56123 v1 Distortions to Agricultural Incentives in Turkey Alison Burrell and Marianne Kurzweil Wageningen University and World Bank Alison.Burrell@wur.nl mkurzweil@worldbank.org Agricultural Distortions Working Paper 10, December 2007 This is a product of a research project on Distortions to Agricultural Incentives, under the leadership of Kym Anderson of the World Bank's Development Research Group (www.worldbank.org/agdistortions). The authors are grateful for helpful comments from workshop participants and for funding from the World Bank's Europe and Central Asia Department and Trust Funds provided by the governments of Ireland, Japan, the Netherlands (BNPP) and the United Kingdom (DfID). This Working Paper series is designed to promptly disseminate the findings of work in progress for comment before they are finalized. The views expressed are the authors' alone and not necessarily those of the World Bank and its Executive Directors, nor the countries they represent, nor of the institutions providing funds for this research project. Executive summary · In the last 35 years, Turkey's population doubled, and its real GNP trebled. The share of the employed workforce engaged in agriculture fell from 60 percent to 34 percent and agriculture's share of GNP fell from 31 percent to under 12 percent. The sector experienced recurrent negative real income growth. · In 2002, 36 percent of the population dependent on agriculture was below the poverty line (about double the rate in manufacturing and distribution). Nearly 20 percent of the rural population had consumption levels less half than the national average. Also within agriculture, the productivity and income gap between the richer and poorer regions has been increasing. · The 1930s and 1940s were characterized by state control of the economy and public services. The 1950s saw a shift towards more liberal domestic and external economic policies, which led to macroeconomic instability and strong stabilization measures. Five- year economic plans began in 1961, and for two decades the economy followed import substitution policies. After the economic crisis of the late 1970s, there was a switch to export-led growth policies and progressive trade liberalization. The period from the later 1980s has been marked by continuing macroeconomic instability, with two major currency collapses and deep recessions. The causes lie in deep-rooted weaknesses in Turkey's economic structure and public finance system, and in endemic distributional conflicts that democratically elected governments have difficulty containing. · The agricultural sector as a whole has been heavily supported since the latter 1980s (in contrast to earlier decades), with a strong bias towards import-competing farm products at the expense of exportables. Total annual transfers to the sector exceeded 5 percent of GDP in seven years during the period 1986-2004. Much of this support did not reach farmers but rather was captured higher up the value chain. · The nominal rate of agricultural assistance was less than that for non-agriculture until the end of the 1980s, but the transition from strong disincentives for farmers to strong positive incentives was very rapid over the 1980s and 1990s. · The average MFN tariff on agricultural imports (WTO definition) in 2004 was over 21 percent, but this average masks wide variation in rates between different commodities. The highest rates were for livestock products, fruits and sugar (over 90 percent). Cereals and cereal products were in the range 50-60 percent, whereas oilseeds, beverages, and cotton had average MFN rates between zero and 11 percent. The weighted average nominal rate of assistance (NRA) calculated for a group of major products covering around two-thirds of primary production averaged 23 percent for the period 1986-2004. · In 2000, an agricultural policy reform program replaced most government-determined prices and product and input subsidies with non-product-specific direct income support, but without any adjustment of external protection. A major objective of the program was to dismantle or reform the inefficient parastatal marketing system through which much of the agricultural support had been channelled. · The NRA has followed a strong cyclical pattern since the mid-1980s. Input subsidies paid directly to farmers increased the NRA by 8-22 percentage points above the protection or taxation exclusively resulting from border measures and production subsidies. This gap fell to about 8 percentage points in the early 2000s, after the switch to direct income 2 support. · The correlation between tariffs and NRAs based on the price gap between farm-gate producer prices and comparable border prices is poor for all individual commodities studied, suggesting a combination of factors such as non-tariff barriers, weak spatial price transmission, inefficiencies in transport and first-round handling, and high market share of state-owned enterprises purchasing at government-controlled prices. Caution is therefore needed in interpreting Turkey's NRAs. · The extent to which government intervention has distorted farmer incentives in the short term is probably small. Late policy announcements, inconsistent price changes and severely delayed payments in a context of high inflation have blunted signals and created uncertainty. In the medium term, the pattern of intervention has been a disincentive for investment and modernization, although it has allowed more marginal farmers to survive. · Reducing imperfections in the economic environment in which farmers operate, and improving the medium term consistency of policy implementation, would raise the transfer efficiency of existing support, which ­ to the extent that it is targeted appropriately ­ would help alleviate rural poverty. · More specifically, reducing policy-induced distortions alone may have a worsening effect on rural poverty in the absence of accompanying improvements in labor market flexibility, massive investment in training and job creation (particularly in rural areas), and restructuring of down-stream and off-farm sectors. · In October 2005, Turkey began negotiations for EU membership. Negotiations are still at a very early stage. It is expected that Turkey will align its border protection for agriculture and its domestic agricultural policies more closely with those of the EU, but not necessarily immediately and possibly not until nearer the end of those negotiations. Also, inflows of FDI triggered by the prospect of EU entry could stimulate job creation in other sectors that would absorb some of agriculture's surplus labor. In addition, pre- accession EU funds might be used to finance programs that would lift parts of the rural population out of poverty. Distortions to Agricultural Incentives in Turkey Alison Burrell and Marianne Kurzweil Introduction and summary The aim of this study is to give an overview of the evolution of Turkey's agricultural policies in recent decades, to examine the extent to which they may have distorted incentives, and to attempt an explanation of the underlying forces that have driven the process and conditioned the results. The review period for the assessment begins in 1961, and continues up to the present. However, in order to explain the situation prevailing at the start of this period, various trends and policy preferences are traced back to earlier decades. Turkey experienced rapid population growth in the second half of the twentieth century (Table 1). Although the population more than trebled between 1950 and 2004, heavy out- migration to urban areas and overseas meant that the rural population grew at half that rate. The agricultural workforce fell from 84 percent of the working population in 1950 to 60 percent in 1970 and 34 percent in 2004. Despite healthy real growth rates for the economy as a whole, the economic performance of agriculture was poor, with average annual real growth rates of 1.6 percent in the 1970s, 0.6 percent in the 1980s, and 1.5 percent in the 1990s. In 1950, agriculture's share of GNP was 41 percent, but by 2004 it stood at just 12 percent. Figure 1 shows that in the last 25 years of the period reported, agriculture experienced negative real growth in more than one year out of three. In 2002, nearly 20 percent of the rural population had consumption levels 50 percent or more below the national average, and 36 percent of the population dependent on agriculture was below the food and non-food poverty line (SIS 2004). And yet Turkish agriculture was highly supported throughout the closing decades of the twentieth century, with total support to the sector exceeding 5 percent of GDP in some years. The following study attempts to throw light on this enigma. 2 Overview of the Turkish economy and agricultural sector By way of background, this section first summarizes developments in Turkey's macroeconomy over the past half-century before looking more specifically at the agricultural sector. Macroeconomic developments Trends in agricultural policy over the review period are closely linked to the evolution of the economy as a whole, and cannot be assessed in isolation from the macroeconomic context. The period from the 1930s to 1950 was characterized in Turkey by state control of the economy and public services (etatism). The 1950s saw a shift towards more liberal domestic and external economic policies, which led to macroeconomic instability that required strong stabilisation measures before the end of the decade. Five-year economic plans began in the early 1960s, and for two decades the economy followed import substitution policies. After a severe economic crisis in the late 1970s, there was a switch to export-led growth policies and progressive trade liberalization. Thereafter, four sub-periods can be distinguished: 1980-88, 1989-94, 1994-99, and from 2000 until the present. The 1980s began with a period of post-crisis rehabilitation, under a 3-year military government. The 1980 stabilisation package included a large devaluation, together with stringent measures to curb inflation (running at over 100 percent) and to reduce a public sector borrowing requirement of 10 percent of GDP. Policies throughout the economy were targeted on export promotion and structural adjustments whose burden fell heavily on wage labor and agriculture. The share of manufactures in exports rose sharply, and manufacturing profit margins increased (Taymaz 1999). It was a period of low growth but relative stability for agriculture. During this period, the strict import controls, which had operated under the previous import substitution program, were gradually relaxed -- including those for agricultural commodities. Macroeconomic changes in response to populist pressures improved real wages significantly in the period 1988-1994. Agriculture shared in this trend, and saw its border protection and output subsidies significantly increased. Currency convertibility and deregulation of the capital market in 1989 allowed the authorities to resort increasingly to international capital markets to finance chronic budget deficits. Interest payments on escalating foreign debt pushed 3 up total spending requirements. As Demir (2002) points out, distributional conflict, in which groups disadvantaged by post-crisis measures attempt to restore their income shares, was not new in Turkey's 20th century history, but for the first time it took place in the context of full capital account liberalization. The result was that the period from the latter 1980s until 2001 was marked by more unstable macroeconomic cycles, with two major currency collapses and deep recessions. Growing deficits and exchange rate pressure continued until the economic meltdown of early 1994, which involved a massive currency devaluation and a new stabilisation package. International agencies downgraded Turkey's credit worthiness, restricting its external borrowing. From 1994 to 1999, the rest of the economy entered another period of export promotion built on a low wage policy. This time the agricultural sector did not bear the heaviest burden of adjustment: its border protection increased for a number of key products, and spending on the sector appeared to run out of control. In fact, during this particular cycle, the budgetary burden of spending on agriculture actively contributed to the impending macroeconomic collapse. During the second half of the 1990s, the government resorted increasingly to Central Bank borrowing for finance, fuelling inflation and spiralling interest rates. Before long the government had also returned to the international money market. By 2000, the ratio of interest payments on government debt to tax revenues was 77 percent, and in 2001 public debt was 91 percent of GNP (OECD 2002). Starting in 1999, several financial crises triggered another program of structural adjustment for the economy as a whole 1, in which agricultural policy reform had a prominent role. In early 2001, the 3-year exchange-rate based stabilisation program backed by the IMF collapsed, only 14 months after its launch, the crawling peg was abandoned and the Lira was floated. In 2002, Turkey experienced its deepest recession since the second world war (Demir 2004; Akyüz and Boratav 2003). The underlying causes of these cyclical patterns are complex. One is the apparent difficulty of establishing a secure, broadly based source of tax revenue. Various authors have documented the consequences of the `tax impotence' of successive Turkish governments (see, for example, Candemır 1994; Harrison, Rutherford and Tarr 1993). In particular, an inability to tax corporate profits and capital appropriately has meant that middle- and lower-income groups 4 are taxed disproportionately. This has discouraged labor force participation, with participation rates below 50 percent of the working-age population. Furthermore, over half the workforce operates in the informal (unregistered) economy, thereby further reducing the tax base (OECD 2004). Another consequence is that methods of public finance, together with inefficient credit allocation, have for years crowded out private investment and held back productivity growth and job creation. This has impeded mobility out of agriculture and deepened rural poverty. A pre-condition for coherent agricultural policies is macroeconomic stability: without it, prices cannot act as reliable signals and longer-term strategies for investment and development are thwarted. This pre-condition has not been consistently met in recent decades. Macroeconomic instability also has implications for the measurement of sectoral distortions. Against such a background, conventional methodologies for measuring distortions cannot easily distinguish between intended policy-induced effects or the desired outcomes of successful lobbying, on the one hand, and the fortuitous or unplanned results generated by unpredictable or uncontrollable cyclical movements on the other hand. This should be borne in mind in the following sections of the study. Turkey's agricultural sector Turkey uses 38-41 million hectares of land for agriculture, of which 22-26 million hectares are cultivated and the rest is permanent pasture, including common grazing lands. A wide variety of agricultural commodities can be produced in Turkey's diverse climatological and topographical conditions. Fruit and vegetables account for over 40 percent of the value of sectoral output, and field crops for one third. Over half the cultivated land is used for cereals. High-value protected crops are grown on about 50 thousand hectares. Livestock products comprise less than a quarter of total output value. Structural and resource weaknesses have hampered the performance of the sector for decades. Nearly two-thirds of farms are smaller than 5 hectares. Farms at all points in the size spectrum tend to be fragmented, with nearly one-quarter consisting of six or more disjoint parcels. Average parcel sizes have continued to decline in recent decades. Turkish inheritance law is a major factor in this trend. Investment per worker and per hectare is low, and yields for 1 According to Demir (2002), it was the seventeenth IMF rescue package for Turkey in 54 years. 