at the local level reached approximately 3.3 percent of GDP, while those accruing during the third quarter of 1998 at the national level reached 3 1/4 percent of GDP, with additional pension arrears of 1 percent of GDP (IMF 1999/100).
Table 6 Consolidated Government Expenditure Arrears, Selected FSU Countries (percent of GDP) 1996 1994 1995 Country total total Total Wages Armenia(1) +6.3 -2.2 -0.7 Azerbaijan(2) 2.6 0.5 Belarus 2.1 2.3 Georgia 0.4 0.6 Kazakhstan 5.4 0.7 Kyrgyz Republic 2.9 1.9 0.8 Latvia 0.0 0.0 Moldova 8.3 7.9 10.9 1.3 Russia 3.0 0.5 Tajikistan 10.3 5.0 Turkmenistan 0.0 0.6 Ukraine 1.2 4.6 1.5 Uzbekistan 0.8 3.6 1996 Social Sup- 1997 1998 Country sector pliers total total Armenia(1) +1.4 Azerbaijan(2) Belarus 0.4 Georgia 1.0 Kazakhstan 1.8 2.9 3.9 Kyrgyz Republic 0.9 1.2 Latvia Moldova 4.6 5.1 6.7 11.0 Russia 0.5 2.0 3.3 3.2 Tajikistan Turkmenistan Ukraine 2.2 0.9 2.8 2.6 Uzbekistan (1) Figures reflect net flows rather than levels. (2) Includes only social protection fund. Arrears for government operations are expected to be substantial. Sources: Ramos 1998, tables 1 and 2; and IMF Country Reports, all countries, various years.
Ukraine has also made some progress in clearing arrears; however, pension and social benefit arrears were estimated at 2.6 percent of GDP in 1998, down from 2.8 in 1997 and 3.7 in 1996 (IMF 1999/42).(26) General budget arrears there are most heavily concentrated in education, health, and housing, with the majority accruing at the regional and local levels. Additions to general budget arrears in Moldova have been dramatic, with the flow equal to 2 percent of GDP for 1994, 1.9 percent in 1995, 4.1 percent in 1996, followed by a 2.8 percent decline in 1997 (IMF 1998/58). The stock of general budget arrears in Moldova was 16.1 percent of GDP in 1993, 10.9 percent in 1996, 6.7 percent in 1997, and 11 percent in 1998 (IMF 1999/ 110). The Moldovan social fund continues to have substantial expenditure arrears, increasing by 2 percent of GDP in 1997. While information on the stock of arrears in Armenia is not available, additions to general budget arrears equaled approximately 1.4 percent of GDP through the third quarter of 1997, up from two consecutive years of net reductions (IMF 1998/22).
In other FSU nations, official arrears have been considerably limited. For example, wage and pension arrears in Georgia equaled. 18 percent of GDP at the start of 1998 (with additions to total general government arrears of.35 percent during 1997) (IMF 1998/99). The Kyrgyz Republic had made steady progress in reducing the stock of wage, transfer, and pension arrears from 2.9 percent of GDP in 1994 to .9 percent by year-end 1997, however, significant backtracking occurred in late 1998 (IMF 1999/31).(27)
Budgetary controls apply to the economic articles in budgets and bind spending units to predetermined expenditure mixes. Because the allocations to ministries often occur within the ministry of finance, not in the budget approved by the parliament, the controls are both nontransparent and overly restrictive. Excessive control of spending across line-items (for example, economic article) during execution unnecessarily binds spending agencies and ministries to fixed production processes both in budget preparation and in execution. Units have minimal, if any, capacity to move funds among economic articles and often are given explicit directions as to how spending is to be reduced if there are revenue shortfalls. The capacity of agencies to manage expenditure execution to deliver services in an efficient fashion is therefore limited. Use of extrabudgetary funds mitigates the problem by giving agencies greater operating flexibility, but reduces political and financial control over the agency.
More flexibility is needed in establishing the allocation across line-items (including staffing and salaries). The countries need mechanisms that allow spending unit discretion in the allocation of resources during budget execution, possibly with formal provisions to allow the reprogramming of a set amount of funds between economic article and expenditure programs without prior approval. Spending unit administrators must, however, be held personally accountable for the appropriateness of such reprogramming and for all funds they control.
In the early days of the transition, the FSU governments lacked a central treasury system. Spending agencies had separate accounts divided among the many parts of the nation in which they operated and revenues accumulated in disaggregated local accounts. A ministry could have money in an account in one part of the country, payments to be made somewhere else in the country, and no easy way to move the money from its resting place to where it was needed. There was no national system to move funds around from the places in which revenues accrued to the places at which bills had to be paid. The governments lacked a single, comprehensive national treasury account that afforded a unified and secure system for receiving government revenues, for mobilizing revenue for government use anywhere in the nation, for moving revenue from the jurisdiction of collection to the jurisdiction legally entitled to those funds, for orderly payments of governments and transfers, for managing and controlling debt issued by the government and for insuring timely servicing of that debt, and for accurate and timely reporting of financial transactions of the government.
