Uneven governance and fiscal failure : the adjustment experience in Turkey

PSD Occasional Paper No. 17 September 1995 Uneven Governance and Fiscal Failure: The Adjustment Experience in Turkey Izak Atiyas - TThe World Bank Private Sector Development Department   Private Sector Development Department Occasional Paper No. 17 Uneven Governance and Fiscal Failure: The Adjustment Experience in Turkey Izak Atiyas September 1995 BuildinCntt.cefomp The World Bank Private Sector Development Department  C:  Contents 1. Introduction .................................................... 1 2. Background: The State, the Economy, and Party Politics before 1980 ................................................... 3 3. The Adjustment Period ...................... ............................. 11 Economic Reform: An Overview ........................................................ 12 The Limits of Income Redistribution ........................................................ 14 Impact of Distributional Pressures on Public Finance ................................... 21 The Crisis and the Stabilization Program of 1944: Does History Repeat Itself? ........................................................ 26 Does History Repeat Itself? ...................... .................................. 26 4. Governance in the Economic Reform Period ................ ............... 29 Distributional Conflict and the Problem of Cooperation in Politics ........................................................ 33 5. Concluding Remarks ................................................... 37 Bibliography .41 Appendix .47 . .   1 Introduction 1.1 Starting in 1980, Turkey embarked on an ambitious program of stabilization and structural adjustment. The speed with which macroeconomic stability was established and the rapid response of the private sector to the new regime made Turkey a success story among adjusting countries. Turkey has succeeded in liberalizing foreign trade, the financial sector, and the capital account, and in transforming a protected, inward-oriented economy into an export- oriented and private-sector-driven economy with a number of sectors that can compete internationally. 1.2 Nevertheless, fourteen years later, in early 1994, Turkey was once again faced with a major macroeconomic crisis, and had to embark on a new stabilization program. The crisis was driven by unsustainable fiscal deficits-overall public sector borrowing requirements (PSBR) reached almost 12 percent of gross national product (GNP) in 1993. The purpose of this paper is to examine the economic reform experience in Turkey and attempt to explain the factors that led to the failure in fiscal policy. 1.3 The main argument of the paper is that the failure in fiscal policy originates in a failure of govemance capacity. The aspect of governance that is most salient in explaining the Turkish experience is the inability of the state to resolve problems of cooperation between political actors. It is argued that the Turkish experience with fiscal policy, or, more specifically, the way state resources are used as instrument of political competition, lends itself to conceptualization as a problem of cooperation. The likely outcome is the uncooperative one. In the context of competition to acquire or maintain political power, political actors behave in a predatory manner and distribute the resources of the state to political constituencies, causing inflation and macroeconomic instability. In the more desirable outcome, by contrast, political competition is less predatory, has less destructive fiscal implications, and the political actors agree not to resort to "excessive" transfers and subsidies as a means of competition. Given the incentives to deviate into uncooperative behavior, however, the realization of this more desirable outcome requires either the existence of norms that restrain predatory behavior or mechanisms that allow the actors to commit themselves not to deviate. The governance failure is associated with the absence of the norms and mechanisms that would support cooperative outcomes. 1.4 The absence of such norms and mechanisms preceded the period of economic reform. Indeed, the inability of the mainstream political actors to reach compromises and to cooperate to resolve fundamental problems facing the country was an important factor in the economic and I  2 Uneven Governance and Fiscal Failure political crisis that afflicted Turkey in the 1970s. It seems that the failure is partially associated with a party structure that is excessively sensitive to the particularistic demands of constituencies. In any case, it is argued that this deficiency in governance was not addressed by governments that pursued economic reform in the 1980s. Instead, the centralization of governmental authority and an increase in the discretion over state resources during this period probably aggravated the problem of cooperation. 