5 field crops and livestock are comparable with those of extensive farming systems in Australia or Argentina. In 2003, 18 percent of the agricultural workforce was illiterate (8.5 percent of males, 28.5 percent of females), and only 11 percent of people working in agriculture had progressed beyond primary school (OECD 1994; Dervis et al. 2004). Regional economic inequalities are marked in Turkey and have been increasing. Average per capita GDP in the richest of Turkey's 26 regions is over 5 times higher than in the poorest region, with prosperity tending to decline as one moves from the Northwest towards the East and Southeast. This is matched by inequality in educational attainment, health status and provision, and life expectancy. The World Bank (2000) found that low incomes in the poorest regions were due mainly to the large share of agriculture in their local economies. However, whereas productivity appeared to be improving faster in the poorer regions for non-agricultural sectors, the reverse was true for agriculture where the productivity gap between poorer and richer regions was increasing. Rather than helping to redress this imbalance, agricultural support throughout the period has been biased towards richer regions and larger farmers. Although the majority of holdings produce for subsistence and for local informal markets, some areas and sub-sectors are fully integrated into national or international markets. The horticultural sector, concentrated in the Southern and Western coastal areas, competes successfully in export markets. 2 Poultry production has increased more than four-fold since 1980, and the sector is characterized by modern production units operating to EU standards. A similar duality is observed in the food supply chain. A significant share of output, particularly for livestock products, bypasses formal marketing channels altogether. Traditional marketing chains typically have a number of intermediate stages operated by small enterprises with weak financial structure and lacking modernised handling facilities. The processing sector suffers from low capacity utilisation. Marketing channels for some commodities have been dominated by inefficient parastatal enterprises for decades. At the same time, supermarkets and discount stores are gaining ground (Sirtioglu 2004; van Berkum 2005), although producer contracts and integrated chain management are not yet widespread. Figure 2 shows the evolution of Turkey's agricultural trade flows. The agricultural trade balance is always in surplus, whereas total trade typically shows a deficit. Agriculture's shares in 2 Turkey has been the world's third largest exporter of fruit and vegetables, after the USA and EU-15, for many years. 6 exports and imports were 88 and 21 percent respectively, and have fallen steadily to 9 and 5 percent in 2004. These sharply declining shares are consistent with rapid growth in the rest of the economy. Agricultural policies Agriculture has been heavily supported in Turkey for decades. The main policy instruments have traditionally been output price support and input subsidies, against a background of high border protection. This section describes the institutional setting, trade measures used and domestic policy instrumentation. Institutional setting The aims of Turkey's agricultural policies, as set out in successive Five-Year Development Plans, have focused on securing the availability and stability of food supply, enhancing output and yield growth, raising self-sufficiency and exploiting export potential, providing stable and sustainable income levels in agriculture, and fostering rural development. Each Plan has set targets and guidelines for achieving them. Formal authority for the formulation of annual programs involving specific agricultural policy measures resides with the Council of Ministers, in consultation with the ministries concerned, the State Planning Office and the Treasury. The prime responsibility for implementation belongs to the Ministry for Agriculture and Rural Affairs (MARA), with some responsibilities allocated to the Ministry for Industry and Trade (MIT) and the Ministry of Finance. A key group of institutional players in the agricultural policy arena have been the state economic enterprises (SEEs). It is important to note that SEEs flourished extensively throughout the Turkish economy during most of the period under review, until the recent wave of full or partial privatizations. 3 In agriculture, the earliest SEEs date from the 1930s (TMO for grains, TSFAS for sugar) and the 1940s (TEKEL for tobacco, agriculture and salt, TZDK for fertilizer 3 In the 1950s, SEEs had a 50 percent share of value added in Turkey's manufacturing and a `virtual monopoly' in other sectors (Flam 2003). Collectively, SEEs had nearly 600 thousand employees in 1990 (Candemır 1994). 7 and other inputs). EBK (meat and fish, later also poultry) and the Feed Industry Corporation were created in the 1950s, SEK (milk) in 1963 and ÇAYKUR (tea) in 1971. The OECD (1994) estimates that, for the period 1980-1992, the state-owned sector purchased 100 percent of sugar beet, and depending on the year, 40-85 percent of the tobacco supply, 10-40 percent of the marketed output of wheat, and up to two-thirds of marketed barley. Purchase prices were fixed by the government, and SEEs were subject to rigid protocols on trading operations. SEEs carried out manufacturing and commercial activities on behalf of the state, in line with strategic plans and annual directives from relevant government bodies. Beginning in the early 1980s, some SEEs lost their monopoly or monopsony powers, and there was a move to allow SEEs more autonomy in fixing prices. However, as government retained its right to set prices, until recent years SEEs were not managed fully in line with commercial principles. SEEs have also played a significant role in providing social services (social security, housing, guaranteed employment) (Demir 2002). The trading losses and capital needs of these organisations were regularly met from public funds. For the years 1991-95, the annual average duty losses of TMO, TEKEL and TSFAS taken together were USD 622 million, and they rose to an annual average of over USD 1.7 billion during 1996-2001. In addition, the government began writing off the debt of agricultural SEEs in the mid-1990s. The average annual debt write-off for TMO, TEKEL, TSFAS and ÇAYKUR during 1996-2001 was USD 550 million, whilst equity injections from the Treasury to agricultural SEEs averaged USD 150 million during the same period. Agricultural SEEs' pricing policies tended to reflect the political cycle. In pre-election periods, increases in selling prices were held below the rate of inflation, catching up only months later. Such practices were one of many factors contributing to their losses. For most of the second half of the twentieth century, Agricultural Sales Co-operatives Unions (ASCUs) also played a key role in the implementation of agricultural policy. ASCUs date from the 1930s, and mostly cover just one product, including important crops such as cotton, hazelnuts, sunflower, olive oil, raisins and sultanas. They provide warehousing, primary and/or secondary processing, packaging and marketing services to their members. According to OECD (1994), during the period 1980-1992, the three cotton ASCUs purchased between 24 and 92 percent of annual marketed output. For hazelnuts, sunflower seeds and soybeans, the relevant ASCUs' share reached 60-70 percent during that period. In 1993, there were 17 ASCUs with 387 8 member co-operatives (ASCs), and 685 thousand individual members. The ASCs were controlled by their respective ASCU, acting as its agents and having little autonomy or direct farmer control. By 2000, the number of ASCUs had fallen to 16 and member co-operatives to 330, whereas individual membership stood at 750 thousand and ASCUs and ASCs together had over 16.5 thousand employees. In the early 1960s, the state began to use ASCUs as agents for the support purchasing of a few commodities, and product coverage increased in the following years. The relevant legislation was amended to formalise this role for the ASCUs in 1985. ASCUs could obtain subsidised credit for part of the cost of support purchasing, and any duty losses were covered by government. ASCUs were also instrumental in channelling subsidised inputs to farmers. After 1994, ASCUs still declared prices to their members, although these prices were not set by government and purchasing was no longer done on the government's behalf. However, the government still appointed ASCU directors and key staff. Budget transfers to ASCUs for the years 1995-2000 averaged over USD 600 million per annum. Agricultural Credit Cooperatives (ACCs) and the state-owned Agricultural Bank also played a role in policy implementation, providing concessional credit to the industry. The Agricultural Bank dealt mainly with large farmers, SEEs and ASCUs, while the ACCs focused on smaller farmers. In 2000, the Agricultural Bank was restructured as `a joint-stock company whose structural and operational characteristics are those of a private sector concern but whose capital happens to be state-owned' (Ziraat Bank 2003). The network of the ACCs extends throughout rural Turkey. Their main role over the period has been to supply farm inputs, usually on a credit basis. Losses due to repayment default and inefficient management were covered by payments channelled through the Agricultural Bank. Agricultural trade policy A crop failure in 1954, coupled with difficulties in expanding agricultural output and monetary expansion, led the government to introduce a stabilization program in 1958. A military intervention in May 1960 terminated democratic party rule, and the new era of Five-Year plans began. With the start of the First Five-Year Plan (1963-67), import substitution became the official development strategy. Under the import substitution regime, most agricultural products could only be imported by a SEE. Moreover, only SEEs could import agricultural inputs such as 9 fertilizer and pesticides, often at an advantageous (i.e. overvalued) exchange rate. In January 1980, the first steps towards liberalizing this regime were taken: two sets of products ("liberalization lists") that could be imported were designated. Products not appearing on these lists were still controlled by quotas, which were abolished in 1984. The liberalization lists were then replaced by three lists designating, respectively, prohibited imports, goods requiring an import permit and goods that could be imported without restriction. Agricultural products could be found on each of the three lists. A system of product-specific customs duties was set up, complementing several umbrella levies (such as the Support and Price Stabilisation Fund 4 (DFIF) levy) that had already been introduced in the early 1980s with a revenue-raising objective. Additional special levies were added later in the decade, including the so-called Mass Housing Fund 5 (MHF) levy. By the early 1990s, in addition to regular customs duties, agricultural imports were generally subject to stamp duty (at 10 percent of cif value), wharf tax (at 5 percent of cif value inclusive of customs duty and some other charges), the so-called municipality share tax (15 percent surcharge on the customs duty), the SPSF levy (10 percent of cif value) and the MHF levy (OECD 1994). This complex structure, together with frequent changes in rates and coverage, characterized border protection over the period up to 1995. It is unclear to what extent its main purpose was to manage domestic markets, help producers cope with fluctuating adverse circumstances or raise revenue. The lack of transparency in such a system makes the net extent of border protection hard to evaluate. Ad valorem equivalents (AVEs) for selected products and years in this period are displayed in Table 2. Clearly, for many products, the applied MFN (most favoured nation) tariff provided only a small part of the total border protection. In 1990, the list of goods requiring permission to import was abolished, and in theory all but six agricultural products could be freely imported. In fact, non-automatic, time-delimited import licenses were still required for a wide range of `sensitive' agricultural products. At the start of the implementation period of the Uruguay Round Agreement on Agriculture (URAA) in 1995, all border levies were converted to tariff equivalents and bound. Table 2 shows the applied MFN tariffs for the same selected products for 1997 and 2003, and the 4 An extra budgetary fund that was used, among other purposes, to fund export subsidies and subsidies on agricultural inputs. 5 A specific import tax, varying by product, created as a source of extra-budgetary funding following the Housing Fund Law of 1984. 10 target tariff binding at the end of the 10-year implementation period. 6 Import licences as such are no longer required for most products 7, although import approval procedures and inspection controls apply. Non-tariff barriers against red meat were intensified in the early 1990s, first by restricting imports to unprocessed meat in 1990, and then in 1992 by requiring slaughterhouses shipping red meat and poultry to Turkey to be inspected by Turkish officials (OECD 1994). Since August 1996, Turkey has operated an outright ban on red meat imports. Restricted imports of breeding cattle have been allowed since 1999, but the ban remains in force for meat, feeder and slaughter animals. This ban has been challenged on various occasions in the WTO and is officially defended on sanitary grounds, in particular with respect to BSE risks (Burrell 2005). Turkey's per capita consumption of all meat (including poultry) is about one quarter that of EU-15. Under Turkey's import substitution program of the 1960s and 1970s, exports were strictly controlled. During the 1980s, regulations pertaining to agricultural exports were gradually simplified. Export levies on high-value products like angora wool, dried fruit and nuts, for which Turkey has a large world market share, had been brought in during the 1960s with the aim of raising revenue. These levies were gradually abolished or allowed to erode in value, and had completely ceased in 1995. During the 1980s, exports of a few products (cotton, some cereals including wheat) were alternately taxed or subsidised, apparently according to domestic supply management criteria. Commodities traded by SEEs regularly received implicit subsidies as the Treasury covered their duty losses (including losses on external trade operations). The implicit export subsidies paid in this way in 2000 are estimated at EUR 8 million for barley, EUR 225 million for sugar, EUR 10 million for tea and EUR 100 million for tobacco (Grethe 2004). Other products received subsidies on an occasional basis. In addition, exporters of either unprocessed or processed agricultural products were able to receive export credits of up to 50 percent of the fob value of the consignment at interest rates often well below the rate of inflation. Turkey's URAA schedules list 44 products on which export subsidies may be given, subject to volume and expenditure bindings. In recent years, subsidies have been paid on only 16 6 Under the URAA, Turkey's tariff bindings had to fall by an average of 24 percent over 10 years, with a minimum 10 percent reduction per tariff line. Turkey opted for the minimum 10 percent reduction on many products, including a number of animal products, tea, most grains, flours and cereal preparations, a few vegetables and nuts, fisheries products, sugar and unprocessed tobacco. 