Much effort, usually with IMF assistance, has been expended in the development and restructuring of treasury systems in the FSU. These reforms have tended to center on creating a comprehensive, centralized account system to counter corruption and control cash disbursements, and have been largely successful in this regard; however, progress across the FSU has been somewhat slow.(28) They have been less oriented to supporting effective budget execution. The treasuries have provided a strong constraint against measured government deficits by routinely rejecting legitimate requests for payment consistent with approved budgets, when treasuries lacked cash to make payment. This has resulted in additional forced loans to the government in the form of arrears. While this action allows for aggregate constraint of cash deficits, it diminishes the importance of the budget. It occurs through a failure to properly manage the budget at earlier points in the commitment stream and reflects a fundamental failure in the budget process. While obligations do not appear as cash, they reflect the usage of scarce resources.(29) Difficult choices must be made and enforced. The use of cash controls and automatic sequestration should only occur as a last resort and their prevalence is evidence of the failure of budget processes and procedures (Deppler and Odling-Smee 1997).
Recent reforms in the Kyrgyz Republic, Kazakhstan, and Azerbaijan have begun to address budget management issues. A treasury expenditure warrant system has been (or is being) developed to provide a monthly apportionment of funds and to hold spending units accountable for obligating only to the amount available. The 1998 Kyrgyz budget law also established a cabinet-level budget commission to monitor budget execution (IMF 1999/31). In addition, the 1999 budget law in Russia states that contracts signed by spending units that exceed budgetary limits will not be considered legal obligations of the government. This is intended to reduce incentives of suppliers to provide goods and services to spending units in anticipation of future offsets. Russia has also attempted to address the issues of overcommitment by forcing ministries to develop formal plans for staying within their spending limits (IMF 1999/100).
Many FSU countries have not worked out the distribution of authority between the government and parliament to adjust the budget during execution.(30) An October 1998 amendment to the Kyrgyz constitution prohibits parliament from introducing changes in the budget midyear unless the government agrees (IMF 1999/31). In Latvia, the minister of finance may propose midyear spending reductions if updated revenue projections imply a deficit greater than that provided in the budget law. Spending reduction beyond this amount must be adopted by parliament (IMF 1999/77). In Ukraine, because of parliamentary opposition to necessary budget cuts in 1998, the president reduced spending limits by decree and the ministry of finance was instructed to introduce monthly spending controls and develop a list of priority cash expenditures approved on a daily basis (IMF 1999/42).
Audit and Verification
The budget process requires reporting and auditing as a final stage, but both remain underdeveloped in the FSU countries.(31) Such requirements are new to the FSU. Audit was not a concern in Soviet times because resources of state enterprises and institutions were in controlled accounts in the state bank (Gosbank). Entity accounts were segregated by type of usage (such as, wages fund, materials fund) and funds were not convertible across use. Payment occurred on presentation of a transfer order, but only after bank verification that such use of the account was authorized. With that control in place, post-expenditure audit was seen as irrelevant. In the Soviet system, reporting of audit findings to the public would have been preposterous.
Reporting and audit are critical to maintaining focus on the intended outcomes of the other components of the budgetary system (Campos and Pradhan 1997). The concept of an audit is not well-developed, institutions for independent audit and their reporting lines are not clearly established, and financial reports are often not prepared on a timely and adequate basis. Further, audit is no substitute for proper management controls, and a strengthening of management control systems must occur simultaneously with establishing capacity for independent verification of financial stewardship. To be sure, one of the primary objectives of verification is to stimulate management control improvements in the form of improved physical controls, accounting controls, process controls, procurement controls, and internal audit capacity (Schiavo-Campo and Tommasi 1999, ch. 9).
Several FSU countries have established or are in the process of establishing (for example, Russia, Kyrgyz Republic, Azerbaijan, and Kazakhstan) a supreme audit institution responsible for auditing executive branch accounts on behalf of the parliament. FSU countries have generally had internal audit units housed in each ministry and reporting to the ministry of finance; however, they have generally been seriously understaffed and have not been capable of performing regular, systematic and comprehensive audits of spending unit activity. Progress has been slow. In the past, all FSU countries have provided execution reports; however, a reliable means of verifying execution data has been lacking without an independent audit institution--and the capabilities of independent audit are seriously compromised due to the limitations of internal audit and management controls. The role of various institutions of government in monitoring the execution of the state budget often needs to be specified, as does the balance between the roles of parliament and executive bodies.