1.5 The existence of such a governance gap has important implications for the conduct of macroeconomic policy. One of the conclusions of the paper is that a sustained resolution of the problem of macroeconomic instability may require structural reforms that change the way the state manages distributional conflicts in society. It is argued that in the case of Turkey, attempts to establish macroeconomic stability by suppressing distributional demands temporarily, without undertaking such structural reforms, has preserved the vulnerability of the fiscal accounts to the process of political competition. 1.6 This failure, however, was coupled with considerable success in another dimension. By the end of the 1980s there was a wide consensus in the country that the preferred model of economic development would be predominantly market-oriented; the country would not be isolated, but instead would become increasingly integrated with the rest of the world. Moreover, there has been a significant change in the business culture: on the one hand, entrepreneurial activity is more appreciated. On the other, protectionist demands from industry find less support from public opinion. 1.7 This paper argues that this consensus builds an irreversibility into economic reform and discusses the contribution of Ozal, the main architect of economic reform and the head of the government during the reforms of the 1980s, in the development of this consensus. 1.8 This combination of factors puts Turkey in an interesting situation. Fifteen years of reform have created a vibrant private sector, and a state that is lagging behind in attending to important regulatory and social functions. Having reformed the aspects of the economy that relate to the economic activities of the private sector, it seems that the majority of the remaining tasks have to do with reforming the state. 1.9 The rest of the paper is organized as follows. The next section provides a brief historical background. The third section describes the stabilization and adjustment program and documents the fiscal deterioration. The fourth section discusses the governance aspects of the stabilization and adjustment period. The fifth section concludes the paper.  2 Background: The State, the Economy, and Party Politics before 1980 2.1 During the first two decades of the Republic, political power in Turkey was monopolized by the Republican People's Party (Cumhuriyet Halk Partisi, CHP), which represented a coalition of military-bureaucratic elites at the center, and traditional notables of the otherwise disconnected periphery. Because of the devastating impact of the Great Depression in 1929 and the perception that the weak entrepreneurial base, concentrated mainly in commerce and dependent on foreign markets, could not promote structural transfornation and industrialization, the 1930s witnessed a systematic effort of state-led social change. In the realm of economic policy, this entailed an array of state activities, including imposition of exchange controls, major public investments in manufacturing, nationalization of foreign and local companies delivering public services, and imposition of import duties and quotas to protect industry. Private manufacturing started to develop under the tutelage of the state, and with the active participation of the bureaucracy (Keyder 1987). 2.2 The transition to multiparty politics occurred in 1945, when the CHP allowed the formation of an opposition party, the Democratic Party (Demokrat Parti, DP). The introduction of political competition and the ascendance to political power of the newly formed DP in 1950 represented a milestone in the country's political and economic development. With the introduction of political competition, peripheral social groups, especially the peasantry, were actively courted by political parties, especially the DP, for the first time. The political system became sensitive to the demands of diverse groups, especially those in the rural areas, creating incentives for these groups to participate in the political process. Whereas during the single-party era patronage was a form of exchange between the state and local notables and did not entail the active participation of peasants in the political process, political competition enabled peasants to acquire power and to bargain for services. Although the leadership of the DP was from the central elite, instrumental in their succession to political power was their ability to forge and mobilize a wide-ranging coalition of landed interests, small peasantry, and urban mercantile groups. The DP advocated political liberalism and promised an end to the oppressive reach of the state. An appeal to religious sentiments found resonance, especially in rural areas alienated by the top-down secularism of the CHP (Sunar and Sayari 1986, p. 173). 