7 But see the US complaint against Turkey regarding import licences for rice (WTO, 2005). 11 products, with the aim of `developing export potential'. Subsidies are paid on a fraction of each consignment, ranging from 18 percent (biscuits) to 100 percent (fish). Subsidies are not provided to exporters in cash, but rather in the form of deductions in payments to public bodies and corporations for items such as social security, insurance and telecommunications. Exporters have to present the necessary documentation to MIT, which arranges the refund with the public organisation from the central budget. In January 1996, Turkey and the EU formed a customs union. Since over half of Turkey's agricultural exports are destined for the EU, and about one third of its agricultural imports come from the EU, the arrangements for agricultural trade with the EU need to be taken into account when assessing border protection for Turkey's agriculture. Unprocessed and processed agricultural products, as well as fisheries products, remain outside the customs union. Turkey has set up tariff rate quotas for a large number of agricultural imports from the EU, but for most of them, the in-quota tariff reductions are very small. Of those products for which the in-quota tariff is significantly reduced or eliminated, fill rates tend to be over 100 percent, so these imports pay the MFN tariff at the margin. For imported processed products with an agricultural and an industrial component, the industrial component is treated in line with the provisions of the customs union, whereas the agricultural component is taxed according to Turkey's own schedule for that product. In 2001 Turkey's agricultural exports to the EU comprised products with no MFN barrier (7 percent by value), those with an MFN barrier and for which Turkey receives no preferential treatment (2 percent), those with an MFN barrier which is reduced for Turkey (36 percent), and those where the MFN barrier is completely waived for Turkey (54 percent) (Grethe 2004). Turkey continues to subsidise exports to the EU for products that are not in free circulation within the customs union. Domestic agricultural policies For decades, the main agricultural policy instruments used were price support for crop products and input subsidies. Input subsidies became heavily used in the early 1960s when the government began to promote agriculture through subsidies for credit, agricultural chemicals, 12 seeds, and irrigation. In the early 1970s, fertilizer was added to the list of subsidized inputs. By contrast, livestock production has been supported mainly by border measures. Although the rates of support on products and input use fluctuated considerably prior to 2000, there were no fundamental changes to the kind of policies and delivery mechanisms used. In 2000-2001, an ambitious program to restructure domestic agricultural policy began, involving policy re-instrumentation and reduced budget outlays. The following paragraphs describe in more detail the policies in place up to 2000. The most recent period is then discussed. Intervention buying of commodities at support prices began in the early 1930s with wheat. Other grains, cotton and tobacco were added in the 1940s. With the first Five-Year Plan (1963-67) the number of crops supported in this way rose to nine, and by the early 1970s it had reached 30. The reforms introduced in 1980 brought the number back to 17, and the share of price support in agricultural GNP declined. By 1990, only 11 commodities were covered (wheat, barley, rye, maize, oats, sugar beet, tobacco, mohair, silk cocoons and poppy seed) but by 1992, with the addition of some horticultural crops and pulses, soybeans, groundnuts, sunflower seeds and rice, the total number of crops with price support was up to 25 (OECD 1994). In 1992, the total cost of agricultural subsidies was USD 3.1 billion, with the greatest shares taken by wheat (13 percent), cotton (22 percent), tobacco (18 percent), sugar beet (17 percent), hazelnuts (9 percent) and sunflower seed (8 percent). By 1999, the total cost of commodity price support alone was over USD 4 billion (about 2.2 percent of GDP) (World Bank 2001). Intervention buying was operated by the SEEs (grains and pulses, sugar, tobacco, tea) and the ASCUs (horticultural crops, cotton, oilseeds, nuts, olive oil). Support prices were announced after planting, and payments were usually made one year or more after harvest and delivery. The short-run production incentives provided by these payments were weakened by the timing of the price announcement, and by the fact that high rates of inflation throughout the period eroded their value and exacerbated real price uncertainty. However, the sustained use of these measures over time undoubtedly distorted regional cropping patterns, supported inefficient production structures and shielded the sector from competitive forces. Once announced, support prices could not be adjusted for changes in market conditions during the growing season or post-harvest, and many did not reflect differences in product quality. SEEs had to act as guaranteed buyer of last resort, so that when purchases exceeded storage or processing capacity, they made distress sales on the world market. The statutory 13 conditions regulating TMO's operations prevented it from passing on transport, handling and storage costs when grain was resold on the domestic market. Such constraints, plus the overstaffing and lack of incentive to operate efficiently induced by soft budget constraints, help to explain the huge losses of the parastals. Controls on area planted were introduced for three commodities (hazelnuts, tobacco and tea) in the mid-1980s, under the authority of the relevant ASCU or SEE. They worked badly due to ineffective enforcement (OECD 1994). Stricter controls and compensation-backed incentives were adopted in 1994. From 1994 onwards, tea growers were also required to cut back part of their plantation each year, in order to improve the quality of the leaf. A `pruning premium' was introduced to compensate them for lost volume. Over the period 1996-2000, tea pruning payments averaged USD 17 million annually. In addition, informal area controls operated for sugar beet. By contrast, the livestock sector has had little direct intervention. Since 1986, producers delivering milk to dairies that are certified as meeting certain technical standards have received an extra payment per litre, the `milk incentive premium'. This premium is currently less than 10 percent of the domestic milk price and is paid on around 25 percent of total production. Otherwise, support for dairy products has been provided by border measures. Currently, tariffs on most dairy products are bound at 180 percent (lower for some cheeses). Applied MFN tariffs were significantly below these bindings in the late 1990s, but moved closer to bound levels in the early 2000s. Apart from temporary intervention buying of live animals in the drought of 1989, support for bovine meat has relied on border measures. In 1995, MFN tariffs on red meat stood at just 15 percent, but shortly afterwards was raised to 165 percent and are now at their bound levels of 225 percent. Since the ban on red meat imports in 1996, insulation of the domestic red meat market is virtually total. A meat incentive premium was paid in 1990-1, and again in 1994-5, per kilogram of beef and sheepmeat on animals delivered to abattoirs satisfying modern hygiene standards 8. Support to input use has been extensive. Incentives for capital investment were paid to farmers during the 1980-85 period largely in the form of reductions in customs duty on imported machinery and other tax deductions. From 1985 onwards grants were paid for various investment 8 It will be paid again in 2006 to boost low producer incomes, which have been hit by high feed prices and weak demand. 14 projects such as establishing feedlots. This form of aid ceased in 1994. MARA also funded on- farm development work (such as field levelling, soil improvements and so on), with costs averaging USD 23 million for 1986-90, USD 52 million for 1991-95 and USD million 63 for 1996-2000. This expenditure has continued at similar levels into the 2000s. Until 1999, credit to farmers was heavily subsidised, and delivered either directly by the Agricultural Bank (or to sugar beet growers through the Sugar Bank, at that time, cooperative- owned) or indirectly through the ACCs and ASCUs. The government also provided cheap credit to the agricultural input industries. Rates to farmers tended to be 40-60 percent below commercial rates, and from the late 1970s until 1998, real interest rates on loans to farmers were negative. From 1986 onwards, larger interest concessions were given for livestock production. In 1992-93, for example, the real rate of interest on loans for crop production was -16 percent, and for animal husbandry -24 percent. In 1994 the average real interest rate on agricultural loans reached -45 percent (OECD 1994; World Bank 2004b). The use of credit subsidies to agriculture peaked in the period 1994-99, averaging over USD 1.3 billion per year. And yet, at the end of the 1990s, the ratio of agricultural lending to agricultural GDP averaged only 14 percent, well below the 30 percent typically observed in comparable countries (World Bank 2004b). The Turkish Competition Authority has estimated that two-thirds of small farms were using informal credit from illegal brokers. World Bank (2004b, p.23) noted that from the mid-1990s cheap and abundant credit encouraged credit delinquency. It has been estimated that only 80 percent of the implicit subsidies reached farmers, due to the high administrative costs and inefficiency of the delivery agencies 9. As mentioned previously, subsidies for the domestic production and consumption of fertilizers began in 1961. Until 1986, ex-factory and farmer prices were controlled by the state at levels above and below the world market price respectively (Niron 1986). Fertilizer distribution was in the hands of TZDK and (for sugar beet growers) TSFAS, and their duty losses were met from government funds. In 1986, all restrictions on fertilizer imports were lifted, and prices were determined competitively in domestic markets, albeit still protected by tariffs. TZDK's share of 9 World Bank (2000) cites a claim by the Farmers' Association (TZOB) in 1997 that one in three or four borrowers of subsidised agricultural credit might not be farmers at all. 15 the fertilizer market fell from 90 percent in 1985 to 10 percent in 1992 (OECD 1994). At the end of the 1990s, 35 percent of the fertilizer sold to farmers was marketed by ACCs. From 1986 onwards, fertilizer subsidies were paid by government to fertilizer distributors via the Agricultural Bank. For a brief period (1994-97), subsidies were paid direct to farmers, upon presentation of the sales invoice, but this was reversed due to the scheme's administrative burden and susceptibility to fraud (World Bank 2004b). During 1990-97 annual expenditure on fertilizer subsidies averaged USD 363 million. The fertilizer subsidy was 39 percent of the market price in 1993, and 50 percent in 1997. The ratio of farmer price to import price over the period did not reflect these high rates of subsidy, except during the period 1996-97 when subsidies were paid direct to farmers. In fact, during 1990-95, the farmer/import price ratio fluctuated between 78 and 108 percent. In 1997, the government began phasing out the fertilizer subsidy, and it ceased completely at the end of 2001 (World Bank 2004b). Agriculture's use of pesticides has been supported in two ways. First, the government assumes the cost of protective measures taken when epidemic crop diseases or pest infestations occur. Second, from 1987 onwards the Agricultural Bank was authorised to pay a rebate of 20 percent on the value of pesticides bought by farmers themselves. Over the period 1996-2001, annual disbursements by government on this item averaged 26 million dollars. Starting in 1985, a subsidy was paid to certified producers of hybrid maize, hybrid sunflower, soybeans and nitrogen-fixing bacteria (OECD 1994). Total payments under this scheme fell during the 1990s from their peak of USD 31 million in 1987 to low levels in the early 2000s. Subsidies have also been paid to farmers, at various times, for seeds and animal feed. Ownership and exploitation rights for water are vested in the state. There has been considerable public investment in irrigation since the 1960s, expanding the irrigated area by about 800 thousand hectares per decade. More than 4.9 million hectares are now irrigated. Irrigation schemes are under the responsibility of the General Directorate of State Hydraulic Works (DSI) for large-scale schemes, and MARA for smaller schemes. During the review period, the management of the smaller schemes was generally transferred on completion to village cooperatives. By contrast, at the start of the 1990s, DSI was operating most of its schemes itself and attempting to recover operating costs, together with a contribution to capital 16 costs, from farmers. It is difficult to obtain data on the full cost of providing irrigation, or the amount of subsidy that farmers have received over the years in this respect. OECD (1994) described how farmers' payments for irrigation use were fixed too low in money terms, were further eroded by inflation and then undermined by poor recovery rates. 