Completion of the budget process requires that the government report its financial operations. The general rule is for the ministry of finance to prepare a summary annual budget execution report several months after year-end. However, these reports do not clearly classify spending by ministry or by economic category, and are generally not verified for accuracy by any independent audit body. They are not yet consistent with important principles of completeness, reliability, relevance, consistency, or timeliness (Schiavo-Campo and Tommasi 1999, ch. 11). Consolidation of budget and extrabudgetary funds in reports has been progressing but is far from complete, the information and presentation contained is limited, and reports do not consistently include the consolidated accounts of subnational units. Reliability is also questionable due to the lack of external verification. Relevance is often lacking due to the summary nature of reports and their failure to be produced in a timely manner. Ideally, budget execution reports should be continual, including monthly appropriation, variance, apportionment, commitment, payment, and arrears accounting, and at least preliminary year-end execution reporting should occur within two months of the close of the budget year. In many cases, the final components of budget preparation are being completed before preliminary execution reports have been made available. More than a small portion of reporting problems stem from a lag in implementing accounting structures consistent with basic accountability and transparency needs.
Financial reporting is important for control and evaluation and most FSU nations are improving their practices. Complete annual reporting statements need to be developed and, to make these reports meaningful and persuasive, the final statement must be audited by an institution independent of the government. FSU countries are far from this standard.
Independent Supreme Audit Institution
In several FSU countries, progress has been made in establishing an independent audit function, but the autonomy and independence of this institution from the executive is less than fully assured. Effective auditing is necessary to identify weaknesses in management controls and detect irregularities in the use of funds that may threaten effective budget implementation, to establish the reliability of reports and financial data, and to identify waste (Schiavo-Campo and Tommasi 1999, ch. 9). The primary vehicle for this is ex post auditing (in the form of both compliance and financial audits), with public reporting of findings and annual review and reporting of the status of all government accounts. For this to occur, an audit institution fully independent of the executive is necessary. This must include budgetary independence, independence in determining the scope of work, access to all necessary data, and a staff of appropriate size and training as essential elements to assure unbiased findings. The development of the full capacity of an audit system requires time and is dependent upon the simultaneous development of requisite input systems in the form of management controls and accounting. Only then is it possible for audit to provide reasonable assurance of the veracity of accounting and reporting statements, to offer effective final oversight of budget expenditure accounts to prevent inappropriate and fraudulent usage of funds, and to assure that the budget is executed consistent with legislative intentions as specified.
An audit institution is vital for holding the executive accountable to the legislature and the public. As a supreme audit entity, this institution should report directly to parliament and the membership of its executive body should be appointed by the parliament for fixed terms with protected tenure to assure its independence. A number of FSU countries have lagged in establishing satisfactory audit procedures, failed to provide sufficient and competent staffing, and have not clarified the executive/ legislative reporting relationship and have supported reporting to the executive or jointly to legislative and executive.(32) The adequacy of funding is questionable, as is access to information and documents, and the principle that findings be publicized and disseminated is not fully established. These shortcomings seriously compromise the independence and capacity of supreme external audit institutions, jeopardize their ability to perform the necessary surveillance role, and limit their ability to stimulate reform in critical supporting systems.(33)
Conclusions and Directions for Reform
While reforms have improved the budget processes, procedures, and institutions in former Soviet republics, much remains to be done to transform fiscal systems, initially designed for command and control economic systems, to those capable of responding to the requirements of the dynamic environment of reformed economic and political systems. Reforms underway across the FSU are addressing (or have addressed) a variety of needs in these areas. However, this is a long-term process. The procedures and institutions required and the social and political mores needed to reinforce them cannot be developed overnight. Notwithstanding substantial improvements, budgets in FSU countries are typically not comprehensive; ministries are provided few alternatives in responding to public service demands; priorities are not adequately articulated within hard and realistic resource constraints; budget development fails to provide a transparent articulation of government intentions and there is rarely an external verification of how budgeted resources have actually been spent.
Budget reform must focus on the interrelated dimensions that make up a public budgeting system. Figure 1 displays elements that require specific and deliberate attention in the FSU. These include comprehensiveness, institutional structures for linking substantive policy and priority setting to budget development, expenditure planning and forecasting, mechanisms for assuring managerial efficiency, and subsystems to assure accountability and control. The status of FSU budget systems regarding these elements (as described above) is underdeveloped.
