Initial Conditions

Initial conditions in the CIS-7 countries are less varied than in the transition economies as a whole, because all seven had the shared experience of being Soviet republics until December 1991. The Caucasus countries (Armenia, Azerbaijan, and Georgia) had a greater sense of national identity than the Central Asian countries (the Kyrgyz Republic, Tajikistan, and Uzbekistan), but none of the CIS-7 countries had a recent memory of statehood or of a market economy. By geography, all of the CIS-7 countries are disadvantaged relative to the Central and Eastern European transition economies, but it is difficult to assess the degree of disadvantage among them. Georgia has Black Sea ports, and all of the Caucasus countries have easier geographic access to Western Europe (as well as to Turkey and Iran) than do the Central Asian countries, which are geographically isolated and whose inherited transport and pipeline links all run north to Russia. Moldova is geographically further west, but it borders only Romania and Ukraine and is not on major east-west transport routes.

All Soviet republics’ economies were integrated into the Soviet system, with very little external trade. The Central Asian republics supplied raw materials, especially cotton. Azerbaijan’s economic importance for most of the twentieth century lay in its oil, but its resources were not exploited very actively in the final decades of the Soviet era. All the CIS-7 countries had industrial sectors that were part of the integrated Soviet economy, with poor prospects of prospering as self-standing industrial sectors. An immediate shock following the dissolution of the Soviet Union was the shift from artificial prices, with energy underpriced and manufactures overpriced, to world prices. Moldova suffered the greatest negative terms-of-trade shock, but Armenia and Georgia and to a lesser extent Azerbaijan and Tajikistan also suffered negative terms-of-trade effects.²

The most important, and most difficult to measure, initial condition was social and political stability. Azerbaijan, Georgia, and Moldova all inherited secessionist problems that led to conflict in the early 1990s, and the status of Abkhazia, Nagorno-Karabakh, and Transnistria remains unresolved. Tajikistan’s civil war was ended by a 1997 peace accord, but the central government still does not control the national territory, and policy implementation is compromised. The Kyrgyz Republic and Uzbekistan have had internal peace, but their governments have used external threats, notably that of Islamic fundamentalism, to justify limiting human rights and democratic processes.

Policies and Explanations for Growth Failures

Macroeconomic stabilization has been a significant explanatory variable in crosscountry studies of growth in transition economies. All the CIS-7 countries suffered from hyperinflation and from the disintegration of the ruble zone in 1992–93, and all have achieved a degree of stability despite further external shocks—especially the Russian financial crisis of 1998. Annual inflation was brought below 50 percent in 1995 in the Kyrgyz Republic and Moldova; in 1996 in Armenia, Azerbaijan, and Georgia; and in 1998 in Tajikistan and Uzbekistan. The three Caucasus countries have established credible price stabilization records over the past five years. The other countries’ inflation records have been uneven. Although none has slipped back above the 50 percent mark, Moldova and Tajikistan had inflation rates over 30 percent in 2000, and Uzbekistan’s inflation rate is repressed by foreign exchange and other price controls. Several CIS-7 countries have external debts that could threaten their macroeconomic stability.

The pattern of initial conditions and macroeconomic stabilization of the CIS-7 countries is consistent with the broader evidence from transition economies. Given the initial conditions and mixed macro stabilization records, growth records of the transition economies can be related to the pace of structural reform. Here, however, the picture is less clear. Why did the Kyrgyz Republic and Moldova, countries with good reform records, especially in the first half of the 1990s, perform much more poorly than Uzbekistan, with its more limited structural reforms? Indeed, why has Uzbekistan enjoyed the best output performance of all the new independent states since the dissolution of the Soviet Union?

One possible explanation is that structural reform has been far from complete. Reforms that might be part of an optimal across-the-board package may be welfare-reducing when introduced piecemeal. This explanation is unconvincing, however, because there is strong international evidence that structural reforms in each of the areas undertaken are desirable, and although there are likely to be losers as well as winners from the reforms, negative resource allocation consequences are uncommon.

A second explanation is that the areas of structural reform are not equally important. The view that Uzbekistan is a laggard rests on lack of reforms in currency convertibility, price liberalization (mainly the state orders), governance and enterprise restructuring, banking, telecommunications, and roads and water, but Uzbekistan rates highly on railways and air transport. More generally, infrastructure and public services seem to work well in Uzbekistan compared with the other CIS-7 countries.

By contrast, the Kyrgyz Republic, despite its strong reformist credentials for most of the 1990s, had terrible infrastructure. Bishkek airport was closed to international flights for most of the decade, and internal communications between the north and south of the country relied on a poor road. Geography does not help the Kyrgyz Republic, but without decent infrastructure it was difficult either to create a well-functioning integrated national economy or to benefit from outward-oriented growth. This view acknowledges the potential benefits of structural reform but makes success conditional on preconditions such as price stabilization, peace and order, and good infrastructure.