10 In 1993, DSI began transferring operation and maintenance of its schemes to Water Users' Associations (WUAs), semi-public bodies that are set up under Municipality Law with leadership consisting of elected members and local officials. Most of the 1.6 million hectares transferred by DSI to water users' groups have gone to WUAs. After transfer, WUAs are expected to recover all operating and maintenance costs from farmers. Where such management transfers have taken place, operation and maintenance costs have fallen by about 40 percent, and yields from irrigated land are reported to have risen by 60 percent (World Bank 2004b). Agricultural policy reform, 2000 to 2005 In 2000, as part of the seventh Five-Year Development Plan, the Turkish government adopted an ambitious program of agricultural policy reform. It involved not only the reinstrumentation of policy and a change of policy delivery systems, but also the dismantling or fundamental reform of failing agricultural sector institutions. Price-fixing by government was discontinued. Product and input subsidies were phased out, and replaced by direct income support (DIS). SEEs were to be restructured and privatised, and ASCUs would become financially autonomous member- controlled cooperatives. To underpin this reform, a World Bank loan of USD 600 million was secured to help fund DIS payments and incentive payments to shift farmers out of producing surplus commodities (World Bank 2001). Activities under this loan agreement constitute the Agricultural Reform Implementation Project (ARIP). The initial impacts of the reform were substantial. On the output side, real agricultural prices fell by 13 percent between 1999 and 2002 alone, and by 22 percent relative to non- agricultural prices (World Bank 2004b). Surprisingly, price falls were greater in the livestock 10 OECD's PSE database shows subsidised electricity and water in connection with irrigation schemes varying between about USD 8 and 33 million over the period 1986-96, between about USD 40 and 60 million between 1997 and 2002, 17 sector, which previously received virtually no price subsidies, than for crops. This is a reminder that not all the price fall was due to agricultural policy changes: demand for livestock products was more vulnerable to the deep economic recession of the early 2000s, because of the closed markets for these products and their higher income elasticity of demand. Subsidies for fertilizer and pesticides disappeared in 2001 and 2002 respectively. The phasing out of credit subsidies was completed by 2002. Table 3 indicates that, after a temporary drop, the aggregate use of fertilizer and credit rebounded to or exceeded previous levels. Farm incomes fell by 16 percent between 1999 and 2002, 11 four fifths of which was due to subsidy removal and the remainder to a 4 percent reduction in output (World Bank 2004b). The DIS payment was not intended to compensate producers fully for price cuts, nor to relieve rural poverty, but rather as a transitional measure to cushion the immediate impact of reform on farm incomes. It has been estimated that, on average, DIS payments compensated farmers for about half their short-term income loss (World Bank 2004b). DIS is inadequate to compensate the income loss of many intensive fruit and vegetable producers, whose area is small and who have lost disproportionately on fertilizer and pesticide subsidies. However, DIS compensation has been more adequate for farmers in the poor regions of the East and Southeast, where pre- reform incomes and subsidised input use were lower and farm sizes tend to be larger (Bayaner and Bor 2006). Table 4 gives an overview of the scheme up to 2005. Implementation of the DIS program has been slowed down because the national land registry is not complete in rural areas. To implement the scheme, MARA has built up its own Farm Registry System (FRS). Not all the 17 million hectares registered in the FRS is covered yet in the national land registry. Inability to prove ownership of land, and disputes concerning ownership, are given as reasons for the incompleteness of both these registration systems, and the resulting incomplete coverage of the DIS scheme. The 50-hectare ceiling for payments has led to some larger farms being divided amongst family members (Bayaner and Bor 2006). 12 Thus, this relatively decoupled measure whose distortionary potential was considered to be small may nevertheless introduce unexpected secondary distortions. thereafter falling to under USD 2 million for 2003 and 2004. 11 By US$2.7 billion, or $1.25 billion after payment of DIS (World Bank 2004b). 12 Since farms above 50 hectares are required to keep books and be assessed for income tax instead of paying a flat rate, this incentive already existed. 18 Hard budget constraints were imposed on the SEEs and ASCUs. TMO now functions as a commercial organisation in the grain sector: its annual price declarations are based on domestic and world market expectations, and are independent of government. Real producer prices for cereals had been falling since their peak in 1996, but this trend appears to have bottomed out in the post-reform period. At the start of the reform, sugar and tobacco support prices both fell by 20-25 percent. Since the early 2000s, prices for these commodities have been set by agreement between growers and processing factories. For sugar beet, a quota scheme similar to that of the EU, with A, B and C quota (for export only, amounting to 1-2 percent of production), was introduced in 2001: quota rights are allocated to factories, and transferred to individual farmers. Sugar beet deliveries fell by 25 percent in the first two years of the reform, and have fluctuated since then. Tobacco area restrictions are more strictly enforced than hitherto, which, together with the price falls, has brought production down to less than half of its 1993 peak of 340 thousand tons. ÇAYKUR now has a reduced role as a tea-processing enterprise. Despite the intention to discontinue the tea-pruning payment, it was stopped for the 2004 crop only. Prices for cotton, sunflower seed and soybeans were less affected as price premiums continued for these products, although it was initially intended that they too would be removed (World Bank 2001). Restructuring measures have been less successful. The ARIP incentive scheme to shift farmers out of surplus crop production was applied to the tobacco and hazelnut sectors. In both cases, uptake was poor. Tobacco farmers had already reacted to price falls by switching 60 thousand hectares out of tobacco in 2000-1, before the ARIP compensation payments became available. Hazelnut producers have been reluctant to grub up established orchards and less than 1 percent of the target 100 thousand hectares was abandoned (World Bank 2004a). In the second instalment of the ARIP loan, the scope of funds earmarked for farmer transition was broadened to cover village-based participatory investments under the new Rural Development plan (e.g. small-scale cooperative processing, storage, village cooling facilities), capacity building for farmers' organisations, and farmers in environmentally fragile areas who want to switch out of crops in over supply. Privatisation of some of the largest SEEs has been slow. For example, by the end of 2006, of the five original units of TEKEL (alcohol, tobacco and salt), only the alcohol enterprise had been sold, and the cigarette enterprise had gone for tender twice unsuccessfully. Only two of TSFAS's initial 27 sugar factories had been privatised, with another three included in the 19 privatisation program for 2007. The infant private sugar processing sector had just six factories, operating with more modern technology and flexible marketing strategies. By contrast, the factories remaining in the state sector have old technology and low capacity. With the expected expansion of the private sector, it might be rational simply to scrap some existing TSFAS capacity. However, the least viable production units tend to be in the poorest areas, and their disappearance would have a high social cost. This situation exemplifies some of the difficulties encountered by attempts to reduce distortions by dismantling an outdated state sector. The early budgetary impact of this reform program was spectacular: a net cut in agricultural subsidies (after DIS payments) of 2.3 percent of GDP (USD 5.5 billion) (World Bank 2004b). As expected, it has not been possible to sustain this rate of cut as coverage of DIS payments has increased, and the financial burden of cleaning up the parastatals is still continuing. In its assessment of the ARIP loan, the World Bank (2001) identified two sets of concerns related to the project's aims. Regarding the transition of the ASCUs to fully fledged market-oriented trading companies, it was uncertain whether farmer members would want to take over control, and whether some ASCUs might prove unfit to survive in competitive markets. As of 2006, all 16 ASCUs were still operating commercially and independently of government, with ASCU directors appointed by owner-members. Over-staffing was lower, and existing ASCU debt was still being rescheduled in exchange for restructuring compliance. The World Bank appraisal also identified as `substantial' the risk of a `political backlash' as long-standing support was withdrawn from well-established client groups, creating `pressure for reversal or non-implementation of measures, or starting new subsidy schemes'. The reform program has entered a phase where the risk of such pressures accumulating is higher, and recent developments are not wholly encouraging. As already mentioned, price premiums are still paid on cotton, oilseeds, and olives, and are scheduled to continue until 2010. Maize started receiving a premium in 2006, and other payments based on output were introduced for cereals in 2007. In 2004, some concessional credit became available again (about USD 30.5 million in 2004), albeit under strict conditions that target producers aiming for higher quality output, such as those producing according to EurepGAP protocols or using higher quality livestock breeds). In 2006, a new insurance subsidy was introduced, which is open to all producers regardless of commodity produced. Under this scheme, the state reimburses 50 percent of insurance premiums direct to the insurance providers. In 2006, the scheme was operating in just 90 of the poorest districts, out of a 20 total of 950 districts. The Agricultural Strategy adopted at the end of 2004 as part of the next Five-Year Development Plan appeared to re-couple part of the DIS payment to particular schemes and targets, although details were not given; the `fertilizer payments', based on land area with rates varying by product groups that were introduced in 2007, are in line with this intention. As of May 2006, the rate of DIS payment for 2005 had still not been announced, which suggests a falling behind with this commitment to farmers that may harden attitudes against continuation of the reform. It is unclear whether these departures from the general principles of the reform might be first signals of the `backsliding' that the World Bank identified as a risk. Estimates of distortions since 1961 Inflation of the Turkish Lira (TL) averaged about 40 percent per annum during 1960-2004, which hampers interpretation of nominal values. Monetary amounts reported in this section have been converted to constant 2004 values using the GDP deflator. In January 2005, Turkey reformed its currency, replacing the Lira (TL) by the New Turkish Lira (YTL), with a conversion rate of 1 YTL= 1,000,000 TL. To facilitate interpretation, figures in constant 2004 values are expressed in their YTL equivalent, although the YTL was not the official currency until 2005.13 Annual producer and border prices, and NRAs, are reported in the Appendix, with the FAO being the source up to 1985 and the OECD thereafter. Import protection and support for agricultural and other products First, we give a snapshot of recent average protection rates for agrifood products, relative to other sectors. In 2004, average trade-weighted MFN tariffs were 40 percent for primary agricultural products 14, 18.9 percent for processed food products and beverages, 1.9 percent for other primary products, and 3.2 percent for manufacturing (Togan 2005, Table 5). In 1996, on forming a customs union with the EU, Turkey aligned its tariffs for non-agricultural commodities 13 The official dollar exchange rate for 2004 (2005) (annual averages) were USD 1=YTL 1.426 (1.341). 14 For all agricultural products (WTO definition), the average trade-weighted MFN tariff was 21.4; the average applied tariff for agricultural products from the EU was 19.8, and for other preferential trading partners 21.2 (Togan 2005). 21 with the common external tariffs of the EU, which have been stable over the period at very low levels. Thus, for the last 10 years, there has been very little offsetting protection in other sectors to take into account when assessing the distortions in the agricultural sector due to agricultural tariffs (see also Figure 5). We next examine intersectoral distortions as captured by the PSE and TSE estimates, which have been calculated by OECD for Turkish agriculture from 1986 onwards. The PSE represents the transfers from consumers and taxpayers to agricultural producers due to agricultural policies, and comprises market price support plus all direct payments from government to farmers. It is derived by calculating PSEs for individual products, which are then aggregated. The standard PSE products account for less than 50 percent of Turkey's agricultural output by value (Kasnakoglu and Cakmak 2000). To reflect the greater share of field crops and horticultural products in Turkey's output mix, potatoes, tomatoes, tobacco, grapes, apples and cotton are added into the calculation of aggregate market price support, which is then `grossed up' to represent market price for the sector as a whole. This extended product selection covered about 58 percent of output in the late 1980s, and thereafter about 63-64 percent. 15 The Total Support Estimate (TSE) consists of the PSE (Producer Support Estimate) and the GSSE (General Services Support Estimate), the latter being non-commodity-specific transfers to the agricultural sector as a whole, not accruing directly to farmers. Figure 3 shows that total agricultural support has fluctuated in line with macroeconomic cycles. The increase after 1988 reflects the shift in income distribution towards workers and farmers as Turkey emerged from the economic restraints set up after the 1980 crisis. The sharp falls in 1994 and 2001 correspond to macroeconomic crises, which were characterized by strong currency devaluations (63 and 49 percent respectively against the US dollar) and negative real income growth (-4.7 and ­7.5 percent respectively). In most non-crisis years, the PSE accounted for 80 percent or more of total support (Figure 3a). However, for 1995-2001, the GSSE took up over 30 percent of the total. The GSSE involves transfers whose aim is to improve the functioning of the sector. Measures include investments in research and development, agricultural schools, infrastructure, marketing and promotion, and public stockholding. For the OECD countries as a whole, these 15 Although OECD member countries in Europe and Oceania have a coverage well above 70 percent, the current PSE coverage for the USA and Japan is on a par with that of Turkey. 