Figure 1 Dimensions of Budget System Reform [ILLUSTRATION OMITTED] Transparency, requires clear roles and responsibilities, comprehensive and reliable fiscal information, open budget preparation, execution and reporting and independent assurances of integrity. Transparency cannot be meaningfully achieved without supporting budget reforms and institutional capacity. Comprehensiveness requires all revenues and expenditures be reflected in budgetary submissions and revenue and expenditure accounts and the full accounting, if not elimination, of all "extra" budgetary transactions, including all general and dedicated revenues, all special accounts, all charge revenue, all fund transfers and subsidies, all operating expenditures, all capital expenditures, all loans, grants and contingent liabilities, and the use of dedicated and extrabudgetary funds. Policy setting and prioritizing requires comprehensiveness and includes a medium-term financial framework, expenditure estimation tied to service output and outcomes, budget classification mapping expenditures to purposes and organizational units, mechanisms requiring spending units to budget within the availability of resources, and execution controls to assure adherence to priorities. Expenditure planning also requires enforced spending envelopes based on priorities, budget classification mapping resources to service outputs, expenditure estimation relating spending to service levels, and flexibility in the mix of factor inputs. Managerial efficiency requires expenditure estimation based priorities and service objectives, flexibility in service-delivery vehicles, accountability for service levels and outputs, and discretion in the relative use of inputs (supplies, labor, capital, and quantities) and in the management of resources. Accountability and control requires comprehensiveness and priority setting, budget classification, and accounting which maps resources to spending units, systematic budget and expenditure reviews, execution controls assuring pre-expenditure control and post-execution accountability, and comprehensive, audited annual reports.
The double-headed arrows in figure 1 reflect the fact that each of these components reinforces the others and emphasizes the need to pursue budget reform from a systematic perspective. For example, it is not sufficient to embark on treasury reform without developing medium-term forecasting capacity or without a redefinition of agency expenditure responsibility. Individual components rarely stand alone. Sub-system reforms must always be considered within the context of related elements and institutions. The implications of this are significant regarding the level of ambition required for meaningful adjustments.
Specific reform emphasis in the FSU needs to be placed on: (1) development of a fully comprehensive government budget and severing nongovernmental functions from governmental organizations; (2) the policy and prioritizing roles of the budget with less attention to physical plans; (3) expenditure planning, short- and medium-term expenditure forecasting, and expenditure ceilings for spending unit budget development with all analysis done on the basis of realistic forecasts of finances instead of normative plans; (4) appropriate budget classification along with a focus on services rather than input norms and flexible deployment of resources to meet population service needs; and (5) clear accountability for budget execution with external verification of the budget program.
Table 5 Tax Arrears and In-Kind (and Netting) Payment Offsets in States of the Former Soviet Union(1) Country Stock of tax arrears as percent of GDP 1995 1996 1997 1998 Armenia(2) 2.0 2.3/2.6 Azerbaijan Belarus(3) 2.1 2.2/3.0 2.2 Estonia Georgia 1.1 2.1 2.5 Kazakhstan(4) 1.4 2.7 2.1 Kyrgyz Republic 0.9 2.3 1.4 Latvia Lithuania Moldova(5) 10.9 13.7 11.2 14.7 Russia(6) +3.9 +6.5 +4.9 +4.9 Tajikistan Turkmenistan 6.2 Ukraine 21.0 3.6 5.3 13.2 Uzbekistan 2.0 2.2 Country Annual tax offsets as percent of total revenue 1995 1996 1997 1998 Armenia(2) 17.0 21.0 16.5 Azerbaijan Belarus(3) Estonia Georgia Kazakhstan(4) 5.9 Kyrgyz Republic 21.0 30.1 21.8 24.1 Latvia Lithuania Moldova(5) 35.2 36.8 35.8 Russia(6) 14.8 26.1 18.8 15.6 Tajikistan Turkmenistan 30.0 51.0 Ukraine 20.0 30.0 28.0 Uzbekistan (1) Data for both the stock of arrears and offsets are not systematically or comprehensively reported. The absence of reported figures in no way reflects the absence of arrears and offsets, and the numbers actually reported likely reflect only partial accounting of actual magnitudes. (2) Offsets are as reported in text of IMF 1998/22 as a percent of tax revenues. However, these figures are inconsistent with supporting tables, calculation from which show 1995 offsets of 8.8 percent of consolidated total general revenue, 14.5 percent of consolidated tax revenue, 13.0 percent of total state revenue, and 17.2 percent of state tax revenue; 1996 calculated figures are 16.7 percent, 20.9 percent, 19.8 percent and 25.3 percent; and 1997 calculations are 29.1 percent, 31.7 percent, 36.1 percent and 38.6 percent. (3) The 3.0 percent figure for 1996 arrears includes "deferred taxes." (4) Arrears for 1998 reflect only those reported for pensions. (5) In-kind collections occur most heavily in the latter half of the year and often are greatest for the social fund. Second half in-kind and offset collections for 1997 were 37 percent for the consolidated budget and 50 percent for the social fund, 1996 second half figures were 41 percent and 30 percent. (6) Tax arrears figures are presented as changes as a percent of GDP. Beginning stock levels were not reported; however, calculations of stock levels derived from the text of IMF 1999/100 and the included graphs result in varied estimates. Offsets reflect only those of the general central government (including ruble offsets), local government tax offsets are as much as 50 percent of tax revenue. IMF text statements of offsets imply significantly greater levels than reported in tables. Sources: IMF Country Reports, all countries, various years.