2.3 The multiparty system and political competition created incentives to use the resources of the state to win political support. Political parties, especially their local organizations, became 3  4 Uneven Governance and Fiscal Failure avenues for deals exchanging support during elections for a variety of benefits, ranging from employment in the public sector, roads, credit, waterways, and 'political factories" to purely personal assistance in alleviating the burdens of bureaucratic red tape. Political entrepreneurship-the ability to represent the interests of constituents in party organizations, and in local and even national political and administrative forums, and the ability to make deals with constituents-became a critical trait for success in politics. With accelerating migration to the cities, particularly after the 1950s, patronage networks in urban settings became very important elements of the political process as well.' 2.4 The DP was more liberal in its economic policy than the CHP; it came to power as the representative of deeply rooted antibureaucratic sentiment. While the 1 950s witnessed a booming private sector, however, the scope of state intervention did not diminish significantly.2 If anything, perhaps the political principles guiding the allocation of resources commanded by the state changed to allow the newly represented groups to receive a larger share, and the logic of political competition gained prominence.3 2.5 In 1960 the DP regime was overthrown with a coup. The new Constitution envisaged 'planned development," and the State Planning Organization was formed. Political power was soon transferred to civilians. The center-right Justice Party (Adalet Partisi, AP) became the inheritor of the DP tradition, while the CHP gradually came to resemble a social democratic party. The two competitors differed in the emphasis to be placed on growth and equity and the desired sizes of the private and public sectors of the economy, but at least until the late 1970s there was a consensus among the main political actors on the basic premises and soundness of what has been called import-substituting industrialization (ISI). The goal of industrialization was paramount, and it was supported by business and labor. In the Turkish version of the ISI, almost no economic group was left out, and each obtained something from the state. Business and labor participated as the agents of industrialization. Peasants participated because they had the power to vote. 2.6 The consensus over ISI entailed a large extent of state intervention: state-owned banks allocated subsidized credit to the private sector, especially to industry and agriculture. While state economic enterprises (SEEs) were originally conceived as the main engines of industrialization in the 1930s and 1940s, when the private sector was very weak, over time they came to play a I See Sayari 1977; Sunar 1990, 1991; and Sunar and Sayari 1986 for discussions of patronage systems in Turkey and their link to the development of the multiparty system. Leder (1977) discusses the impact of political competition in rural areas. Karpat (1974) traces the development of the political entrepreneur in the urban shantytowns. 2 The DP experimented with a more liberal import regime in the early years of its rule, but returned to rigid import controls by 1955. 3 Keyder (1987) describes Menderes, the leader of the DP and the prime minister, in the following way: "he seemed to despise the ordinary constraints of accounting and was endlessly inaugurating public works projects without regard to cost" (p. 134). These were financed through the Central Bank, which in turn caused a doubling of prices between 1955 and 1959.  Background: The State, the Economy, and Party Politics before 1980 5 crucial role in the development of the private sector. They were dominant in the production of critical intermediate inputs, such as chemicals and petrochemicals, which were sold to the private sector at subsidized prices. They were also instrumental in subsidizing agriculture through the production and subsidized allocation of fertilizers. Finally, they were an important source of employment for surplus labor. 2.7 Import licenses were another significant source of rents for the private sector. Import- substituting private industry thus benefited from protection, from the rents associated with the quantitative restrictions that characterized the import regime, subsidized inputs, various forms of fiscal incentives, and subsidized financial resources. The agricultural sector, in addition to receiving subsidized credit, also benifited from support prices-which, on average, surpassed world prices-and subsidized fertilizers and seeds. 