22 measures contribute almost 20 percent of total support. These are considered relatively benign transfers whose potential for distortion is undefined but thought to be low. By contrast, in Turkey, the GSSE has consisted largely of bail-out payments to the SEEs and ASCUs. During 1995-2002, these payments never fell below 85 percent of the GSSE, and over the same period they averaged one third of total support. Even since the 2000-1 reform, the cost of sanitizing these organisations continued to require considerable transfers. How distortionary these transfers to SEEs and ASCUs were, and whose incentives they distorted, is difficult to discern. To the extent that duty losses arose because SEEs were required to act uncommercially (e.g. absorb transport costs, accept all produce offered, etc.) in order to assist farmers, then their losses amount to hidden producer subsidies. However, the parastatals' soft budget constraint and recurrent political interference in their functioning encouraged overstaffing and inefficiency (Olgun, 1991; Demir, 2002), and undoubtedly acted as a catalyst for rent-seeking and non-market-oriented behavior in the food supply chain. Moreover, the size of the transfers in the later 1990s meant that they were not neutral with respect to the macroeconomy. Thus, the distorting impact of this component of total support in Turkey has not been confined to agriculture, but it probably distorted agricultural incentives less than the total amounts would suggest. The low farm support levels in 1994 and 2001 occur primarily because market price support was squeezed from above by falls in domestic producer prices, rather than by sharp increases in border prices due to currency devaluation or by marked changes in border protection. This raises the question of what is actually measured by this price gap and what factors determine it. The evidence suggests that, in these crisis years, domestic prices may have been affected by lower demand due to reduced incomes (certainly the case for livestock products, which have high income elasticities of demand) or that payments to farmers were reduced, whether to soften the impact of the crisis on consumer food prices, or because of cash flow problems of the parastatals. If the squeezing of the price gap is in fact due to internal demand phenomena or to slippage between announced policy measures and implementation, it is striking and informative that such large internal price movements were not corrected by price arbitrage between domestic and foreign markets. Whatever the case, one must be cautious in interpreting these gaps and their changes as the sole result of official policies and under the control of policy makers. 23 Relative to GDP, the peak in support in the early 1990s was slightly higher (7.3 percent) than the one occurring in 1998-99 (around 7 percent) (Figure 3b). However, it was the latter that triggered fundamental policy reforms. During the 1990s, there was a shift in the relative shares of support coming from (less `visible') market price support and from the government budget (consisting of more `visible' direct subsidies to farmers plus general support to the industry). In 1998-9 the latter reached historic levels both in real terms and as a percentage of GDP, 16 and accounted for more than half of the total support. 17 Given the government's severe deficit financing problems, the situation was unsustainable. The need to control agricultural spending was yet another trigger for the urgent reform of public finance that began in 1999. Nominal rates of assistance (NRA) for agricultural commodities In accordance with the methodology as defined in Anderson et al. (2008), Nominal Rates of Assistance (NRAs) have been calculated for key products, based on price gaps. The price gap for an unprocessed commodity at the farm level is defined as the domestic producer price at farm level, plus all transport costs, handling costs and marketing margins incurred in getting the domestic product to the port, minus the relevant fob or cif price of the exported or imported equivalent product. Production subsidies and support granted to intermediate input use, such as fertilizer etc. are also taken into account. The price gap thus calculated is assumed to pick up the effect of all tariff and non-tariff barriers associated with an imported product, or export subsidies and other aids for an exported product. The resulting product-specific NRA represents the price gap expressed as a percentage of the border price. An NRA reflecting the protection or taxation of a country's total agricultural sector is calculated via a weighted average of the individual products' NRAs and of the `guesstimated' NRA for the remaining commodities' aggregate, using as weights the undistorted gross value of production. The NRA for total agriculture also includes non-product-specific support payments. 16 In 1999, it was 3.5 percent of GDP. For comparison, we note that in 2005 Turkey's total educational spending was 3.5 percent of GDP, when over 20 percent of the population was between 5 and 15 years of age. 17 As a percentage of total expenditure from the consolidated budget, agricultural spending was 44, 37 and 42 percent in 1997, 1998 and 1999 respectively. 24 Due to the government's regulation of the foreign exchange market, a black market rate arose parallel to the official rate. In the 1960s the black market premium amounted to about 45 percent, but it declined subsequently and after the mid-1980s remained at very low levels. In order to account for these irregularities, an equilibrium exchange rate was calculated as the weighted average of the official and the black market rate, with weights based on an estimate of the proportion of foreign currency that is sold on the parallel market (see Anderson et al. 2008). The exchange rate distortion as a part of the protection to import-competing goods is then the difference between the exchange rate the importer faces and the equilibrium exchange rate. With regard to exportable products, the exchange rate distortion is measured by the difference between the exporter's exchange rate and the equilibrium rate. The data for calculating the individual product's NRAs stem from two sources. For the period 1961-85, all information on prices and quantities is taken from the FAO agricultural database (FAOSTAT). For this time period no data on distortions with regard to intermediate input use, non-product-specific payments or production subsidies was available. Thus, NRAs for 1960-85 reflect a pure price wedge between the domestic producer price and the border price. Information for the remaining time period, 1986-2005, comes from the OECD's PSE/CSE database, and so output subsidies and payments based on input use are included in the calculated product-specific NRAs. More precisely, the latter include payments based on the use of variable inputs (fertilizer, hybrid seed, pesticides, etc.) and support coupled to the use of on-farm services and investments. Non-product-specific subsidies such as payments based on area planted and animal numbers, historic entitlements, input constraints, overall farm income and finally, miscellaneous payments are included only in the NRA for the total Turkish agricultural sector. Rice and hazelnuts are the only products not covered by the OECD PSE/CSE database. The data for these two commodities come from FAOSTAT even for the post-1985 period. The information on data used for the calculation of the equilibrium exchange rate are obtained from Easterly (2006), which provides the official exchange rate as well as the black market premium for various countries for the period 1960-2005. In the absence of other information, the proportion of the foreign currency sold on the parallel market is assumed to be 50 percent in the years when the market was active because the official exchange rate was overvalued. 25 Nominal rate of assistance for non-agricultural commodities For the purpose of intersectoral comparative evaluation, an NRA for the total non-agricultural tradables sector has been calculated in addition to the NRA estimate with regard to overall agriculture. This indicator is determined via NRA guesstimates for the import-competing, exportable and non-tradable parts of the total non-agricultural sector. The exchange rate distortion applying to importable and exportable goods ­ as explained above ­ is then added to the corresponding NRA guesstimate. The data on non-agricultural distortions come from various sources. For the time period 1960-1983, they were taken from the precursor of this study which was conducted by Krueger, Schiff and Valdés (1991) (KSV). KSV used the average tariff on non-agricultural products as an approximation for the distortions to the non-agricultural import-competing sector, while for exportables and non-tradables (services) an NRA of zero was assumed. For the 1986-2005 period, further tariff information was taken from the World Bank's World Development Indicators (WDI). Since a complete time series of tariff data is not available, extrapolation was used from one data point to another. The division of the total non-agricultural sector into importables, exportables and non-tradables is based on assumed shares for importables and exportables, while for non-tradables we used the share of services in GDP. Relative rate of assistance (RRA) and the trade bias index (TBI) The Relative Rate of Assistance (RRA) provides an indication of the assistance to agriculture relative to the rest of the economy. As explained in Anderson et al. (2008), the RRA compares the total NRA for tradable agriculture with the total NRA of the tradable non-agricultural sectors. 18 A negative RRA indicates that agriculture is either less subsidized or more taxed than non-agricultural industries, suggesting an anti-agricultural bias is in place, and conversely for a positive RRA. The Trade Bias Index (TBI) is calculated by comparing the NRA for agricultural 18 The Relative Rate of Assistance, RRA, is defined as: RRA = 100[(1+NRAagt/100)/(1+NRAnonagt/100) ­ 1], where NRAagt and NRAnonagt are the average percentage NRAs for the tradables parts of the agricultural and non- agricultural sectors, respectively. 26 exportables with the NRA for agricultural import-competing products. 19 Typically, this has been negative, reflecting a trade policy bias against agriculture. Our calculated NRAs fluctuate considerably over the period. When the only support is via border measures, and providing horizontal and vertical price transmission is good, the price gap represents the tariffs and non-tariff barriers. In countries like Turkey with less well integrated markets and poor marketing and processing infrastructures, the component representing distribution costs and marketing margins for farm products may be substantial. If this component is underestimated, part of these costs will be confounded with the price distortion due to border measures. However, these margins are probably relatively constant or slowly trending, and it is unlikely that they have generated the sharp swings observed in the calculated price gaps for Turkey. The domestic price-fixing mechanisms used by the Turkish government until 2000 could be expected to distort the relationship between domestic and border prices, if administered prices diverged from world market price trends. The extent of this distortion would depend on the degree of monopsony exercised by the state-controlled purchasing bodies (SEEs and ASCUs), and whether their market share varied markedly between years. If state purchasing bodies had only weak monopsony powers, then either the link between domestic and border prices would be quite strong (assuming good transmission between the border and domestic markets) or domestic prices would reflect internal market conditions (assuming weak price transmission). Turning to the data, the price gap instability comes at least as much from producer prices as from border prices, and for the most part is unrelated to any changes in tariff levels. For the export commodities tomatoes and grapes, there is considerable movement in both border and producer prices, and the price gap fluctuates on both sides of zero. For wheat, oilseeds and poultry, the domestic price shows some response to border price in some sub-periods, but even then the response is weak and with a variable lag. Less surprisingly, for beef and sugar, movements of the two prices are unrelated. Little, and then no, foreign trade in beef meant complete isolation of the domestic market for this product, whereas for sugar there was little foreign trade and the first round purchasing conditions were close to a perfect monopsony. In general, transmission from border price to producer price is poor. It is unclear how much this is 19 The Trade Bias Index is defined as: TBI = 100[(1+NRAagx/100)/(1+NRAagm/100) ­ 1], where NRAagm and NRAagx are the average percentage NRAs for the import-competing and exportables parts of the agricultural sector. 27 because policies themselves have acted so as to effectively cut the link with border prices, and how much it is due to exogenous factors like poor but variable price arbitrage within the country. Figure 4 shows the development of the aggregate NRAs of import-competing, exportable and total products covered by this study. From 1961 to 1979, agricultural interventions experienced a steadily increasing trend, which is in line with the pattern of interventions in agricultural prices described above. In 1961, when the government's First Five-Year Plan was adopted, increased intervention in agricultural pricing followed. In addition to price interventions the government also expanded the coverage under the support purchase program and increased subsidies on inputs. From the beginning of the 1970s, subsidies to users of fertilizer were implemented. Within the 1970-74 period, import-competing agricultural sectors changed from being slightly taxed to being protected. This protection peaked with an NRA of 20 percent in the 1975-79 period. The taxation of agricultural exports, shown by a negative NRA in Figure 4, also decreased during the 1961-1974 period, while in the following period a stronger negative trend can be observed again. The implicit taxation on agriculture as a whole was approximately 20 percent in the 1960s but decreased to 7 percent during the 1970s. The increase in taxation in the following five-year period (down to a rate of about 30 percent for agriculture as a whole) is predominantly the result of the acute macroeconomic stabilization phase between January 1980 and July 1981. As already mentioned, this period was characterized by a heavy currency devaluation and the reduction of input and production support programs. However, the pro- agricultural policy that emerged after November 1981 reversed that trend so that in 1985-89 agriculture experienced positive assistance again: NRAs for import-competing commodities averaged 28 percent while exportables were only slightly taxed (by 8 percent). Protection for agriculture overall averaged 4 percent, and increased to just over 20 percent throughout the 1990s and into the present decade, before rising in 2005 to 26 percent. The development of NRAs for individual commodities is shown in Table 5. As described above, total agriculture was taxed from 1961 to 1980. The products subject to the highest levels of taxation were maize, tomatoes, tobacco, hazelnuts and apples (36-45 percent). Other crops such as sunflowers, barley, eggs and cotton also experienced negative NRAs but to a lesser extent (19-24 percent), while wheat and grapes show the lowest taxation (6-15 percent). The only products that received protection during this time period were potatoes and milk (4-41 percent). 28 Furthermore Table 5 shows that, over the period 1985-2005 as a whole, NRAs for dairy, beef and sugar were the highest (above 70 percent), followed by barley, eggs, poultry and potatoes (42-52 percent), and sunflowers, maize and wheat (32-39 percent). Finally, tomatoes, grapes, tobacco and sheepmeat received the lowest levels of support (5-20 percent). With some low positive or negative NRAs, the export crops apples, hazelnuts and cotton were subject on average to taxation (6-46 percent) over the 1985-2005 period. 