The authors would like to thank Helga Muller (sector manager--Public Finance, Europe and Central Asia Region, Division of Poverty Reduction and Economic Management) for comments on an earlier version of this manuscript. Errors and misstatements remaining in the article are the sole responsibility of the authors.
(1.) The states of the former Soviet Union are Estonia, Latvia, Lithuania, Moldova, Ukraine, Tajikistan, Turkmenistan, Uzbekistan, the Kyrgyz Republic, Kazakhstan, Armenia, Azerbaijan, Georgia, Ukraine, and the Russian Federation. They proclaimed independence on various dates in the last half of 1991, with the end of the Soviet Union formally declared on December 25, 1991. The three Baltic states became part of the Soviet Union much more recently than the rest. Because they have different histories and represent a special case, they are excluded from much of the discussion here.
(2.) For an accounting of budgetary difficulties experienced in transition at the local level in Hungary see Straussman and Fabian (1994) and Thurmaier (1994) for the experience of local government in Poland.
(3.) For a recent critique of policies advanced in transition as a result of the Washington Consensus and the elaboration of a new set of policies, which highlight the need for conscious and concerted institutional reform (including strengthened budget systems and prioritized expenditure restructuring), see Kolodko (1999).
(4.) In many ways, the FSU situation typifies the stylized description of a public sector institutional environment found in developing countries (Shah 1997).
(5.) Recently there has been some rebound in expenditures as a percent of GDP. For example, in Belarus, expenditures of the general government increased 4.2 percent between 1996 and 1997 and were budgeted to increase an additional 1.5 percent in 1998 (IMF 1998/108).
(6.) Huge price distortions and economic change make comparisons between current expenditures and early-transition expenditures hazardous. However, the magnitude of these changes is still telling, even with cautious interpretation.
(7.) It has also been suggested in some countries, most notably Russia and Ukraine, that further spending reductions are in order and that attempts in transition economies to move to a public spending complement similar to the Organization for Economic Cooperation and Development nations are premature and unsustainable (Deppler and Odling-Smee 1997). However, at the same time that economic capacity is lacking, demographics indicate an aging and growing dependent population that will strain the capacity for social spending (de Castello Branco 1998). Others have argued that much of the strain on social spending in transitional economies results from loose eligibility criteria (Fakin and de Crombrugghe 1997).
(8.) For example, budget procedures and structures in the Republic of Azerbaijan are governed by the budget systems law (no. 391) adopted December 1, 1992. This law was in the process of substantial overhaul during the spring and summer of 1998. The Turkmen law was last revised in June of 1996. The Kazak budget system law dates from December of 1996 and has also recently been in the process of significant revision, with extensive redrafting in 1998. This is part of a comprehensive government budget system reform program established by resolution (no. 1533) of the government, December 13, 1996. The Kyrgyz have adopted the most comprehensive of budget reforms in the Central Asian region to date. Their existing budget procedures law is by far the most comprehensive and was adopted during November of 1997 as the result of an extensive government reform program begun in 1996. Restructuring and reform has been slower to develop in Uzbekistan. The budget principles law for Russia was adopted in 1991 (Law of the Russian Federation No. 1735-1 of October 10, 1991 on the Principles of the Budget Organization and Budget Process of the Russian Federation) with amendments in 1994 and 1995. A major revision is in process because the system has been overtaken by the development of a federal state and changes in the relationships between governmental bodies at the federation level. Substantial revisions are also currently underway in Lithuania, seeking increased inclusion of extra budgetary funds, limits on line ministries' year-end spending and carryover provisions, and a three-year rolling budget framework (IMF 1999/73).
(9.) Constitution of the Republic of Azerbaijan, Article 95, no. 5.
(10.) Calculations are the authors', based on data from IMF 1999/ 100.
(11.) Earmarking can introduce market-like forces into determination of public expenditures or preserve an earnings-related social insurance program. More often, however, the schemes simply complicate fiscal management and control and reduce the ability of government to move resources toward greatest public priority (Bird 1997).
(12.) This only accounts for fee systems officially adopted by institutions. Ministerial officials acknowledge substantial additional under-the-table payments for services.