2.8 For business, personal contacts with members of the cabinet or with the bureaucracy have been important for access to incentives, subsidized credit, and government contracts. Apart from personal contacts by leading entrepreneurs, business associations have generally been weak in exerting direct influence on the policymaking process (Bugra 1994; Onis and Webb 1994). Nevertheless, these associations were influential as voting blocks in elections. The voting power of small business and agriculture probably explains the very narrow tax base of public revenues in Turkey, as well as persistent efforts of the center-right parties to dominate the elections of business associations. 2.9 In agriculture, the predominance of smallholdings precluded direct influence on the government or its ministers. The main channels of influence were elections, local party congresses, and contacts with local party leaders. To the extent that organizations existed, their survival and any influence they had was mainly the product of the center-right parties', especially the AP's, establishment of control over these groups by placement of party sympathizers in their governing bodies (Erguder 1981, 1991). 2.10 The transition to multiparty competition affected the bureaucracy as well. The CHP regime had cultivated a bureaucratic elite, one that was loyal to the mission of state-led modernization and Westernization. It evolved to see the formulation of public policy as its legitimate responsibility, and it did not have a conception of public interest that would arise out of a competitive political process: 'Thus, it was a responsibility of the bureaucrats to carry out public services on the basis of 'objective' criteria, and not in accordance with the 'whims' and 'illegal' interventions of the politicians" (Heper 1990, p. 609; see also Heper 1987, 1989). The bureaucratic elite also saw itself as the guardian of 'Kemalism" and its statist orientation toward public policy, and viewed the new political elite as opportunists who misled and exploited an uneducated population. The political elites portrayed the bureaucracy as alienated elitists who represented an authoritarian state and were unresponsive to the expression of popular demands. The political class attempted to subjugate the higher civil service rather than devise clear mechanisms of authority and delegation, if only to turn the civil service into a passive and obedient agency that distributed the politically directed spoils. This conflict generated several outcomes, especially in the 1970s. First, the bureaucracy became extremely politicized, and the civil service ranks came to be filled by party clients and partisans. Changes in governments were followed by major reshuffling of the civil service. The civil service also became weak financially,  6 Uneven Governance and Fiscal Failure and real salaries declined significantly over the years, which discouraged talented people from joining the bureaucracy. The bureaucracy, in turn, tried to protect its power by engaging in 'pathological bureaucratization" (the term is from Heper 1977, p. 80)-policies imposed by governments would be sabotaged in a maze of rules and regulations. 2.11 By the end of the 1970s, on the eve of the adoption of a major program of structural adjustment and a military takeover, Turkey was in a deep economic and political crisis. There was a severe crisis of governance. The party system had become polarized and fragmented, leading to weak governing coalitions at the mercy of small parties. As new and more ideologically oriented parties emerged, the mainstream parties on the right and the left became unable to cooperate sufficiently to address some of the basic problems afflicting the country- most notably, the need to curb political violence and to find solutions to a mounting balance of payments crisis. The inability of the parties in the center-right and center-left to cooperate allowed ideologically oriented parties to bargain for political power incommensurate with their electoral power (Erguder 1988, pp. 122, 124). 2.12 A variety of reasons have been proposed to explain the failure of governance that increasingly characterized the 1970s. Such explanations have included the emergence of new cleavages generated by the immense social transformation of Turkey as a result of rapid economic development; a tendency among political leaders to see political power as absolute; and the attempts of the mainstream parties to counter competition from new, radical groups, both on the right and the left (see Erguder 1988; Ozbudun 1988; Turan 1988). What is especially interesting for the purposes of this paper is that the combination of the state's discretionary command over a large amount of resources and a party system that became extremely responsive to the localized and particularistic demands of its constituencies inhibited the political system's ability to address matters of public policy. The first component of this combination is straightforward, but the second may require further exploration. 