A third explanation, to some extent a generalization of the previous one, is that an effective state is the crucial precondition for successful transition. The CIS-7 countries include the five Soviet successor states that endured the most violent passage to nationhood. Georgia, Moldova, and Tajikistan went through civil wars, and even today some areas of the countries are not under central government control. Armenia and Azerbaijan fought a bitter war in 1992–94, which caused much suffering and economic disruption but did not undermine state legitimacy. The Kyrgyz Republic and Uzbekistan had peaceful transitions to nationhood, although the Kyrgyz Republic is riven by political and economic divisions between north and south.

Output performance, measured by real GDP in 2001 as a percentage of real GDP in 1989, follows a threefold division that reflects these histories much more closely than it matches the extent of structural reforms (see Table 4.1). Moldova, Georgia, and Tajikistan have the worst performance; Azerbaijan, Armenia, and the Kyrgyz Republic are intermediate; and Uzbekistan is the best performer. Failure to create an effective state is evident in the three poor performers.³ Uzbekistan’s central government, by contrast, has largely achieved its aim of being technocratically efficient. This last explanation of course begs the question of what makes for an effective state in the context of the CIS-7 countries, an issue discussed later in the chapter.

Liberalization of Internal Markets and International Trade

The Kyrgyz Republic (since 1998), Georgia (2000), and Moldova (2001) are members of the World Trade Organization (WTO). The other four countries have applied for membership. But even among the WTO members, the operation of the law of one price and commitment to market mechanisms are imperfect.

• The Kyrgyz Republic’s international trade is hampered by the country’s landlocked status. Almost all overland transport must transit Uzbekistan or Kazakhstan plus another country. The Kyrgyz domestic market is not integrated. There are large differences in incomes between the north and south, but the labor flows that would respond to and whittle down the wage differentials are not happening.⁵

• In 2001, Georgia, in breach of WTO rules, introduced export restrictions on scrap metal, its largest export, and on timber. Although the restrictions were removed in 2002, the episode raised questions about the government’s commitment to WTO obligations. In both Georgia and Moldova, secessionist forces limit the government’s control over parts of the national territory and its ability to monitor borders.

• Armenia and Azerbaijan are not yet WTO members, but their records on trade and foreign exchange liberalization do not fall much short of the CIS-7 countries that are WTO members. But their international trade is hindered by regional tensions that prevent normal relations between them and hamper development of regional initiatives. Despite a ceasefire in 1994, a formal agreement remains elusive. In Tajikistan, even since the formal end of the war in 1997, the central government has had difficulty exercising control over the entire national territory. Domestic commerce remains restricted by poor transport infrastructure, local government intervention (requiring permits and charging fees for crossing district boundaries), and licensing requirements.

• Uzbekistan is the laggard in price liberalization. Uzbekistan has a liberal tariff schedule, but currency convertibility has been severely restricted and key goods like cotton are still marketed through a state trading system. Both the foreign exchange controls and the state trading requirements are being liberalized, and Uzbekistan is committed to establishing current account convertibility. This pattern of delayed but eventual reform is typical of Uzbekistan. It resisted International Monetary Fund (IMF)-style macroeconomic stabilization in 1992–93, but after the government accepted the desirability of tightening monetary policies, implementation was purposeful in 1994. If the government fulfills its promises about currency convertibility, Uzbekistan will have a fairly liberal trade regime. The remaining holdout against price liberalization will then be the agricultural sector. The government recognizes the allocative efficiency arguments for ending state marketing, but it remains dependent on cotton for substantial revenue,⁶ so it would need to introduce alternative revenue-raising measures that could be politically costly.

Enterprise Reforms and Enforcement of Budget Constraints

Privatization of small companies and of housing was effectively completed during the 1990s, but privatization of large enterprises has proceeded much more slowly, especially for strategic enterprises, which often have included infrastructure providers. Even when formal changes in ownership status have occurred, management change and restructuring have rarely followed.

Measures of private sector share in GDP could be landmark indicators of progress in transition, but the EBRD’s rounded numbers are very approximate estimates (see Table 4.2). In the CIS-7 countries, agriculture is a major sector in which ownership status is often opaque. Small and medium-sized enterprises operate on the margins of the official economy and are underreported. Ownership of a few major enterprises can affect the aggregate figure in the small CIS-7 economies—the Kumtor gold mine, jointly run by a Canadian company and a state entity, accounted for 16 percent of the Kyrgyz Republic’s GDP in 2001.⁷ The EBRD’s transition indicators for large-scale privatization are also imprecise.

• The Kyrgyz Republic was one of the most rapid reformers in the CIS during the early 1990s. Reforms included both small enterprise privatization and transfer of most large and medium-sized enterprises into private ownership through a voucher scheme.