20 The NRAs for beef, potatoes, eggs and sugar have risen significantly, while protection and taxation for dairy and apples have fallen. Table 8 summarizes the overall sectoral results with respect to agriculture and non- agricultural sectors. As previously explained, we follow the OECD's assumption that agricultural products not covered by this study were subject to the same level of protection or taxation as the commodities analyzed individually. Thus, for the years where no non-product-specific supports were in place, the NRA for the total agricultural sector is identical to the NRA of the covered products. As Table 8 shows, agriculture received non-product-specific support in the 2000-05 period, reaching 5 percent in 2005 (following the adoption of the DIS in 2001). Accordingly, for this period the NRA for agriculture as a whole (including product-specific and non-product- specific measures) is higher than the total NRA for the commodities covered. Table 8 also reports the trade bias index (TBI) for Turkey, which is negative over the whole period whilst showing quite erratic movements. In the years from 1961-84, the reason for this steady anti-trade bias was the heavy taxation of agricultural exports whereas imports were either taxed less or slightly subsidized. The TBI remained strongly negative over the remaining periods even though taxes on exportables decreased significantly and even turned into subsidies. The negative trend in the TBI persisted because from the mid-1980s importables also became increasingly protected, so that exportables were still discouraged relative to import-competing products. With regard to non-agricultural industries, Table 8 shows a high level of protection particularly in the period 1961-79. The peak in the later 1960s reflects the very high tariffs on importable non-agricultural products. Thereafter, non-agricultural protection declined gradually, and almost reached a level of zero in 2000-05. The relative rate of assistance to agriculture 29 (RRA) compares the levels of taxation or protection of the agricultural sector with that of the overall non-agricultural sector. Table 8 and Figure 5 show a negative RRA of about 50 percent in the 1961-84 period. During this time not only was agriculture heavily taxed directly but also indirectly via high non-agricultural protection via significant import tariffs. Then, with non- agricultural protection coming down and agricultural taxation declining, the RRA became virtually zero by the end of the 1980s, reflecting an equal rate of assistance to the agricultural and non-agricultural sectors. The trend towards stronger support to agriculture led to a positive RRA in 1990-94. For 2000-4, the RRA averaged about 20 percent, and reached 30 percent in 2005. For the period from 1986, information on product-specific input subsidies and non- product-specific support measures is available. Figure 6 depicts NRAs for importables, exportables and total agriculture including and excluding product-specific input support. This reveals that subsidies to intermediate input use were higher for import-competing products than for export goods. Fertilizer initially took the highest share of product-specific input support, but in line with the decision to reduce fertilizer subsidies this share steadily decreased and concessional loans became the input support measure with the highest share. As Figure 6 shows, with the start of the ARIP reform in 2000, NRAs including and excluding product-specific input subsidies converge. This is a reflection of the reduction in coupled production support under the ARIP reform program. In 2001, the DIS began and more than replaced the direct product- specific support of the late 1990s. Other studies measuring distortions to agriculture To complete the picture on protection rates, we refer to the work of Togan, Bayaner and Nash (2005). They calculated NRPs and effective rates of protection (ERPs) by aggregated commodity groups and industries for 2002, using a 1996 input-output table. The main conclusions from their analysis are, first, that NRPs and ERPs are relatively close for primary agricultural production, which is consistent with the high share of non-traded items in the cost structure of Turkish agriculture, and second, that for the food processing industries, NRPs and ERPs diverge 20 Exportable surpluses are about 4 percent, 11 percent and 60 percent per year of production for apples, tomatoes and cotton, respectively, but export subsidies were used for tomatoes and apples for a few years in the late 1990s 30 considerably, with ERP usually exceeding NRP. This is what would be expected given the relatively high rates of protection on many of these processed products 21 and the much lower rates of protection on other inputs, and assuming relatively small cost shares of raw materials. The precursor of the present study, conducted by Krueger, Schiff and Valdés (1991) (KSV), examined the evolution of distortions to agriculture in 16 countries over the period 1960- 83. Turkey was part of this study, in which seven agricultural commodities were covered. The Nominal Rates of Protection (NRPs) obtained by KSV represent an index comparable to the NRA , although their study is based on a slightly different methodology and different data. One main methodological difference is the distinction between direct and indirect interventions. KSV defined effects that are due to agricultural policies (e.g., price controls, border measures, production subsidies) as direct effects and those resulting from non-agricultural and economy- wide policies as indirect effects. Another issue that is handled differently is the exchange rate used to convert border prices into local currency. In the present study, following Anderson et al. (2008), an equilibrium exchange rate is determined and exchange rate distortions are included in each product's NRA. Consequently, our commodity specific NRAs are not identical to those estimated by KSV. However, the overall trends in agricultural protection and taxation show the same pattern in both studies. Finally, the OECD's percentage Producer Support Estimate (PSE) offers another indicator comparable to the NRAs resulting from this study. As already mentioned above, for the 1986-2004 time period the price and quantity data used to calculate NRAs for most products were taken from the OECD's PSE/CSE database. Nevertheless, although the same data were used, neither the commodity specific NRAs and PSEs nor the NRA and PSE for total agriculture, are absolutely identical. The reason for the discrepancies lies again in the methodological approach. In contrast to Anderson et al's (2008) methodology, the OECD converts the US dollar border prices at the official exchange rate and thus does not consider the influence of the parallel market rate. In addition, the PSE is expressed as a percentage of distorted prices whereas our NRA is expressed as a percentage of the undistorted border price. only. 21 Togan (2005) gives averages for 2005 of 75 percent for products made from meat, fish etc, 93 percent for sugar and sweets, 31 percent cereals products, flour, etc, and 534 percent for processed fruit and vegetables. 31 Distortionary impact of policies In an environment with variable rates of double-digit inflation, imperfect or incomplete markets and high levels of uncertainty, it is unclear to what extent the price information received by farmers reflects government policy intentions. Moreover, when policies change from year to year as in Turkey, short-run supply responses are likely to be very cautious. Indeed it is difficult to reconstruct a rationale that would have deliberately led to the pattern of support price changes shown in Table 6, where a shaded cell represents a fall in real price in order to emphasise the erratic pattern of real price movements for individual commodities. Presumably risk-averse farmers reacted to such a climate of uncertainty by adopting risk- minimising strategies such as under-use of variable inputs in the short term and under-investment in the medium term. Investigating the supply response of Turkish wheat producers during the period 1960-2002, Bayaner and Bor (2006) found that output price had no explanatory power in their model, but that lagged gross wheat revenue and current fertilizer price had the expected effects, although elasticities were low. This supports the idea that output support in Turkey probably raised the medium-term trend level of output, but that annual supply fluctuations (to the extent that they are controlled by farmers and not caused by weather) are a response to changes in variable input prices. Table 7 summarises the various policy related influences affecting Turkish agriculture over recent decades. Statistics such as the NRA are designed to measure distortions of type 2 in contexts relatively free of other interference. The NRA also includes the effects of exchange rate policies (type 1). However, the effect on incentives of distortions of type 3 are hardly captured by our measures, or ­ if contained therein ­ are difficult to disentangle from those of formal policy measures. The exogenous influences of type B may also be picked up to an unknown extent by these measures. Furthermore, the degree to which those exogenous influences have themselves distorted what was intended by policy makers when legislating policies is unknown. Distortions of type 4 arise from the way support is distributed within agriculture, which is not picked up by our sector-averaged measures. Poverty is a feature of the lower end of a (usually highly skewed) income distribution. The distribution of support amongst Turkey's agricultural producers has been very unequal. In the early 2000s, the two-thirds of producers with less than 5 hectares of land used only 22 percent of the land (30 percent of the irrigated 32 land), 35 percent of the tractors, had less than 45 percent of the sheep and little over 50 percent of the cattle (Cakmak 2003). Larger farmers used more inputs, and until the abolition of input subsidies, they had better access to subsidised inputs or subsidised credit to obtain them. Consequently, yields were and remain significantly higher on larger farms, and support delivered through market price still favors the larger, richer farmers. Now that most subsidies have been replaced by DIS, paid per hectare of the land holding, a bias still exists in favour of larger farms although there is a cap of 500 decars in the DIS system aimed at limiting the extent of payment concentration. Historically, poverty has given farmers an incentive to leave farming and rural areas to provide cheap labor for other sectors of the economy. However, the transferable human capital of Turkey's farmers is very low, and unemployment in other sectors is high due to long-term under- investment in job creation, effectively narrowing this escape route. When considering the impact of agricultural support on poverty levels within agriculture, it should be remembered that the transfer efficiency of agricultural support policies to the sector as a whole has been low. The delivery systems used for transfers in the past created opportunities for a high rate of `leakage', and this still continues to a lesser extent. Therefore, measured support and transfers over-estimate the amounts that actually reached farmers as a whole, even before their distribution within agriculture is considered. Consumers have been the source of 70-80 percent of the transfer to agriculture, and this burden is particularly regressive. In the late 1990s, the share of average household expenditure on food, beverages and tobacco was 36 percent for Turkey as a whole, but for the poorest 20 percent of households, it was 45 percent in urban areas and 56 percent in rural areas (Cakmak 2003). The tax burden of the direct transfers to farmers and agricultural parastatals from government revenue also falls heavily on the middle- and lower-income groups. It is true that many of the lowest income workers are in the informal sector and therefore do not pay direct taxes. However, the tax take from the highest income groups in Turkey is also disproportionately low for various reasons. The food consumer tax equivalent (CTE) of Turkey's agricultural policies has been very high for a developing country. Turkey's average CTE of just over 25 percent in the 1990s (see Table 9) was similar to the EU's, but since then the EU has moved much more from price 33 support to direct income payments for farmers. Thus by 2005 the CTE was just 17 percent in the EU-25 whereas it was 27 percent for Turkey. Taking all these factors into account, one cannot assume that reforming Turkey's agricultural policies in a way that lowers prices received by farmers must have a commensurate, positive impact on poor food-consuming households. For this to occur, price reductions have to be transmitted along the supply chain to consumers. The relevance of this warning to Turkey's situation is illustrated by the case of milk. Turkish farmgate prices for milk for most of the period 1990-2001 were about 75 percent of the EU milk price. At the same time, wholesale prices for butter and skim milk powder were considerably higher than EU prices, in some years double or more, indicating large inefficiencies and high transaction costs downstream from the farm gate (Grethe 2004). No evidence is available on wholesale-retail margins, but the weak structure (fragmented, many stages) and poor performance (over-capacity, inefficiency) of the food distribution chain suggest that this margin is also significant. Thus, the chance of a fall in the price of raw milk failing to reach the consumer of dairy products is strong. Since most agricultural commodities undergo storage, grading, processing and/or transportation before reaching the consumer, the effectiveness of the whole downstream sector in passing on policy reforms is crucial. Conclusions Measured transfers to the agricultural sector in Turkey have been high relative to GDP throughout the review period. They reached historic levels in the second half of the 1990s. However, the institutional set-up for delivering support to agriculture siphoned off or wasted a significant portion of the transfers captured from other economic sectors, and little benefit has reached poor farmers. Agricultural policy reform, begun in 2000, has reduced the budgetary cost of agricultural support. The inefficient state marketing sector has been partly restructured, and its role in the marketing and distribution of agricultural commodities has diminished. Furthermore, the switch from coupled input and output subsidies to direct income support has reduced the potential for 34 farm support to distort production decisions. As the most recent figures suggest, the level of support granted to agriculture since 2000 is lower relative to that of the 1990s. This is particularly true with respect to import-competing products. Even though in the beginning of the new century there was a slight increase in subsidies, the level of support for exportables and total agriculture remained stable at rates of around 5 and 20 percent, respectively. However, estimates for 2005 show a small increase in support for exportables and for agriculture as a whole, while support to import-competing products remains high at 50 percent. The massive agricultural support structure consisting of state-owned and state-controlled enterprises, which dominated agricultural markets for most of the review period, had its origins in the etatist policies