(13.) Recent research finds no significant relationship between public capital expenditures in the FSU and economic growth during transition, possibly because of gross misallocation of the spending or mismeasurement of difficult data (Taube and Zettelmeyer 1998; Zettelmeyer 1998).
(14.) Revenue forecasts in FSU countries have been notoriously overly optimistic, as the next section describes. But even these control numbers have not been provided to the ministries. There has often been no constraint in budget development.
(15.) Exceptions are in the Kyrgyz Republic and Kazakhstan, where allocations down to the ministerial level appear in the budget as adopted by parliament. This occurred for the first time during construction of 1998 budgets.
(16.) Azerbaijan, Kazakhstan, and the Kyrgyz Republic currently adjust budget execution by controlling the level of monthly expenditure warrants issued to spending units. These current practices appear orderly and have generally proven effective in allocating resources across the budget year. Provisions for more substantial adjustment have not been formalized in Kazakhstan or Azerbaijan.
(17.) See specifically IMF 1999/100, Annex II. Estimates of 1997 wage arrears in regional capitals along the Volga River average 30 percent of the monthly wage bills with a high of 50 percent in Kazan (de Melo and Ofer 1999).
(18.) In-kind pension payments in Moldova reached 34.7 percent of total payments in 1997 and 47.8 percent in 1998 (IMF 1999/110). Tax offsets in Ukraine included 400,000 tons of grain as payment to the pension fund.
(19.) These offsets have largely provided a form of forced subsidy to the apple farmers.
(20.) The 1998 state budget in Ukraine officially allowed offsets only for specific rural projects and defense, a policy ignored during budget execution. In fact, past tax amnesties for the agricultural sector and frequent tax restructuring may have encouraged tax arrears (IMF 1999/42).
(21.) There is some discrepancy between IMF report text and tables. Authors' calculations based on tax collection figures contained in tables suggest that offsets equaled 32 percent of consolidated tax revenues for the third quarter of 1997. An official "tax arrears payment scheme" introduced in 1997 was seen as failing to stem the tide of arrears as few firms elected to participate in the program.
(22.) In June 1998, a presidential decree in the Kyrgyz Republic was issued prohibiting the writing off of tax arrears, reducing the stock of tax arrears from 2.3 to 1.4 percent of GDP between year-end 1997 and end-September 1998 (IMF 1999/ 31). However, as a result of natural disasters and the Russian crisis, Kyrgyz tax offsets rose from 21.8 percent of tax collections in 1997 to 24.1 through the third quarter of 1998. Offsets for 1996 equaled 30 percent of collections.
(23.) The FSU countries lack access to short-term capital markets, either external or domestic, to cover shortfalls in the way that western countries would and printing money--the solution in Soviet times--has had terrible effects on prices.
(24.) The problem of arrears has progressively been shifted down as central governments have delegated many spending responsibilities to subnational governments.
(25.) Arrears also produce opportunities for corruption. Unpaid suppliers may attempt to influence the order in which payments are cleared by bribing line ministries or treasury officials to move up the list (Schiavo-Campo and Tommasi 1999, ch. 7).
(26.) Arrears for 1996 equaled 4.6 percent of GDP when those to suppliers are included (Ramos 1998).
(27.) One method of dealing with budgetary arrears is "securitization," for example, the issuance of a marketable (possibly interest beating) form of I.O.U. to creditors (Ramos 1998).
(28.) While this success has been substantial, significant problems remain. For example, in Russia military expenditures (representing one quarter of the 1999 budget) do not flow through the treasury system and in Ukraine the pension fund and three security agencies are excluded (IMF 1999/100 and IMF 1999/42). The first regional treasury began operation in Moldova in January of 1998 and was expected to include all 10 regions by April of that year (IMF 1998/58).
(29.) Recent budget execution presentations attempt to incorporate some commitments as expenditures in execution reporting in Russia (IMF 1999/100) and Ukraine (IMF 1999/42).
(30.) Schiavo-Campo and Tommasi (1999, ch. 7) discuss possible parameters for establishing relative levels of responsibility for mid-year budget adjustments. A recent Moscow Times article (November 11, 1999) advances the existence of a "secret budget" to describe the budget actually executed by the government compared to that passed by parliament. A district court has ruled against the Finance Ministry's application of mid-year adjustments.
(31.) With characteristic Russian forthrightness, a hint of the less-than-total devotion to transparency and audit among some senior Russian officials can be found in a Moscow Times (September 29, 1999) quote from Viktor Geraschenko (head of the Central Bank of Russia) concerning the external audit of bank operations: "Had we not conducted an audit, nobody would know anything at all."
(32.) Additional confusion has been engendered by audit chambers sometimes focusing on questions of legislative choices, rather than establishing adherence to legislative intent.