• In Moldova, voucher-based mass privatization in 1994–96 led to whole or partial divestiture of more than a thousand large and medium-sized state enterprises, but it was considered unsuccessful because it resulted in fragmented ownership. The emphasis then shifted to cash sales by tender or by auction.

• In Armenia, despite early and rapid privatization of land, privatization of large and medium-sized enterprises followed a path similar to that in Moldova. A voucher period in 1995–99 was followed by an emphasis on cash sales, but large-scale privatization remains incomplete, and restructuring has been limited.

• In Azerbaijan, although small and medium-sized enterprises have been privatized, there has been only limited progress in privatizing large enterprises. Foreign purchase, while invited, is often discouraged by packaging unattractive options with desirable assets and by requiring investors to settle enterprises’ past debts. Restructuring has been slow, and high levels of interenterprise arrears and nonperforming loans owed by state enterprises have inhibited financial sector development and strategic foreign investment.

• In Uzbekistan, small and medium-sized enterprises were privatized in 1992–93, but privatization of large enterprises has been slow. A list of 258 large companies to be sold by tender was drawn up in 1998 but aroused little interest owing to high prices and restrictions on changing the output mix. A revised list in December 1999 included 162 enterprises, but for some enterprises less than 50 percent of shares were offered for sale. In sum, the state retains whole or majority ownership of many large enterprises. Even when the state is not a majority shareholder, it can postpone strategic decisions in shareholder meetings if the decisions do not accord with its interests.

• In Tajikistan, formal privatization of large enterprises has been very slow, although large parts of the state sector effectively have collapsed as many assets were stolen. Privatization of large and medium-sized enterprises was launched in 1992, but by the end of 1997 only 30 enterprises had sold shares, and even these enterprises had 60 percent of their shares unsold.

In general, privatization has not led to desirable restructuring in the CIS-7 countries, especially when the incumbent managers retain control. Reforms seeking to displace the incumbents face determined resistance from that group.

Moldova has tried an innovative approach to inclusive restructuring. The Agency for Restructuring and Enterprise Assistance (ARIA), established in 1995/96 with World Bank support, had restructured 90 enterprises by 2001, creating more than 450 small and medium-sized enterprises.⁸ A key feature was ARIA’s commitment to not having permanent staff or providing finance, but rather encouraging its clients to solve their own problems and generate their own funding by sale or leasing of equipment or space. Thus creditors did not view ARIA as a cash cow that would service their nonperforming loans and had to accept a share of the loss from misallocated capital. Small and medium-sized enterprises were created by spinning them off from the oversized state enterprises or by leasing space to enterprises seeking to expand. Although the aim is internal restructuring rather than foreign investment, ARIA facilitates trade links so that the small and medium-sized enterprises can make subcontracting deals with European Union (EU) companies and also arranges for managers and workers to visit firms in the near West. The ARIA approach has been adopted by Armenia and Georgia.

Banking and Other Financial Market Reforms

All the CIS-7 countries replaced the monobank system of the centrally planned economy with two-tier banking systems, but none has created an efficient financial sector yet. The typical picture is one of too many undercapitalized new banks coexisting with successors to the old state bank, with all failing to provide effective financial intermediation. Indicators of financial development such as M3 (a measure of the broad money supply)/GDP are low (Table 4.4), and access to credit is difficult for small and medium-sized enterprises. EBRD indicators of financial reform show no progress in Tajikistan and very limited change (indices from 2− to 2+) in the other six countries. Some initial steps toward strengthening the banking sector are now being taken, although little has been done to develop securities markets or nonbank financial intermediaries.

Table 4.4.

Financial Indicators for the CI5-7, 2000

Source. EBRD (2002b).

¹M2 is used rather than M3.

Table 4.4.

Financial Indicators for the CI5-7, 2000

Indicator		Armenia	Azerbaijan	Georgia	Kyrgyz Republic	Moldova	Tajikistan	Uzbekistan
Number of banks		31	59	30	22	20	17	34
Number Foreign owned		11	5	8	6	11	4	6
Asset share of state-
	owned banks (percent)	2.6	60.4	0.0	—	9.8	6.8	77.5
Nonperforming loans as
	share of total loans (percent)	6.2	—	5.6	16.4	20.6	10.8	0.0
Stock market capitalization
	(percent of GDP)	0.8	—	—	0.3	0.1	—	1.1
Broad money (M3)/GDP
	(percent)	14.6	11.6	—	11.8	20.3	8.8	11.91

Source. EBRD (2002b).

¹M2 is used rather than M3.

View Table

Table 4.4.