2.13 This excessive sensitivity4 has been noted by many political scientists in Turkey, and its origins are likely to lie in the particulars of the patronage system that developed in response to the introduction of political competition. It seems that in addition to the dynamics of interparty competition, patterns of intraparty competition and bargaining also played a role. Standing as a candidate for parliamentary elections depended on the candidate's ability to obtain the support of party delegates at the subprovincial level, where support was made available in return for contingent claims on various benefits if the candidate was successful in the elections. The delegates acted as "brokers" between the party and the constituency, and they were appointed by the party leaders because they could mobilize constituencies, provided, of course, that the broker himself could deliver to the constituencies. To honor the liabilities, the candidate relied on the party leadership. While this granted substantial authority to the leaders of government parties, as well as an ability to control the actions of their deputies, the same mechanism implied that the maintenance of leadership status was contingent on the government's ability to cater to the liabilities: the deputy could defect if the party leadership failed to grant the service. Party leaders, 4 Gunes-Ayata (1994, p. 36) calls this the 'subservience" of party officials.  Background: The State, the Economy, and Party Politics before 1980 7 including those in the government, became preoccupied with keeping their parties together, rather than attending to pressing public policy. The legislature ceased to be an institution of deliberation and legislation, and was in constant stalemate.5 The domination of constituency concerns over others in the behavior of the legislators affected the functioning of the GNA [Grand National Assembly, the Parliament] in several ways: First, many of the sessions of the legislature could not start because most legislators were so busy that they could not attend sessions, usually making it impossible for a quorum to obtain. Second, parliamentary duties turned into instruments which a deputy used for the exclusive advancement of his own career. For example, a position on the committee on State Economic Enterprises which is supposed to exercise oversight on the activities of these institutions, became a channel through which jobs could be found for constituents. Membership in the Committee of the Budget and Economic Planning became an arena where, mostly, pet constituency projects could be pushed. In summary, the Parliament became a grand-spoils operation, little concerned with other functions. [Turan 1988, p. 91] 2.14 It seems that the system suffered from a severe public goods problem: whenever a given liability was honored, all the benefits accrued to the specific constituency, but the cost (ultimately macroeconomic disequilibria and inflation) was socialized. This created an incentive to overspend. The institutional structure of fiscal expenditure management conformed to the political necessities. Information flow was extremely poor, and there was no system to monitor the efficiency of fiscal expenditures. The Central Bank was essentially subordinated to the government: fiscal deficits were financed mainly through monetary expansion, and the system also entailed substantial quasi-fiscal subsidies distributed by the Central Bank and state-owned banks. 2.15 Another important characteristic of the system was that it made political cooperation between different parties difficult. The increasing inadequacy of cohesion within the parties and the need of political leaders to juggle to cater to party members made it difficult to bargain successfully with other party leaders.6 Difficulty in cooperation exacerbated the public goods problem mentioned above, and with it, fiscal and quasi-fiscal balances. 2.16 The economic performance of the system was, on average, quite satisfactory until 1977. There were several instances of balance of payments problems, but they were resolved without straining the boundaries of the ISI framework. Between 1950 and 1975 real GNP increased about 4.5 times (an average growth rate of 6.2 percent). The economy went through significant structural change; the share of industry in gross domestic product (GDP) increased from 11 percent in 1950, to 18 percent in 1965, and to 22 percent in 1975. Growth was inward- oriented; between 1975 and 1980 the export/GNP ratio was less than 5 percent. 5 The Parliament's inability to elect a president in 1980 was one of the most visible indicators of its failure to function. 6 See Strom 1994 on the importance of intraparty bargaining in the degree of success of interparty bargaining in political coalitions.  