of the early decades of the Republic (1923-1960). The role of the state- owned sector was enhanced when it became an indispensable agent of government in implementing successive five-year plans, from the 1960s onwards. It seems to have flourished under both the import substitution regime of the 1960s and 1970s, and the export promotion policies of the 1980s. The agricultural parastatals were amongst the most influential lobbyists on behalf of the sector, and were already successful in capturing transfers from other sectors when Turkey's income was far lower than at present. These organisations also played a role in providing social benefits (conspicuously lacking from central government), and thus enjoyed support in local rural areas. OECD (1994) estimated that in the early 1990s the state-owned sugar corporation, TSFAS, alone affected the livelihood of 3 million people. Much has been written about the distributional conflict that is endemic in Turkey. Virtually all occupational and interest groups (except for government employees and consumers) are organised, and the various distributional coalitions all operate very effectively politically (Olgun 1991). As Olgun notes, rent-seeking is pervasive in Turkey, and the success of political parties has often appeared to depend on maximising support amongst these powerful interest groups, if not simultaneously then sequentially (see, for example, Demir, 2002). The political economy of Turkey's agricultural policies has to be seen in this general context. Given that `agricultural interests' found their most powerful (and successful) proponents and lobbyists in the agricultural parastatals and agribusiness, it is not surprising that, even when the pendulum was swinging in favor of agriculture, the small farmer (that is, the majority of primary producers) was not well served. Furthermore, the dismantling or disempowering of these well-entrenched interest groups is problematic in countries in transit to parliamentary democracy, or with 35 relatively fragile political coalitions, who need an electoral mandate for their reform but may depend politically on the very groups who will lose from it (Demir 2002). It is also clear that macroeconomic stability is a necessary condition for setting agricultural policy on a more consistent and effective future path. Taking the previous 20 years' experience in Turkey as a starting point, Rodrik (1999) argued that the greater the latent distributional conflicts among social groups, and the weaker the political institutions for managing these conflicts, the greater is the (negative) impact of external shocks on the internal growth process. An external shock, such as the oil price crises of the 1970s, requires adjustment processes that are rarely without distributional consequences. When the latent rivalry amongst different social groups is strong, such a shock can trigger an outbreak of distributional conflict whose repercussions may take years to stabilise. When this occurs, `inadequate adjustment condemns the country to foreign exchange bottlenecks, import compression, debt crises and bouts of high inflation'. In the resulting chaos, some interest groups will fare better than others for reasons that lie outside the realm of economic analysis. A prime casualty is the medium-term rate of growth. This theory sheds light on general macroeconomic developments in Turkey over the last 30 years, developments that have consistently overshadowed the design, implementation and outcomes of sectoral policies. However, the re-instrumentation of agricultural policy and the liberalization of markets, which were introduced in 2001, represent important necessary conditions for the reduction of rural poverty in Turkey. A revision of economy-wide conditions will further improve future prospects for agriculture and the rural population by establishing a basis that allows sectoral policy changes to be more effective. Even though OECD (2004) describes the recovery from the 2001 crisis as job-poor, the recent new legislation aiming for greater flexibility of labor markets can be seen as a first step. The need to broaden the tax base and encourage private investment, not least foreign direct investment, are also regularly cited as pre-conditions for the kind of growth that would provide a background against which Turkish agriculture can modernise and restructure. In October 2005, Turkey began negotiations for EU membership. Negotiations are still at a very early stage. It is expected that Turkey will align its border protection for agriculture and its domestic agricultural policies more closely with those of the EU, but not necessarily immediately and possibly not until nearer the end of those negotiations. Also, inflows of FDI 36 triggered by the prospect of EU entry could stimulate job creation in other sectors that would absorb some of agriculture's surplus labor. In addition, pre-accession EU funds might be used to finance programs that would lift parts of the rural population out of the poverty trap. The initiative must come from the Turkish authorities themselves to prioritise this expenditure so that it maximises the longer-term prospects of the sector. In conclusion, the distortions that impoverish Turkish farmers are structural and institutional, and are compounded by the low level of human capital throughout the sector. Distortions elsewhere in the economy (rigid labor markets, persistent low rates of job-creating private investment due to distorted financial markets resulting from large public borrowing requirements), together with a high birth rate, have created a situation where it is impossible for agriculture to shed population fast enough and to achieve any degree of spontaneous restructuring independently. Distortionary agricultural policies cannot be singled out as the main cause of rural poverty in Turkey. But the recent developments towards a lower level of intervention represents a first step in the right direction. Since the reduction of agricultural support does not represent a panacea for ameliorating entrenched poverty in Turkey's rural areas, however, it is also necessary to promote macroeconomic stability and reform in order to provide a more conducive context for sectoral policy liberalization. References Akyüz, Y. and K. Boratav (2003), "The Making of the Turkish Financial Crisis", World Development 31: 1549-1566. Anderson, K., M. Kurzweil, W. Martin, D. Sandri and E. Valenzuela (2008), "Methodology for Measuring Distortions to Agricultural Incentives", Appendix 2 to this volume. Bayaner, A. and Ö. Bor (2006), "Do the Policies Always have the Same Consequences? The Impact of Direct Income Support on Wheat Production: the Case of Turkey", New Medit 5(1): 15-20. 37 Burrell, A. (2005), "Turkey's Foreign Trade Position", Chapter 8 in A.M. Burrell and A.J. Oskam (eds.), Turkey in the European Union: Implications for Agriculture, Food and Structural Policy, Oxford: CABI Publishing. Cakmak, E. (2003), "Evaluation of the Past and Future Agricultural Policies in Turkey: Are they Capable of Achieving Sustainability?", Options Méditerranéennes Série A 52: 1-11. Candemır, H.B. (1994), "External Debt and Internal Transfer Problem: The Case of Turkey: 1980-1990", Discussion Paper 9404, Research Department, Central Bank of the Republic of Turkey, January. Demir, F. (2002), "Politics, Society, Labour Markets and Patterns of Distribution during Transition: the Paths of Retransformation in the Post-liberalization Era in Turkey", Seminar on Labour Market, Work and Welfare during the Transition and Integration Processes. Demir, F. (2004), "A Failure Story: Political and Financial Liberalization in Turkey, Revisiting the Revolving Door Hypothesis" , World Development 32: 851-869. Dervis, K., M. Emerson, D. Gros and S. Ülgen (2004), The European Transformation of Modern Turkey, Brussels: CEPS. Easterly, W. (2006), Global Development Network Growth Database. Available at www.nyu.edu/fas/institute/dri/global%20development%20network%20growth%20databa se.htm, accessed June 2007. Flam, H. (2003), "Turkey and the EU: Politics and the Economics of Accession", Seminar Paper No. 718, Institute for International Economic Studies, Stockholm. Grethe, H. (2004), Effects of Including Agricultural Products in the Customs Union between Turkey and the EU: A Partial Equilibrium Analysis for Turkey, Berlin: Peter Lang Publishing Group. Harrison, G.W., T.F. Rutherford and D.G. Tarr (1993), "Piecemeal Trade Reform in the Partially Liberalized Economy of Turkey", The World Bank Economic Review 7: 191-217. Kasnakoglu, H. and Cakmak, E. (2000), "Turkey", Chapter 5 in A. Valdés, (ed.), Agricultural Support Policies in Transition Economies, World Bank Technical Paper No. 470, Washington DC. Krueger, A, M. Schiff, and A. Valdés (1991), The Political Economy of Agricultural Pricing Policy, Vol 3: Africa and the Mediterranean, Baltimore: John Hopkins University Press for the World Bank. 38 Niron, S. (1986), "Fertilizer Pricing in Turkey", in E. L. Segura, Y.T. Shetty and M. Nishimizu (eds.), Fertilizer Producer Pricing in Developing Countries: Issues and Approaches, Washington DC: World Bank. OECD (1994), National Policies and Agricultural Trade, Country Study: Turkey, Paris: OECD. OECD (2002), "Economic Survey of Turkey", Policy Brief, OECD Observer, Paris. OECD (2004), OECD Economic Surveys: Turkey 2004, Paris: OECD. Olgun, H. (1991), "Turkey", Chapter 6 in A.O. Krueger, M. Schiff and A. Valdés (Eds.), The Political Economy of Agricultural Pricing Policy, Volume 3: Africa and the Mediterranean, Baltimore: Johns Hopkins University Press. Rodrik, D. (1999), "Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses", Journal of Economic Growth 4: 385-412. Sirtioglu, I. (2004), Turkey Retail Food Sector Report 2004, GAIN report No TU4005, Ankara: USDA FAS. State Institute of Statistics (SIS) (2004), "Poverty Study", Press Release 13/04/2004. Downloaded from www.die.gov.tr/english/SONIST/YOKSL/200404/200404.htm. Taymaz, E. (1999), "Trade Liberalization and Employment Generation: The experience of Turkey in the 1980s", in A. Revenga (Ed.), Turkey: Economic Reforms, Living Standards and Social Welfare Study, Vol 2 Technical Papers, Washington DC: World Bank. Togan, S. (2005), "Turkey: Trade Policy Review", The World Economy 28(9): 1229-1262. Togan, S., A. Bayaner and J. Nash (2005), "Analysis of the Impact of EU Enlargement on the Agricultural Markets and Incomes of Turkey", Chapter 2 in B. Hoekman and S. Togan (eds.), Turkey: Economic Reform and Accession to the European Union, Washington DC: World Bank. van Berkum, S. (2005), "Agricultural and Food Industry Structure", Chapter 6 in A.M. Burrell and A.J. Oskam (eds.), Turkey in the European Union: Implications for Agriculture, Food and Structural Policy, Oxford: CABI Publishing. World Bank (2000), Turkey: Economic Reforms, Living Standards and Social Welfare Study, Poverty Reduction and Economic Management Unit, Report No 20029-TU, Washington DC: World Bank. 39 World Bank (2001), Project Appraisal Document on a Proposed Loan in the Amount of US$ 600 Million to the Republic of Turkey for an Agricultural Reform Implementation Project/Loan, Report No. 21177-TU, Washington DC: World Bank, June. World Bank (2004a), "Turkey: Agricultural Reform Implementation Project. Release of the Second Tranche ­ Full Compliance (Loan 4631-TU)", Memorandum, World Bank, Washington DC. World Bank (2004b), Turkey: A Review of the Impact of the Reform of Agricultural Sector Subsidisation, ESSD Unit, Europe and Central Asia Division, Washington DC. World Bank (2007), World Development Indicators 2007, Washington DC: World Bank. WTO (World Trade Organization) (2005). Turkey ­ measures affecting the importation of rice. WT/DS334/1, 7 November 2005. Ziraat Bank (2003), 2003 Annual Report, Ankara: Ziraat Bank. Table 1: Population and GDP, Turkey, 1950 to 2004 Population Rural2 Agric. GNP Period Population Rural Agricult Real Real population workers3 (current populati ural GNP1 GNP1 USD) on workers3 per capita million million million billion Average growth percent per year 1950 20.8 15.7 7.4 3.4 1960 27.5 19.3 8.3 9.8 1950-60 2.8 2.1 1.2 6.3 3.4 1970 35.3 21.8 8.2 191 1960-70 2.5 1.2 -0.1 5.6 3.0 1980 44.5 25.0 8.4 681 1970-80 2.3 1.4 0.1 4.0 1.6 1990 56.2 22.9 8.7 151 1980-90 2.4 -0.9 0.4 5.2 2.8 2000 67.4 23.8 7.8 200 1990-00 1.9 0.4 -1.1 3.5 1.6 2004 71.7 23.8 7.4 300 2000-04 1.6 0.1 -1.2 3.2 1.6 1. In constant 1987 Turkish Lira. 2. Population living in settlements of less than 20 thousand inhabitants. 3. Including forestry, hunting and fishing. Based on Household Labour Force Survey. Sources: Turkish Statistical Institute, World Bank. Table 2: Applied tariffs and tariff bindings, selected agricultural products, Turkey, 1992 to 2005 1 st st 2004/5 Product As of 1 July 1992 As of 1 July 1994 1997 2003 MFN Border AVE MFN Border AVE Binding tariff price tariff price Average MFN tariff percent percent 2 2 2 2 2 % USD % % USD % S W S W S Beef & veal 1 4246 57 15 4246 59 165 165 228 - 225 Pigmeat 30 1646 139 20 1563 135 200 - 228 - 225 Sheepmeat 1 3105 68 15 3105 63 165 - 228 - 225 Poultry meat 30 1755 60 5 1764 39 64 65 62 63 86 Milk powder 1 1515 105 20 1722 95 130 130 150 150 180 Cheese 73 75 88 99 139 Eggs 15 845 75 20 801 88 23 23 24 24 53 Potatoes 23 23 20 20 19 3 Hazelnuts 5 n.a. n.a. 5 n.a. 40 40 40 44 44 43 Citrus 50 50 50 55 54 3 3 Grapes/raisins 1 n.a. n.a. 10 n.a. n.a. 55 55 56 55 55 3 3 Tea 20 n.a. n.a. 10 n.a. n.a. 145 145 145 145 168 Wheat exempt 141 21 3 162 65 12 12 14 13 180 Rye 5 n.a. 31 5 n.a. 30 30 - 60 60 180 Barley exempt 97 31 3 83 39 10 10 43 43 180 Maize exempt 110 27 3 97 34 18 34 10 20 180 Sorghum 15 n.a. 42 10 n.a. 40 15 15 10 10 180 Rice 1 308 42 26 275 44 33 34 40 39 45 Soybeans exempt 237 2 exempt 246 0 0 0 0 0 23 Rape seed 1 n.a. 26 10 n.a. 25 24 24 9 12 10 Sunflower seeds exempt 233 34 3 233 12 19 19 8 8 26 4 Sugar beet 10 32 37 10 33 11 23 23 20 20 19 3 Olive oil 10 n.a. n.a. 5 n.a. 37 37 37 32 32 31 Palm oil 10 8 10 8 36 Sun-/safflower oil 23 25 26 22 32 Sugar exempt 275 73 3 280 146 135 135 127 98 135 Tobacco 0 4023 75 25 3770 78 25 25 25 25 45 Cigarettes 58 58 39 29 151 4 Hides, skins 2 1 2.23 1 30 Wool exempt 4267 0 2 3738 2 0 0 0 0 6 Cotton exempt 1302 0 - 1297 2 0 0 0 0 6 1. Product classification at the 4-digit HS level, and the for least processed form of the product, unless otherwise stated. 2. S=simple average; W=trade-weighted average. 3. Not possible to calculate an AVE, but the Mass Housing Fund Levy alone for these products was (in USD/t): hazelnuts (1992: 100, 1994: 35), raisins (1992: 100, 1994: 300), tea (1992: 2000, 1994: 3000), olive oil (1992: 200, 1994: 32). 4. Sugar beet: HS code 121291. Hides and skins: HS code 41. Sources: OECD (1994), WITS database. 2 Table 3: Credit and fertilizer use, Turkey, 2000 to 2004 2000 2001 2002 2003 2004 Agricultural credit (million YTL) Total ag. credit given by the Agricultural Bank 3409 2949 2884 3562 4017 Credit originating from the Bank 1575 770 600 534 1293 directly to producers 832 541 444 455 1220 ACCs and other organisations 243 2228 156 7 7 Credit originating from funds 1834 2179 2285 3028 2724 Fertilizer use (thousand tons) Total fertilizer 10425 8293 8696 9762 10153 Source: Turkish Statistical Institute. 