Financial Indicators for the CI5-7, 2000

Indicator		Armenia	Azerbaijan	Georgia	Kyrgyz Republic	Moldova	Tajikistan	Uzbekistan
Number of banks		31	59	30	22	20	17	34
Number Foreign owned		11	5	8	6	11	4	6
Asset share of state-
	owned banks (percent)	2.6	60.4	0.0	—	9.8	6.8	77.5
Nonperforming loans as
	share of total loans (percent)	6.2	—	5.6	16.4	20.6	10.8	0.0
Stock market capitalization
	(percent of GDP)	0.8	—	—	0.3	0.1	—	1.1
Broad money (M3)/GDP
	(percent)	14.6	11.6	—	11.8	20.3	8.8	11.91

Source. EBRD (2002b).

¹M2 is used rather than M3.

• In Armenia, M3/GDP increased from 8 percent in 1995 to 15 percent in 2000, and nonperforming loans fell from 36 percent to 6 percent. However, the banking sector is small. At the start of 2001, banking system capitalization was US$61 million, but by September 2001 it had fallen to US$35 million, with total banking system assets of just over US$400 million.¹⁰ Eight of the 30 banks are under government-appointed administration. At the end of 2000, the last fully state-owned bank, the Armenian Savings Bank, was sold to Armenian investors with limited banking experience, and the bank needs much restructuring. To strengthen the banking sector, minimum capital requirements were increased in 2001 and 2002 and are set to reach US$5 million by 2005. Reforms in 2001 also included a new law for determining when a bank is insolvent and procedures for liquidating insolvent banks.

• Azerbaijan remains undercapitalized and overbanked, even though the number of banks declined from 210 in 1994 to 53 in 2001. Two state-owned banks dominate the sector, accounting for more than half the banking sector’s total assets.¹¹ A sign of the sector’s weakness was the March 2002 withdrawal from retail banking by HSBC, the only major international bank active in that market. In 2001, the Central Bank moved to improve internal audit capacity and accounting policies and procedures and introduced minimum capital requirements of US$5 million for new banks and US$2 million for existing banks (raised to US$2.5 million in 2002). In late 2000, an estimated 60 percent of banks were not in compliance with minimum capital requirements.¹² The Baku Stock Exchange started trading in 2000, but apart from T-bills, trading has been minuscule.

• Georgia has too many banks, and they provide limited financial intermediation (only 3 percent of the population have bank accounts), with big spreads between deposit and lending rates (10–20 percent). The share of nonperforming loans fell from 40 percent in 1995 to 6 percent in 2000, but in late 2000 a number of banks were in financial distress because of management and corporate governance problems. In early 2001 six banks closed. In January 2001, minimum capital requirements were increased from US$ 1 million to US$2.4 million, and a new asset classification system and international accounting standards were required for most banks. In October 2001, a new banking law established a legal framework and procedures for dealing with distressed banks.

• The financial systems in the three Caucasus countries are highly dollarized. The share of bank deposits denominated in U.S. dollars in September 2001 was 81 percent in Azerbaijan, 80 percent in Armenia, and 78 percent in Georgia. Macroeconomic stabilization and low inflation (see Table 4.1) have not generated sufficient confidence in the national currencies. Cash transactions remain ubiquitous and are associated with large shadow economies, whose existence perpetuates low levels of tax collection. Denationalization of the money supply and preference for cash also characterize the Kyrgyz Republic, Moldova, and Tajikistan, but to a lesser degree.

• The Kyrgyz banking sector was transformed between 1993, when it was state-bank dominated and 92 percent of loans were nonperforming, and 1998, when nonperforming loans were negligible. In 1998, M2/GDP was 14 percent, and state-owned banks held only 7 percent of the banking sector’s assets. The banks were fragile, however, and public confidence was undermined by a banking crisis in 1998–99 in which three of the four largest banks were liquidated. Banking sector assets fell from US$ 160 million to US$90 million at the end of 2000. Since the crisis, financial sector reforms have been on hold. The Central Bank retains ownership of key banks. Parliament blocked a tender for privatizing the Savings and Settlement Corporation. In July 2001, the minimum capital requirement was raised to US$2.14 million, but only about a third of banks are complying. The government aims to improve prudential regulation, auditing standards, and public confidence, but it recognizes the need for institutional reforms to encourage bank lending and for technical change to improve efficiency. Reform efforts have been undermined, however, by political interventions halting corrective measures when they threatened favored banks.

• In Moldova, all banks except the Savings Bank became private in 1992. Broad money has been increasing, reaching 23 percent of GDP by the end of 2001. Nonperforming loans fell from 46 percent of loans in 1996 to 21 percent in 2000. Minimum capital requirements were increased in 2001. By mid-2001, 4 of the country’s 20 banks met the minimum capital requirements and could engage in domestic banking, 9 had twice the minimum capital requirements and could engage in foreign exchange operations, and 3 dealt in trusts and equities; the other 4 banks were suspended. In the second half of 2001, loans increased by 36 percent, and there was a further reduction in bad loans. The banking sector remains concentrated, with five banks accounting for two-thirds of total assets.