8 Uneven Governance and Fiscal Failure 2.17 When the first oil shock hit in 1973, Turkish policymakers responded to the worsening current account deficits by running down reserves and borrowing abroad. Expansionary policies continued. The currency remained overvalued, and by 1979 international reserves were depleted. Lack of foreign exchange crippled industrial production (see Table 1). 2.18 Successions of coalition governments did not exhibit any capacity to deal with the political or the economic crisis. Two stand-by agreements with the IMF, in 1978 and 1979, failed because the government was unable to curb expenditures. In November 1979, the AP came to power and Demirel became the prime minister. On January 24, 1980, the government launched a program of economic stabilization and liberalization. The architect of the program was Mr. Ozal, a former civil servant at the State Planning Organization (SPO). Political and social instability continued. The military took over in September 1980. The Parliament was dissolved, and all political parties were closed down. This represented a grand bargaining failure between the mainstream parties on the right and the left, the AP and the CLP. These were the parties that had the most interest in preserving democracy and cooperating to find viable solutions to the economic and political crisis. They failed.  Background: The State, the Economy, and Party Politics before 1980 9 Table 1. Macroeconomic Indicators (Percent) Growth Rates Change in wholesale' Year GNP Industry price index PSBR/GNP 1978 2.9 6.6 52.6 1979 -0.4 -5.6 63.9 1980 -2.3 -6.0 107.2 10.5 1981 5.3 7.4 36.8 4.9 1982 3.7 4.9 25.2 4.3 1983 4.6 8.0 30.5 6.0 1984 7.8 10.1 50.3 5.3 1985 4.5 6.3 43.2 3.5 1986 7.5 8.8 29.6 3.6 1987 9.3 9.5 32.0 6.1 1988 1.5 3.2 70.5 4.8 1989 0.9 3.1 64.0 5.3 1990 9.8 9.0 52.3 7.6 1991 0.5 2.7 55.3 10.3 1992 6.4 6.8 62.1 10.6 1993 8.1 8.7 58.4 11.7 1994 -6.01 -3.9 120.7 8.2 a. 1978-87, State Institute of Statistics, 1981 weights; 1988-93, State Institute of Statistics, 1987 weight. Source: GNP and industry growth rates: 1978-79, OECD (1987) Appendix Table B; 1980-91, OECD (1994) Appendix Table A (new series); 1992-94:,calculated from SIS figures. Percent change in wholesale price index: 1978-81, OECD (1987) Appendix Table E; 1982-94, OECD (1995) Appendix Table E. PSBR/GNP: 1980-83, OECD (1990) Table 21, 1984-94: see Table 2.  10 Uneven Governance and Fiscal Failure  3 The Adjustment Period 3.1 The military takeover changed the rules of the political game. The Parliament was dissolved, and all political parties were closed down. Union activity was suspended. Nevertheless, there was continuity in economic policy. The military government retained Ozal and granted him the main responsibility for conducting economic policy. Following a banking crisis in 1992, Ozal and his team resigned. The break was brief, and in 1983 Ozal returned as prime minister. 3.2 As emphasized by Onis (1992, p. 11), the military regime enjoyed legitimacy and support because it was seen as a temporary solution to the governance crisis that had gripped the country, especially because the military was successful in restoring law and order. Nevertheless, the return of Turkey to fully competitive politics with no entry barriers was gradual, and a brief summary here will be useful. First, general elections took place in 1983.7 The military allowed only a small number of parties to participate in the elections. Former political leaders were banned from participating in politics. Two of the parties that ran in the 1983 elections, one on the right and one of the left, were led by former soldiers and supported by the military. The third, the Motherland Party (Anavatan Partisi, ANAP), was led by Ozal himself. ANAP easily won the elections and Ozal became the prime minister. 3.3 Other parties eventually were allowed to form. On the right, the Great Turkey Party (Buyuk Turkiye Partisi, BTP) was formed in 1983, and was identified as the continuation of the AP. It was closed down by the military in the same year. The AP tradition was then picked up by a new party, the True Path Party (Dogru Yol Partisi, DYP). On the left, the Social Democracy Party was formed. This later was merged with the Populist Party to form the Social Democratic Populist Party (Sosyal Demokrat Halkci Parti, SHP). 3.4 The intensity of political competition dramatically increased after the banned leaders were given the right to engage in politics in a 1987 referendum. Demirel and Ecevit, the former leader of the CHP, and about 100 other politicians reentered the political arena. The majority in the referendum was a slim 100,000 votes. Soon after the referendum, in the general elections of November 1987, ANAP received 36 percent of the votes and, thanks to the electoral law, 63 percent of the seats in the Parliament. After that, electoral support for ANAP decreased. During the local elections of March 1989, ANAP's share dropped to 22 percent, and it emerged as the 7 See the Appendix for a chronology of political developments after the military takeover. 