(33.) Another source of confusion, directly emerging from Soviet times, is the idea that this audit institution should also be the auditor for commercial enterprises and for tax compliance. These audit roles are clearly different, reporting responsibilities are different, and the audit assignment is different. However, in the Soviet era, all these activities were merged through state ownership of enterprises and through the importance of delivering the state plan.
Bird, Richard M. 1997. Analysis of Earmarked Taxes. Tax Notes International 97 (June 23): 120-39.
Campos, Jose Edgardo, and Sanjay Pradhan. 1996. The Impact of Budgetary Institutions on Expenditure Outcomes: Binding Governments to Final Performance. Working paper. The World Bank, Policy Research Department, Washington, DC.
--. 1997. Evaluating Public Expenditure Management Systems: An Experimental Methodology with an Application to the Australia and New Zealand Reforms. Journal of Policy Analysis and Management 16(3): 423-45.
de Castello Branco, Marta. 1998. Pension Reform in the Baltics, Russia and Other Countries of the Former Soviet Union (BRO). Working paper 98/11 and 98/133. International Monetary Fund, Washington, DC.
de Melo, Martha, and Gur Ofer. 1999. The Russian City in Transition: The First Six Years in Ten Volga Capitals. Working paper. The World Bank, Development Research Group, Washington, DC.
Deppler, Michael C., and John Odling-Smee. 1997. Aspects of Fiscal Performance in Some Transition Economies Under Fund-Supported Programs. Working paper 97/31. International Monetary Fund, Washington, DC.
Fakin, Barbara, and Alain de Crombrugghe. 1997. Fiscal Adjustments in Transition Economies: Social Transfers and Efficiency of Public Spending. Sponsored Research Report, The World Bank, Washington, DC.
Freinkman, Lev M., and Irina Starodubrovskaya. 1995. Restructuring of Enterprise Social Assets in Russia: Trends, Problems, Possible Solutions. Working paper, IIASA Conference, Vienna, The World Bank, Washington, DC.
Hinton, Hugh. 1998. Transforming Budgeting from Central Planning to a Market Economy: Dilemmas of Reform in Ukraine. Paper presented at the annual conference of the Association for Budgeting and Financial Management, November 5-7, Washington, DC.
Holland, Stephen, and Tim Ensor. 1996. Sources of Financing for the Health Sector of Turkmenistan. Ashgabat, Turkmenistan: Lukman Health Programme of Turkmenistan.
International Monetary Fund (IMF). 1997/109. Ukraine--Recent Economic Developments. Country Report 97/109, October. Washington, DC: IMF.
--. 1997/111. Republic of Belarus: Recent Economic Developments. Country Report 97/111. Washington, DC: IMF.
--. 1998. World Economic Outlook, ch. II and V. Washington, DC: IMF.
--. 1998/8. Kyrgyz Republic: Recent Economic Developments. Country Report 98/8, January. Washington, DC: IMF.
--. 1998/12. Republic of Estonia: Selected Issues and Statistical Appendix. Country Report 98/12, February. Washington, DC: IMF.
--. 1998/16. Republic of Tajikistan: Recent Economic Developments. Country Report 98/16, February. Washington, DC: IMF.
--. 1998/22. Republic of Armenia: Recent Economic Developments. Country Report 98/22, April. Washington, DC: IMF.
--. 1998/47. Republic of Latvia: Selected Issues and Statistical Appendix. Country Report 98/47, May. Washington, DC: IMF.
--. 1998/58. Republic of Moldova: Recent Economic Developments. Country Report 98/58, June. Washington, DC: IMF.
--. 1998/81. Turkmenistan: Recent Economic Developments. Country Report 98/81, August. Washington, DC: IMF.
--. 1998/83. Azerbaijan Republic: Recent Economic Developments. Country Report 98/83, August. Washington, DC: IMF.
--. 1998/84. Republic of Kazakhstan: Recent Economic Developments. Country Report 98/84, August. Washington, DC: IMF.
--. 1998/99. Georgia: Recent Economic Developments and Selected Issues. Country Report 98/99, September. Washington, DC: IMF.
--. 1998/108. Republic of Belarus: Recent Economic Developments. Country Report 98/108, October. Washington, DC: IMF.
--. 1998/116. Republic of Uzbekistan: Recent Economic Developments. Country Report 98/116, October. Washington, DC: IMF.
--. 1999. World Economic Outlook. Washington, DC.
--. 1999/31. Kyrgyz Republic: Recent Economic Developments. Country Report 99/31, April. Washington, DC: IMF.
--. 1999/42. Republic of Ukraine: Recent Economic Developments. Country Report 99/42, May. Washington, DC: IMF.
--. 1999/59. Republic of Estonia: Staff Report for the 1999 Article IV Consultation. Country Report 99/59, July. Washington, DC: IMF.