• Uzbekistan’s banking sector remains dominated by state banks (see Table 4.4), with the National Bank of Uzbekistan alone holding 55 percent of the assets. Current plans for privatizing it by 2004 appear unrealistic. Much of the government’s national investment plan is operated through the banking system, with credit channeled to state enterprises, often at preferential rates. Farmers also receive payment through the banking system. Many deposit holders have restricted access to funds in their accounts, although the government planned to cancel restrictions on cash withdrawals by the end of 2002.

• In Tajikistan, public confidence in the financial system is low, and bank deposits amounted to less than 3 percent of GDP in 2000. In June 2001, restructuring agreements between the National Bank of Tajikistan and the four largest commercial banks expired. The agreements led to improvements in the banks’ net worth, loan recovery, staff training, and credit assessment procedures, although two large banks (Amonatbank and Agroinvestbank) remain weak. The National Bank of Tajikistan intends to strengthen its supervisory capacity and to increase minimum capital requirements to US$2 million.

The degree of structural reform in the financial sector is obviously far behind that of Eastern European transition economies, but progress is difficult to assess. There is two-way causality: lagging financial sector development has held back the real sector, but banks have been constrained by the limited loan possibilities caused by poor economic conditions and by limited fee-earning activities. Ownership does not seem to be critical in itself; Moldova, for example, quickly reduced the role of state-owned banks, but its financial sector remains fragile. Mismanagement and lack of trained staff appear to be crucial, irrespective of ownership.

Although banks have enjoyed soft budget constraints, many banks (more than 170 in Azerbaijan) have been permitted to fail, and failure was usually connected to insider lending, mismanagement, or poor governance systems. Prudential regulation is improving, but legislation has run ahead of observance and enforcement. One problem is the number and quality of staff in bank supervision departments—and the high turnover. Stricter prudential regulations have often led to nonlending rather than to better lending, so that banks are not playing the financial disciplining role that would be expected in a well-functioning market economy. Nonlending is also connected to the broader institutional issues of the enforcement of contracts and ability to foreclose on delinquent borrowers.

Land Privatization and Redistribution

Although CIS-7 governments have espoused the desirability of increasing farmers’ incentives through land reform, governments have been reluctant to allow full private property rights over land. State and collective farm managers have resisted change strongly in some cases.

The most radical land reform in the CIS-7 has been in Armenia and Georgia, where independent private farms now dominate (see Chapter 7). The redistribution of land on the basis of family size resulted in very small farms and a steep decline in marketed output.¹³ Azerbaijan and Moldova transformed their farm sectors in 1998–99, although legislation permitting private ownership had existed in Moldova since 1991–92 and in Azerbaijan since 1995–96. Both countries issued paper land shares to rural households and then in 1998 started to convert the paper certificates into physical plots. The Kyrgyz Republic allowed private land ownership after a referendum in June 1998, although it was followed by a five-year moratorium on transactions involving land. These five countries have the most liberal land policies in the CIS-7.¹⁴ Tajikistan has enacted a land reform program but has made little progress in implementation owing to the complexity and cost of obtaining a certificate for cession of land and the renewed outbreak of civil war between December 1996 and June 1997. Uzbekistan retains full state ownership of land.

Several countries have introduced leasing arrangements for land or partial decollectivization of state farms. The Kyrgyz Republic established 99-year land use rights for private farms and a land registration system. Uzbekistan introduced 49-year land use leases, but lease transfers are restricted. Despite potential benefits, and references to China’s household responsibility system as a model for granting use rights without private land ownership, such reforms have had limited impact in the CIS-7 countries.

Except in Armenia and Georgia, corporate and other organizational forms retaining many of the attributes of the old collective farms have proliferated, making it difficult to establish how effectively the farm sector has been restructured. A key impediment has been the power of state farm managers to influence legislation and its implementation at the farm level.¹⁵ The division of state farm assets has been controversial, with many stories of farmers out of favor with state farm managers receiving a spare tractor wheel or other useless piece of equipment as their share. In Central Asia and Azerbaijan, access to irrigation is often more important than the land itself. The state’s role in agriculture remains especially strong in Uzbekistan, with state orders for the two main crops (cotton and wheat) and state involvement in the distribution of fertilizer, seed, chemicals, and machinery, as well as state financing and maintenance of the irrigation system. The organization of much of the land area appears to have changed little, though the choice of output mix and of inputs has changed.¹⁶

Transport and Telecommunications

A key impediment to growth in transition economies has been the failure to upgrade infrastructure. For the CIS-7 countries, the inherited networks of railways, urban transport, gas and electric power, heating, water, and wastewater were adequate. The telecommunications systems were primitive, and the newly independent countries fell further behind global standards as they responded weakly to the challenges and opportunities of technological progress. The provision of utilities also has deteriorated owing to lack of maintenance and poor management, closely related to the failure of large-scale privatization and restructuring.