11  12 Uneven Govemance and Fiscal Failure third party. ANAP lost the general elections in 1991, and Demirel, as the leader of the True Path Party, became the prime minister in a coalition government formed with the left-of-center SIP. 3.5 Throughout this period, there were no fundamental divergences from the main trend in economic policy, which was to progressively reduce controls over private sector economic activity, reduce protection from imports, enhance export orientation, and increase Turkey's integration into the world economy. Indeed, as will be stressed in the next section, a consensus emerged that Turkey's future lay in an economic regime that was fundamentally different from the ISI. In apparent contradiction of this general trend, the weight of the public sector in the economy did not diminish. In the context of intensifying political competition, deficits in the public sector remained a particularly intractable problem, endangering macroeconomic stability. The rest of this section will review the political dynamics behind fiscal deterioration. Economic Reform: An Overview 3.6 The January 24 economic reform program had ambitious objectives.8 The program envisaged nothing less than a fundamental change in the mode of economic development, from a protected, inward-oriented approach with extensive state regulation and intervention in many areas of the economy to a stance that was export-oriented, with the private sector as the main engine of growth. 3.7 The initial stabilization package included a steep devaluation, price deregulation for industrial products produced by the private sector, and large increases in the prices of SEE products. Interest rates were deregulated in June 1980. Initial attempts at trade liberalization were undertaken in 1981. By 1982, the public sector borrowing requirements as well as the rate of inflation were reduced, and the economy resumed growth (Table 1). The outstanding aspect of growth was its export orientation-between 1980 and 1982 the exports/GNP ratio doubled, to 10 percent. 3.8 With Ozal's return as prime minister in 1983, economic reform proceeded at full speed, and the next few years witnessed reforms in diverse areas. In the financial sector, banking deregulation was accompanied by better supervision; a foreign exchange market, a stock exchange, and an interbank market were created. Foreign exchange transactions were deregulated, and the capital account was liberalized in 1989 (for details, see Atiyas and Ersel 1994). In the area of foreign trade, in addition to the sustained real depreciation of the currency, there were two important trends.9 The first was the liberalization of imports. Negative import lists were replaced by positive lists in 1983-84. By 1988 only thirty-three items were subject to any import licensing; quantitative restrictions were thus effectively eliminated. Tariffs and other surcharges on imports changed frequently, but there was a general reduction in these as well; ' A very partial list of detailed accounts of the reform process would include Celasun and Rodrik 1989; Krueger and Aktan 1992; Krueger and Turan 1993; Onis and Webb 1994; and Rodrik 1990, 1991. 9 For details of trade reform see Baysan and Blitzer 1990, 1991; and Krueger and Aktan 1992.  The Adjustment Period 13 unweighted average tariffs decreased to 11 percent (17 percent for consumer goods) by 1989 (Onis and Webb 1994, Table 5-5). 3.9 The second important policy tool used in the trade area was the export incentive. In the prereforim regime, exports were discouraged both by overvalued exchange rates and by protection from imports. In order to encourage exports, the government supplemented real depreciation of the Turkish lira with an array of incentives (Krueger and Aktan 1992; Rodrik 1993). These included provisions for foreign exchange retention, tax rebates (later reduced, and eventually phased out in 1989), export credits (with interest rates much lower than the very high commercial rates), foreign exchange allocation schemes, corporate tax reductions, and subsidies from a Support and Price Stabilization Fund. Exporters whose annual export volumes were above a minimum level were rewarded with additional tax rebates. These special incentives were instituted to promote large trading companies (Onis 1992), and they were eliminated in 1989. The amount of resource transfer effected by these incentives was substantial. According to Togan, nominal subsidies (net of indirect taxes) averaged about 32 percent of total exports in 1983, and 24 percent in 1984; they declined to 16 percent in 1989 and to 13 percent in 1990. Yeldan (1994) calculated that between 1982 and 1992, export subsidies were substantially higher than total corporate taxes collected by the government for all years except 1986, when they were almost equal. Arslan (1993) calculates that between 1988 and 1991 the fiscal cost of export incentive programs ranged between 2.0 and 2.4 percent of GNP. 3.10 Fast growth in exports became one of the most important yardsticks of the success of the adjustment program. Between 1980 and 1990, the value of exports increased from US$2.9 billion to US$13.0 billion. The share of manufactured goods in total exports increased from 29 to 72 percent. 