--. 1999/73. Republic of Lithuania: Staff Report for the 1999 Article IV Consultation. Country Report 99/73, August. Washington, DC: IMF.
--. 1999/77. Republic of Latvia: Staff Report for the 1999 Article IV Consultation. Country Report 99/77, August. Washington, DC: IMF.
--. 1999/95. Republic of Kazakhstan: Selected Issues and Statistical Appendix. Country Report 99/95, September. Washington, DC: IMF.
--. 1999/96. Republic of Lithuania: Selected Issues and Statistical Appendix. Country Report 99/96, September. Washington, DC: IMF.
--. 1999/99. Latvia: Selected Issues and Statistical Appendix. Country Report 99/99, September. Washington, DC: IMF.
--. 1999/100. Russian Federation: Recent Economic Developments. Country Report 99/100, September. Washington, DC: IMF.
--. 1999/110. Republic of Moldova: Recent Economic Developments. Country Report 99/110, September. Washington, DC: IMF.
Kaser, Michael. 1998. Stabilization and Reform: Experience of Five Central Asian States. Working paper, Washington, DC: IMF.
Kolodko, Grzegorz W. 1999. Ten Years of Post Transition: The Lessons for Policy Reform. Working paper, Development Economics Group, Washington, DC.
Myers, Roy. 1996. Is There a Key to the Normative Budgeting Lock? Policy Sciences 29: 171-88.
Ramos, Alberto. 1998. Government Expenditure Arrears: Securitization and Other Solutions. Working paper 98/70, International Monetary Fund, Washington, DC.
Schiavo-Campo, Salvatore, and Daniel Tommasi. 1999. Managing Government Expenditure. Washington, DC: Asian Development Bank.
Shah, Anwar. 1997. Balance, Accountability, and Responsiveness: Lessons About Decentralization. Working paper, The World Bank, Washington, DC.
Straussman, Jeffrey, and Katalin Fabian. 1994. Local Government Finance in Hungary: Okund Revisited. Public Budgeting and Finance 14(4): 71-83.
Tanzi, Vito. 1994. Reforming Public Finances in Economies in Transition. International Tax and Public Finance 1(2): 157.
Taube, Gunther, and Jeromin Zettelmyer. 1998. Output Decline and Recovery in Uzbekistan: Past Performance and Future Prospects. Working paper 98/132, International Monetary Fund, Washington, DC.
Thurmaier, Kurt. 1994. The Evolution of Local Government Budgeting in Poland: From Accounting to Policy in a Leap and a Bound. Public Budgeting and Finance 14(4): 84-97.
Wilczynski, J. 1970. The Economics of Socialism. Chicago: Adline Publishing.
World Bank. 1996a. Fiscal Management in Russia. Washington, DC: World Bank.
---. 1996b. Republic of Kazakhstan: Transition of the State, vol. II. Country Operations Division I, Washington, DC: World Bank.
--. 1996c. Georgia--Public Expenditure Review. Country Operations Division II, Washington, DC: World Bank.
--. 1998. Public Expenditure Management Handbook. Washington, DC: World Bank.
--. 1999. Georgia--Education System Realignment. Report No. PID7940. Washington, DC: World Bank.
Zettelmeyer, Jeromin. 1998. The Uzbek Growth Puzzle. Working paper 98/133, September, International Monetary Fund, Washington, DC.
John L. Mikesell is a professor of public finance and policy analysis at Indiana University, specializing in tax policy and administration and government budgeting. He has served as chief fiscal economist with the Barents Group Ukraine fiscal reform project, resident director for intergovernmental finance with the Georgia State University Consortium Russian Federation fiscal reform project, and budget process specialist with World Bank missions to Kazakhstan, Kyrgyzstan, Azerbaijan, and Turkmenistan. He edits Public Budgeting and Finance. Email: email@example.com.
Daniel R. Mullins is a professor of public finance in the School of Public Affairs at American University, specializing in public budgeting and state and local finance. His research has appeared in a variety of academic and professional sources. He has been extensively involved in numerous World Bank and USAID budget and fiscal reform projects in FSU nations, including Russia, Uzbekistan, Kazakhstan, the Kyrgyz Republic, Azerbaijan, and Turkmenistan. He is the managing editor of Public Budgeting and Finance. Email: firstname.lastname@example.org.
|Printer friendly Cite/link Email Feedback|
|Author:||Mikesell, John L.; Mullins, Daniel R.|
|Publication:||Public Administration Review|
|Date:||Sep 1, 2001|
|Previous Article:||New Labour and the Public Sector in Britain(1).|
|Next Article:||Checks, Balances, and Appointments in the Public Service: Israeli Experience in Comparative Perspective.|