Other Utility and Infrastructure Services

Interconnections Among Structural Reforms

Although the key dimensions of structural reform have been treated separately, the interconnections are important. The uncertain status of land privatization discourages banks from lending to farmers who might offer land as a collateral. The significance of large-scale privatization, or its absence, goes far beyond the direct impact. The continuing importance of state-owned enterprises and of large companies with unchanged management has established vested interests that not only oppose further enterprise reform but have also been able to maintain soft budget constraints (mostly through tax and energy arrears) and to forestall reforms that would increase competition. This situation has had economywide effects through poor governance, stunted revenue sources and earmarked public expenditures, and inadequate infrastructure.

The feeble record of new enterprise creation in the CIS-7 countries is tied to the overall business environment and reflects lack of structural reform as well as inadequate protection of property rights and other governance issues. Changing the incentives of old enterprises, promoting the proper development of new ones, and establishing a positive and effective political economy of reform are interconnected.²⁰ In practice, enterprise reform and facilitation of new entry have been complementary. However, the literature attempting to measure the relative impact of bureaucratic red tape, lack of finance and law and order, and other obstacles to entrepreneurship remains inconclusive. The evidence from the CIS-7 countries is unhelpful on this question, because there are too many obstacles and too few successes to identify significant empirical patterns.

Throughout the CIS-7 countries, a culture of nonpayment undermines respect for economic contracts, and the tradition of nondisclosure and tax avoidance impedes the spread of accounting and auditing to international standards. Quasi-fiscal subsidies (such as acquiescence in nonpayment of utility bills) helped poor households cope with the transitional recession of the first half of the 1990s but have become a wasteful approach to social security. A critical issue in the CIS-7 countries is the need to transfer social security provision from the energy sector to the general government budget.

Revenues and Expenditures

Government revenues as a share of GDP vary considerably, but inadequate revenue is a major problem for most of the CIS-7 countries. Only Uzbekistan has managed to maintain public revenues and to keep public expenditures more or less in line with revenues (see Table 4.5), thanks to its success in maintaining and rebuilding state institutions, limited reform of the enterprise sector, and extraction of resources from a farm sector that produces a valuable staple export (cotton). In the Kyrgyz Republic and Moldova, state revenue collection has contracted, and expenditures have not been tailored to the reduced resources. Inflationary pressures are not under control, yet the governments find it difficult to reduce their expenditures in the face of high poverty levels and other demands. Azerbaijan’s oil wealth has enabled the country to avoid serious external debt, but government revenues remain low. In Georgia and Armenia, public revenues amounted to only 15 and 17 percent of GDP, respectively, in 2000, and both countries had substantial budget and current account deficits.

Some of the CIS-7 countries are working to boost public revenue or to reduce expenditures without reducing services, but it is difficult to assess the success of these efforts. Armenia is streamlining the tax system to lower the overall tax burden and reduce incentives for tax evasion. The small business tax code has been simplified, and the new tax code approved in December 2000 put corporate taxes on a progressive scale, with a 20 percent flat rate for large businesses. Budget revenues rose 20 percent in the first five months of 2001 over the same period in 2000, but fell in the second half of 2001, reflecting continuing collection difficulties. Georgia took steps in 2001 to improve revenue collection by reducing corruption among customs officials. Further steps to improve revenue collection included tests for tax officials, new procedures for monitoring tax officers, and changes to the tax code. The Kyrgyz Republic has focused on improving budgetary provisions and cutting costs (closing local offices, reducing duplication of services, cutting the workforce). And all the CIS-7 countries are implementing treasury and public accounts auditing systems.

An important challenge on the expenditure side has been to replace subsidies for loss-making enterprises and their employees with true adjustment assistance. A quick fix in the early transition era was to use quasi-fiscal measures such as energy or tax arrears to cushion the adjustment costs for individuals and enterprises, but this was an inefficient social assistance policy for households and it perpetuated soft budget constraints for enterprises. The philosophical prerequisite for shifting from subsidies to producers—replacing a worker orientation with a consumer orientation—is still a way off in the CIS-7 countries. Targeted social assistance funded out of general government revenue is superior to inegalitarian exemption from taxation or to indirect assistance provided in an untargeted manner by enterprises whose primary purpose should be efficient provision of services. Uzbekistan appears to be relatively advanced in dealing with arrears and in targeting social assistance from the state budget to those in need.