3.11 The weakest component of the adjustment effort was macroeconomic stability. A major element of stabilization in the early 1980s was a significant retrenchment in the public sector. The public sector borrowing requirements (PSBR) were reduced from almost 9 percent of GNP to 4 percent between 1980 and 1981 (Table 1). This drastic reduction was achieved by cutting both the deficit of the central government budget and SEE losses. In the case of the consolidated budget, major reductions were made in personnel expenditures and transfers to SEEs. In the case of the SEEs, the main sources of the decrease were the deregulation of the prices and a relative decrease in wages and salaries (see Celasun 1990 for a review of fiscal policy in this period). A second important component of the decline in SEE borrowing requirements was a reduction in subsidies allocated to agriculture. (The role of SEEs in Turkish agricultural policy will be discussed in more detail below.) 3.12 The initial reduction was generally maintained until the mid-1980s, with an average PSBRIGNP ratio of 4 percent in 1980-86. The first significant jump occurred in 1987 as a result of the fiscal expansion associated with the election year, and the economy recorded an extraordinarily high growth rate of 9.3 percent. This was followed by a minicrisis, and a stabilization program was initiated in February 1988. The PSBR ratio then increased on a sustained basis until it hit a record 11.7 percent in 1993, followed in 1994 by a major crisis, a stabilization effort, and a record 6 percent drop in GNP (Table 1). Below it will be argued that the deterioration in the public accounts was closely linked to increased redistributive pressures on  14 Uneven Govemance and Fiscal Failure government finances. Before that, a closer look at the origins of these pressures in the labor markets and agriculture will be useful. The Limits of Income Redistribution 3.13 The stabilization and adjustment effort of the early 1980s entailed major redistributions of income. There were two conspicuous losers-labor and agricultural producers. In both cases, the political environment in the early 1980s was instrumental in affecting the redistribution of income. The more militant of the two largest trade unions, DISK, was dissolved by the military government. While the other, Turk-Is, continued under military rule, it operated under extremely restrictive rules laid down by the military government. The government ended strikes and banned union activity. Collective bargaining was replaced by the deliberations of the High Arbitration Board, which was controlled by the government. Moreover, the new Constitution banned unions from engaging in political activity or supporting or receiving support from political parties. The resulting decline in real wages was one of the basic pillars of the stabilization effort. For the economy as a whole, the downward flexibility of wages facilitated the real depreciation that was required for outward orientation. For the public sector, it facilitated the fiscal retrenchment that occurred in the early 1980s. 3.14 In the case of agriculture, the suspension of political competition eliminated the main channel of influence that traditionally had served as the vehicle to obtain subsidies from the state. The main channel of influence for agricultural producers, local party organizations, disappeared when the political parties were banned. Moreover, in the early 1980s new parties were formed that did not have the extensive organizations characteristic of the parties active prior to the military takeover. Finally, ANAP did not attempt to build up a rural constituency. The outcome was that there were significant reductions in both the number of commodities subject to support purchases and the real level of support prices offered to farmers. This eliminated one of the most important instruments of income redistribution in the sector.'( 3.15 The opening up of the political regime in the mid-to-late 1980s fundamentally changed the environment for macroeconomic policy. Collective bargaining was reinstituted in 1987 as part of the trend of redemocratization. In the spring of 1989, right after local elections in which ANAP, the governing party, lost, collective bargaining between Turk-Is and various SEEs, involving 600,000 workers, reached a deadlock. Workers were requesting a wage increase of 170 percent to compensate for the real wage losses of the previous decade. After protracted '