Also on the expenditure side, the state should provide primary health care and education, infrastructure, and law and order. Public expenditure on education has been highest in Uzbekistan (8 percent of GDP) and lowest in Armenia, Tajikistan, and Georgia (around 2 percent of GDP). Public expenditure on health has been low in all the CIS-7 countries, ranging from 4 percent in Moldova and Uzbekistan to less than 2 percent in Tajikistan, Azerbaijan, and Georgia.²²

Figures for public expenditures on infrastructure and law and order are not readily available, but one area that is important to several of the CIS-7 countries is the maintenance of the irrigation systems on which much of agriculture depends. In Moldova, the share of irrigated land decreased from 49 percent in 1995 to 4 percent in 2000, as initial disorganization and then malfunctioning of the irrigation system amplified the impact of droughts in 1992, 1994, 1996, and 2000. In Tajikistan, obsolete irrigation systems contributed to the severity of the 2000 drought, the worst in 74 years, and are the proximate explanation for the much greater decline in cotton yields than in neighboring Uzbekistan.

Law and Order

In addition to safety from street crime and violence, in a market economy there is need for an effective judicial system to protect property rights.

Legal and judicial reform is difficult to measure. A study of five transition economies, including Georgia and the Kyrgyz Republic, found progress in the details rather than in the legislation or grand reform packages.²³ The introduction of judicial qualification exams and higher salaries for judges had a dramatic effect on the independence and effectiveness of the judiciary in Georgia, while underpaid police officers were still levying arbitrary fines on motorists. According to the EBRD’s mid-1999 Business Environment and Enterprise Performance Survey, commercial courts are especially corrupt in Azerbaijan, the Kyrgyz Republic, and Moldova, even though the latter two are leaders in legislating judicial reform.

The EBRD’s legal transition indicators (Table 4.6) are based on assessments by participants (lawyers and academics) of the state of legal reform. They cover both the extent to which commercial law and financial regulations have reached internationally acceptable standards and the degree to which the laws are enforced. The legal transition indicators consistently show higher marks for extensiveness of legal transition than for the effectiveness of commercial law or financial regulations.

Table 4.6.

Legal Transition Indicators for the CIS-7 Countries, 2001

(EBRD index)

Source: EBRD (2002b).Note: The measurement scale is from 1, little or no change, to 4+, representing the standards of an industrialized market economy.

Table 4.6.

Legal Transition Indicators for the CIS-7 Countries, 2001

(EBRD index)

Indicator	Armenia	Azerbaijan	Georgia	Kyrgyz, Republic	Moldova	Tajikistan	Uzbekistan
Commercial law
Overall	2+	2+	3	—	4-	2	3
Extensiveness	3-	3	3	—	3+	2	3
Effectiveness	2	2	3	—	4-	2	3
Financial regulations
Overall	3	2	3-	—	3+	2	2 +
Extensiveness	3+	2+	3	—	4	2	3-
Effectiveness	3	2	2 +	—	3	2-	2-

Source: EBRD (2002b).Note: The measurement scale is from 1, little or no change, to 4+, representing the standards of an industrialized market economy.

View Table

Table 4.6.

Legal Transition Indicators for the CIS-7 Countries, 2001

(EBRD index)

Indicator	Armenia	Azerbaijan	Georgia	Kyrgyz, Republic	Moldova	Tajikistan	Uzbekistan
Commercial law
Overall	2+	2+	3	—	4-	2	3
Extensiveness	3-	3	3	—	3+	2	3
Effectiveness	2	2	3	—	4-	2	3
Financial regulations
Overall	3	2	3-	—	3+	2	2 +
Extensiveness	3+	2+	3	—	4	2	3-
Effectiveness	3	2	2 +	—	3	2-	2-

Source: EBRD (2002b).Note: The measurement scale is from 1, little or no change, to 4+, representing the standards of an industrialized market economy.

Reservations about implementation apply to government services in general. By the EBRD measure of administrative corruption (bribe payments as a percentage of reporting firms’ annual revenue), 6 of the CIS-7 countries (Azerbaijan, the Kyrgyz Republic, Armenia, Georgia, and Moldova) ranked among the 6 worst of 20 transition economies.²⁴ Tajikistan and Uzbekistan were not included in the 20, but bribe taking is believed to be extensive in both.

In Armenia, a high-level commission has been set up to fight corruption and bribery, reform the civil service, and increase the transparency of administrative procedures. In Azerbaijan, the Chamber of Accounts has authority to audit government agencies and publish its findings. Recent administrative reforms include merging key ministries, improving the effectiveness of the civil service, and transferring policymaking and regulatory functions from state enterprises to relevant ministries. A key implementation issue is the extent to which the state oil company and the Oil Fund operate at arm’s length from the government. In Georgia, a November 2000 anticorruption strategy aimed to establish fair and transparent tax rules, introduce a code of ethics and higher salaries for government workers, and strengthen the judiciary through tighter selection and testing criteria. The last point appears to have had some success, but there is widespread skepticism about the government’s commitment to fighting corruption.

Although the CIS-7 governments are all willing to acknowledge the existence of corruption and several have active anticorruption policies, corruption remains a major problem. Transparency International’s Corruption Perceptions Index for 2002 included 4 of the CIS-7 countries among the 102 countries covered: Uzbekistan ranked 68, Georgia 85, Moldova 93, and Azerbaijan 95 (with a higher ranking indicating a higher perception of corruption).

The Shadow Economy and an Effective State

Whether an effective state has been created is difficult to establish empirically, but from an economic perspective one indicator is the extent of the shadow economy. Economic agents operating in the shadow economy have chosen to exit the official economy, presumably because the benefits from being in the official economy do not compensate for the restrictions and obligations. Explanations of the shadow economy typically focus on the desire to avoid taxes or burdensome regulations, but there are costs in terms of lost access to public services and the formal financial sector. In most of the CIS-7 countries, the burdens of discretionary regulation by corrupt officials are substantial and the benefits of an effective financial, legal, and social system are lacking, so the shadow economy flourishes (see Chapter 9).

All measures of the size of the shadow economy have weaknesses. The most common approach is to relate total output to some measure of inputs, and the shortfall of officially reported GDP from estimated GDP is ascribed to the shadow economy. Several such measures are reported in Table 4.7. They are roughly consistent in showing high levels of shadow economy activity in six of the CIS-7 countries and lower levels in Uzbekistan.

Table 4.7.

The Shadow Economy as a Share of Official GDP in the CIS-7, 1994–96

(In percent)

Source: As noted in table: Lackó (1996); Johnson, Kaufman, and Shleifer (1997); Eilat and Zinnes (2002).Note: All the approaches reported here are based on electricity intensity, adjusted for changes in other inputs, economic structure, relative prices, and other factors. The absolute measures should be treated with caution, but as an ordinal ranking they provide a useful guide.

¹Numbers in parentheses are rankings among countries studied, from lowest to highest.

Table 4.7.

The Shadow Economy as a Share of Official GDP in the CIS-7, 1994–96

(In percent)

	1994—95 average			1996
Country	Lackó(1996)	Johnson, Kaufman, and Shleifer (1997)	Eilat and Zinnes (2002)	Eilat and Zinnes (2002)1
Armenia	—	—	77	93(6)
Azerbaijan	49	146	49	101 (7)
Georgia	62	171	121	105 (8)
Kyrgyz Republic	37	—	201	199(12)
Moldova	—	61	128	140(11)
Tajikistan	—	—	108	137 (10)
Uzbekistan	29	9	9	15(1)

Source: As noted in table: Lackó (1996); Johnson, Kaufman, and Shleifer (1997); Eilat and Zinnes (2002).Note: All the approaches reported here are based on electricity intensity, adjusted for changes in other inputs, economic structure, relative prices, and other factors. The absolute measures should be treated with caution, but as an ordinal ranking they provide a useful guide.

¹Numbers in parentheses are rankings among countries studied, from lowest to highest.

View Table

Table 4.7.

The Shadow Economy as a Share of Official GDP in the CIS-7, 1994–96

(In percent)

	1994—95 average			1996
Country	Lackó(1996)	Johnson, Kaufman, and Shleifer (1997)	Eilat and Zinnes (2002)	Eilat and Zinnes (2002)1
Armenia	—	—	77	93(6)
Azerbaijan	49	146	49	101 (7)
Georgia	62	171	121	105 (8)
Kyrgyz Republic	37	—	201	199(12)
Moldova	—	61	128	140(11)
Tajikistan	—	—	108	137 (10)
Uzbekistan	29	9	9	15(1)

Source: As noted in table: Lackó (1996); Johnson, Kaufman, and Shleifer (1997); Eilat and Zinnes (2002).Note: All the approaches reported here are based on electricity intensity, adjusted for changes in other inputs, economic structure, relative prices, and other factors. The absolute measures should be treated with caution, but as an ordinal ranking they provide a useful guide.

¹Numbers in parentheses are rankings among countries studied, from lowest to highest.

Creation of an effective state is the most difficult problem facing formerly centrally planned countries. In cross-country studies of transition economies’ performance, variables such as distance from Düsseldorf or years under Communism are correlated with growth, but they suffer from multicolinearity, which makes it difficult to isolate the nature of the Western transition economies’ edge. One dimension of that edge is successful state building. The CIS-7 countries, by contrast, include the transition economies that have gone through the most difficult experience of state building. Among the CIS-7 countries, Uzbekistan stands out as the most successful state builder, and it also enjoyed the best output performance as well.