3 Table 4: Direct income support scheme, Turkey, 2001 to 2005 2001 2002 2003 2004 2005 Payment per hectare (YTL) 100 135 160 160 n.a. Area ceiling (ha per farm) 20 50 50 50 50 Participating hectares (mn) 11.8 16.2 16.5 16.7 16.7 Share of agricultural area ( percent) 48 66 67 68 68 Farming households participating (mn) 2.2 2.6 2.7 2.7 2.7 Share of farming households 54 75 87 87 87 Total cost (YTL bn) 1.18 2.19 2.64 2.66 n.a. Sources: World Bank (2004), MARA. 4 Table 5: Nominal rates of assistance to agricultural industries, Turkey, 1961 to 2005 (percent) 1961-64 1965-69 1970-74 1975-79 1980-84 1985-89 1990-94 1995-99 2000-04 2005 Importables Rice -- -- -- -- -- 27 246 111 98 - Wheat -11 -19 -12 -7 -26 39 45 47 26 39 Maize -- -33 -13 -- -45 12 45 55 40 68 Sunflower -- -6 -20 -27 -61 15 37 54 19 23 Sugar -- -- -- -- -- 15 36 113 116 136 Milk -- 38 47 73 7 62 125 127 57 29 Beef and veal -- -- -- -- -- 20 65 120 133 102 Exportables Barley -15 -44 9 -21 -37 22 59 63 23 51 Potatoes -- 7 3 12 -7 23 27 32 118 315 Poultry -- -- -- -- -- 34 57 62 50 66 Sheepmeat -- -- -- -- -- 14 22 31 11 14 Eggs -- -- -- -- -23 21 32 73 51 121 Grape -- -11 2 -7 -11 6 19 5 29 44 Apple -- -27 -40 -53 -52 -13 -16 6 0 92 Cotton -26 -21 -11 -7 -31 -54 -47 -49 -36 -45 Tobacco -46 -47 -38 -38 -68 0 37 49 -12 16 Tomatoes -- -39 -41 -37 -36 34 9 -13 -10 -31 Hazelnuts -- -- -- -22 -57 -47 -40 -31 -4 - Importables -11 -10 6 20 -20 28 60 81 54 50 Exportables -30 -28 -18 -23 -35 -8 3 -2 3 11 Weighted average of -19 -18 -7 -8 -30 4 20 21 20 26 above Standard Deviation 19 33 33 41 27 41 71 62 57 92 Share of above products in total gross -- 64 63 60 58 50 53 55 54 57 value of agricultural production* * Share is calculated in terms of undistorted prices Source : Authors' calculations based on FAO data to 1986 and OECD data thereafter. 5 Table 6: Real changes in purchase price of selected agricultural commodities, Turkey, 1978 to 2003 (prices deflated by the Wholesale Price Index, annual percentage changes) Wheat Cotton Tobacco Tea Sugar beet Sunflower Hazelnuts 1978 -26.7 -16.2 -26.0 -21.4 -13.2 -14.3 -14.6 1979 -4.1 10.9 -25.8 -26.3 -3.7 -13.8 6.4 1980 -2.1 -6.6 -11.5 -7.8 5.4 20.7 39.8 1981 33.1 -7.6 -10.2 7.6 7.9 -3.1 -17.0 1982 -4.8 -5.1 18.5 3.8 -0.7 6.4 -6.6 1983 -3.3 28.2 2.4 1.0 -11.6 -15.0 -13.8 1984 7.3 -1.3 -11.0 -7.4 -11.9 19.9 -0.3 1985 -6.8 -9.8 -0.7 -3.2 8.6 -9.7 86.5 1986 -2.3 -4.8 11.0 -2.4 -5.8 -8.7 -21.0 1987 -7.6 58.3 40.9 -6.2 0.6 -3.5 30.6 1988 0.2 -17.5 18.0 -4.0 16.6 0.9 -0.3 1989 20.8 13.7 -1.0 -3.6 13.1 10.1 -16.9 1990 0.9 -9.0 7.4 4.5 11.2 -15.3 -18.4 1991 -3.5 0.9 -9.2 2.0 -1.2 12.0 4.8 1992 -5.4 -5.5 -3.0 46.0 0.4 0.6 2.8 1993 0.6 1.5 -19.9 -8.3 -4.2 1.0 -0.4 1994 -13.2 -9.4 -26.2 -20.0 -13.4 7.9 79.6 1995 13.1 20.8 2.2 7.5 43.3 1.6 -21.9 1996 69.0 -4.2 6.0 18.4 -1.3 19.9 43.3 1997 -12.5 12.1 12.7 11.0 39.7 -1.9 41.7 1998 -11.3 -24.4 1.4 1.0 -15.0 -4.0 -15.4 1999 -4.9 -9.7 4.7 -6.0 2.9 -24.0 -8.6 2000 -12.1 7.0 -20.1 -14.7 -15.2 -16.2 -33.9 2001 -6.4 -2.3 -14.2 -3.6 -14.6 37.3 -13.2 2002 -2.8 -21.6 -11.9 -15.8 5.8 -16.3 -29.5 20031 23.4 -1.4 -63.9 -0.5 -1.4 -20.4 26.4 1. Estimate. Source: Authors' calculations based on data from the State Planning Organisation. 6 Table 7: Summary of influences on Turkey's agricultural and food sectors A. Policy-induced distortions B. `Exogenous' influences 1. Macro-economic policies Fragmentation of food Exchange rate policy supply chains Monetary policy (inflation) Poor market infrastructure, 2. Agricultural policies, including poor transport and instruments storage facilities, weak Tariffs, NTBs information transmission etc Product/input subsidies, taxes on market transactions Market imperfections Product/input subsidies, Endemic rent-seeking and taxes direct to farmers corruption Supply quotas Area controls Lack of sufficient education, Area payments training and job creation policies 3. Agricultural policies, delivery mechanisms Inefficient state enterprises Long information and feedback channels Implementation delays and inconsistencies Weak supervision of implementation protocols Weak enforcement of conditions for eligibility 4. Incidence of agricultural policy within agriculture Bias in favour of larger/richer farmers Bias with respect to products or farm types Regional bias Source: Authors' categorization. 7 Table 8: Nominal rates of assistance to agriculture, to non-agricultural sectors, and relative rate of assistance, Turkey, 1961 to 2005 (percent) 1961-64 1965-69 1970-74 1975-79 1980-84 1985-89 1990-94 1995-99 2000-04 2005 Covered Products -19 -18 -7 -8 -30 4 20 21 20 26 Non-covered products -19 -18 -7 -8 -30 4 20 21 20 26 All agric. productsa -19 -18 -7 -8 -30 4 20 21 20 26 Non-product specific input 0 0 2 1 0 0 0 0 3 5 assistance (NPS) Total agriculture incl. -19 -18 -5 -7 -29 4 20 21 23 31 NPSa,b Importablesa -11 -10 6 20 -20 28 60 81 54 50 Exportablesa -30 -28 -18 -23 -35 -8 3 -2 3 11 TBI -21 -16 -17 -35 -17 -28 -35 -44 -32 -26 Tradables All Agriculturea,b -19 -18 -7 -8 -30 4 20 21 23 31 All Non-Agriculture 61 141 50 56 35 20 10 2 1 0 RRAc -46 -64 -37 -36 -50 -13 9 19 22 30 a NRA including product-specific input subsidies b NRA including other (incl. decoupled & non-product-specific) subsidies c The RRA is defined as 100*[(100+NRAagt)/(100+NRAnonagt) - 100], where NRAagt and NRAnonagt are the average percentage NRAs for the tradables part of the agricultural and non-agricultural sectors, respectively Source: Authors' calculations. 8 Table 9: Consumer tax equivalent for food products, Turkey,a 1986 to 2005 (percent) Other Beef Sheep We Wheat Maize Barley Apples Grapes Potatoes Tomatoes Sunflower Sugar Tobacco Poultry Eggs Milk a Grains and veal meat 1986 6 5 4 4 -9 13 -1 22 24 17 21 45 14 4 19 186 1987 35 9 5 5 3 1 48 43 24 16 22 23 28 19 13 122 1988 57 9 -1 -1 10 -4 22 56 -7 -1 -15 -11 9 10 9 69 1989 31 7 0 0 -29 46 40 65 23 -13 -9 1 22 24 47 95 1990 16 15 4 4 -12 13 61 21 22 17 8 46 25 53 29 530 1991 64 20 5 5 -23 2 9 32 41 50 63 124 24 42 9 508 1992 18 18 4 4 0 38 44 15 45 62 24 67 13 63 76 207 1993 10 10 7 7 -14 20 30 12 30 50 33 57 26 81 36 146 1994 11 -4 2 2 -23 35 10 -18 1 -18 38 0 25 29 17 100 1995 -12 3 0 0 12 -6 47 -31 25 40 46 43 28 66 96 64 1996 3 6 1 1 6 -17 70 10 46 44 64 23 -7 29 44 81 1997 30 12 1 1 3 15 2 7 56 117 51 88 0 23 55 135 1998 50 14 3 3 0 24 17 -19 37 123 36 148 26 73 69 113 1999 51 12 3 3 1 0 14 -36 29 183 -4 146 33 54 74 95 2000 19 11 2 2 3 26 32 -1 28 120 -20 131 27 52 78 79 2001 -5 1 0 0 -19 -8 46 -16 23 37 -48 79 -15 19 32 -5 2002 13 6 1 1 -1 30 174 -19 6 90 -16 118 9 47 36 55 2003 56 13 2 2 -1 30 145 5 12 150 6 194 21 53 19 61 2004 22 13 3 3 16 67 194 -20 12 167 2 113 7 75 87 51 2005 35 13 4 4 92 44 315 -31 23 136 16 96 14 66 121 27 a. The negative of the OECD's (2007) Consumer Subsidy Equivalent (CSE), expressed as a percentage of undistorted prices. Source: Authors' calculations, based on (OECD 2007). 9 Figure 1: Growth of real agricultural GDP, Turkey, 1950 to 2005 20 15 10 per cent per year 5 0 -5 -10 -15 1950 1954 1958 1962 1966 1970 1974 1978 1982 1986 1990 1994 1998 2002 Growth rate of real agricultural GNP Source: Turkish Statistical Institute. 10 Figure 2: Agricultural trade, Turkey, 1961 to 2005 8 100 90 7 80 6 70 5 60 USD billion per cent 4 50 40 3 30 2 20 1 10 0 0 1961 1966 1971 1976 1981 1986 1991 1996 2001 Agricultural exports Agricultural imports (X-M)/(X+M) % (right hand axis) Source: FAOSTAT, Turkish Ministry of Industry and Trade. 11 Figure 3: Agricultural Producer Support Estimates, Turkey, 1986 to 2005 (a) Producer Support Estimate, General Services Support Estimate, and Total Support Estimate (real YTL billion) 30 25 20 real YTL (2004) bn 15 10 5 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 PSE GSSE TSE 12 Figure 3 (continued): Agricultural Producer Support Estimates, Turkey, 1986 to 2005 (b) Market Price Support, Total Support Estimate, and budgetary expenditure on agriculture (as percent of GDP) 8 7 6 per cent of GDP 5 4 3 2 1 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 MPS TSE budgetary expenditure Source: OECD PSE database. 13 Figure 4: Nominal rates of assistance (NRA) to exportable and import-competing agricultural industries, and to total agriculture (including product-specific support), Turkey, 1961 to 2005 100 Importables Exportables Weighted average of covered products 80 60 40 20 0 1961-64 1965-69 1970-74 1975-79 1980-84 1985-89 1990-94 1995-99 2000-04 2005 -20 -40 Source: Authors' own calculations. 14 Figure 5: Relative Rate of Assistance to primary agriculture, Turkey, 1961 to 2005 150 All agriculture All non-agriculture RRA 100 50 0 1961-64 1965-69 1970-74 1975-79 1980-84 1985-89 1990-94 1995-99 2000-04 2005 -50 -100 Source: Authors' own calculations. 15 Figure 6: Nominal Rates of Assistance including and excluding product-specific and non-product-specific (NPS) support, Turkey, 1986 to 2005 160 Importables excl. input subsidies Exportables excl. input subsidies Total excl. input subsidies Importables incl. input subsidies Exportables incl. input subsidies Total incl. input subsidies 140 Total incl. input subsidies & NPS 120 100 80 60 40 20 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 -20 Source: Authors' own calculations. 16 Appendix Table A1: Nominal rate of assistance to agric. products, Turkey, 1960 to 2005 Rice Wheat Maize domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 n.a. n.a. n.a. 591 n.a. n.a. n.a. n.a. n.a. 1961 n.a. n.a. n.a. 731 73 -0.20 n.a. 66 n.a. 1962 n.a. n.a. n.a. 811 73 -0.07 n.a. n.a. n.a. 1963 n.a. n.a. n.a. 800 73 -0.07 n.a. 69 n.a. 1964 n.a. n.a. n.a. 798 76 -0.10 n.a. n.a. n.a. 1965 n.a. n.a. n.a. 858 75 -0.01 n.a. n.a. n.a. 1966 1947 n.a. n.a. 877 73 0.04 831 n.a. n.a. 1967 1654 n.a. n.a. 880 138 -0.46 826 n.a. n.a. 1968 1721 n.a. n.a. 898 200 -0.64 838 n.a. n.a. 1969 2074 n.a. n.a. 951 66 0.13 937 110 -0.33 1970 2201 n.a. n.a. 985 63 0.32 953 n.a. n.a. 1971 2347 n.a. n.a. 1029 71 -0.09 923 66 -0.13 1972 2585 n.a. n.a. 1081 80 -0.08 1064 n.a. n.a. 1973 4170 n.a. n.a. 1310 202 -0.56 1460 n.a. n.a. 1974 5120 n.a. n.a. 2330 204 -0.22 2350 n.a. n.a. 1975 5390 n.a. n.a. 2675 224 -0.24 2510 n.a. n.a. 1976 5590 n.a. n.a. 2660 153 0.02 2620 n.a. n.a. 1977 9960 n.a. n.a. 2915 127 0.01 3310 n.a. n.a. 1978 15130 n.a. n.a. 3570 n.a. n.a. 4410 n.a. n.a. 1979 18590 n.a. n.a. 5190 n.a. n.a. 5940 n.a. n.a. 1980 25690 n.a. n.a. 10575 n.a. n.a. 13020 n.a. n.a. 1981 54380 n.a. n.a. 18330 174 -0.18 22450 n.a. n.a. 1982 57720 n.a. n.a. 23015 174 -0.27 27900 188 -0.18 1983 63250 n.a. n.a. 27415 155 -0.28 28230 1600 -0.93 1984 207000 n.a. n.a. 43965 173 -0.31 46790 164 -0.23 1985 302000 571 0.10 63500 153 -0.14 66000 184 -0.25 1986 345000 600 -0.19 79030 111 0.14 70440 96 0.12 1987 425000 1067 -0.56 96760 82 0.53 86360 85 0.24 1988 947000 n.a. n.a. 168090 72 0.78 209710 122 0.22 1989 1902000 322 1.75 327650 116 0.62 310170 125 0.27 1990 2333000 209 3.24 503370 164 0.33 442350 121 0.52 1991 2748000 176 2.58 762590 106 0.89 797770 117 0.68 1992 5308400 216 2.42 1155910 141 0.35 1185080 110 0.63 1993 7238800 184 2.48 1842920 151 0.25 1599740 110 0.40 1994 12893000 169 1.57 3591000 109 0.43 3133000 119 0.03 1995 20382000 190 1.33 7432000 187 0.04 6769000 138 0.23 1996 39510000 205 1.44 18508000 219 0.21 16871000 177 0.34 1997 64273000 197 1.08 35141000 174 0.46 28503000 128 0.58 1998 115953000 223 1.14 53566000 132 0.87 46969000 110 0.94 1999 170804000 265 0.54 75963000 116 0.77 65189000 103 0.64 2000 249248000 199 1.00 97325000 128 0.30 90000000 103 0.46 2001 374372992 181 0.69 157098000 136 -0.04 139840000 108 0.07 2002 561713984 167 1.23 229321000 134 0.15 212834000 120 0.18 2003 669937024 224 0.99 353620000 144 0.64 305652000 126 0.61 2004 n.a. 431 n.a. 350440000 199 0.24 318705000 132 0.69 2005 n.a. n.a. n.a. 331000000 177 0.39 260000000 115 0.68 17 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Sunflowers Sugar Milk domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 n.a. n.a. n.a. 461 n.a. n.a. n.a. n.a. n.a. 1961 n.a. 149 n.a. 376 n.a. n.a. n.a. 43 n.a. 1962 n.a. 198 n.a. 382 n.a. n.a. n.a. 34 n.a. 1963 n.a. 280 n.a. 385 n.a. n.a. n.a. 83 n.a. 1964 n.a. 181 n.a. 427 n.a. n.a. n.a. 105 n.a. 1965 n.a. 172 n.a. 427 n.a. n.a. n.a. 25 n.a. 1966 1735 148 0.01 442 n.a. n.a. 1322 84 0.36 1967 1720 152 -0.04 448 n.a. n.a. 1386 72 0.61 1968 1707 156 -0.14 439 n.a. n.a. 1486 69 0.70 1969 1695 144 -0.07 439 303 n.a. 1593 146 -0.14 1970 1851 164 -0.05 606 n.a. n.a. 1754 92 0.61 1971 2005 165 -0.24 606 n.a. n.a. 2125 109 0.22 1972 2425 164 0.01 606 n.a. n.a. 2288 99 0.59 1973 2380 221 -0.27 848 n.a. n.a. 2702 117 0.57 1974 2940 362 -0.44 1091 670 n.a. 3340 168 0.37 1975 5710 333 0.09 1364 600 n.a. 4750 179 0.69 1976 5490 334 -0.04 1697 700 n.a. 5520 115 1.81 1977 5650 324 -0.23 1879 n.a. n.a. 6520 195 0.47 1978 6440 373 -0.48 2212 569 n.a. 9910 210 0.41 1979 9840 747 -0.70 3364 351 n.a. 14260 261 0.26 1980 19380 561 -0.59 4879 773 n.a. 26190 272 0.13 1981 31340 582 -0.58 11848 771 n.a. 35910 322 -0.13 1982 40150 582 -0.62 15121 183 n.a. 43370 168 0.43 1983 52180 600 -0.64 18000 336 n.a. 54800 306 -0.27 1984 90840 613 -0.60 22788 280 n.a. 82000 188 0.18 1985 151000 375 -0.17 33333 356 n.a. 135000 173 0.62 1986 168000 200 0.33 132145 168 0.33 110860 58 1.17 1987 215000 202 0.31 164944 166 0.24 135100 71 0.64 1988 370000 280 -0.04 363388 258 0.05 245930 102 0.24 1989 668000 257 0.32 678743 369 -0.03 432890 104 0.43 1990 862000 271 0.36 1040634 340 0.24 744230 45 1.54 1991 1512000 258 0.49 1681416 268 0.50 1142930 45 1.26 1992 2582000 260 0.51 2709118 243 0.66 2100970 99 1.29 1993 4016000 280 0.41 4065106 246 0.55 3135400 116 1.28 1994 9528000 320 0.10 8099607 333 -0.13 6654000 113 0.89 1995 18000000 314 0.34 22131070 345 0.52 13000000 173 0.90 1996 35000000 294 0.68 36532000 312 0.56 25678000 174 1.02 1997 65000000 274 0.66 87396000 265 1.19 54122500 152 1.36 1998 110000000 309 0.62 138020265 237 1.46 88424280 159 1.98 1999 130000000 240 0.40 218379013 184 1.91 114794000 140 1.11 2000 165000000 206 0.37 245858516 179 1.25 167041260 150 0.78 2001 370000000 245 0.27 360080058 214 0.41 222193000 192 0.11 2002 460000000 289 0.07 566815847 198 0.93 345188000 148 0.56 2003 485000000 287 0.14 665185965 178 1.55 445932000 185 0.86 2004 515000000 323 0.12 741367403 195 1.67 524853610 244 0.53 2005 500000000 302 0.23 759272218 239 1.36 503600000 295 0.29 18 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Beef and Veal Maize Barley domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 n.a. n.a. n.a. n.a. n.a. n.a. 458 n.a. n.a. 1961 n.a. n.a. n.a. n.a. 66 n.a. 507 46 -0.12 1962 n.a. n.a. n.a. n.a. n.a. n.a. 545 n.a. n.a. 1963 n.a. n.a. n.a. n.a. 69 n.a. 547 49 -0.04 1964 n.a. n.a. n.a. n.a. n.a. n.a. 544 65 -0.28 1965 n.a. n.a. n.a. n.a. n.a. n.a. 615 n.a. n.a. 1966 n.a. n.a. n.a. 831 n.a. n.a. 688 n.a. n.a. 1967 n.a. n.a. n.a. 826 n.a. n.a. 686 103 -0.44 1968 n.a. n.a. n.a. 838 n.a. n.a. 721 n.a. n.a. 1969 n.a. n.a. n.a. 937 110 -0.33 786 n.a. n.a. 1970 n.a. n.a. n.a. 953 n.a. n.a. 818 n.a. n.a. 1971 n.a. n.a. n.a. 923 66 -0.13 847 56 -0.05 1972 n.a. n.a. n.a. 1064 n.a. n.a. 920 51 0.23 1973 n.a. n.a. n.a. 1460 n.a. n.a. 1185 n.a. n.a. 1974 n.a. n.a. n.a. 2350 n.a. n.a. 1995 n.a. n.a. 1975 n.a. n.a. n.a. 2510 n.a. n.a. 2040 n.a. n.a. 1976 n.a. n.a. n.a. 2620 n.a. n.a. 2045 133 -0.10 1977 n.a. n.a. n.a. 3310 n.a. n.a. 2435 137 -0.22 1978 n.a. n.a. n.a. 4410 n.a. n.a. 3320 121 -0.18 1979 n.a. n.a. n.a. 5940 n.a. n.a. 4715 170 -0.36 1980 n.a. n.a. n.a. 13020 n.a. n.a. 8040 156 -0.39 1981 n.a. n.a. n.a. 22450 n.a. n.a. 14860 184 -0.37 1982 n.a. n.a. n.a. 27900 188 -0.18 17050 170 -0.45 1983 n.a. n.a. n.a. 28230 1600 -0.93 20950 146 -0.42 1984 n.a. n.a. n.a. 46790 164 -0.23 39450 137 -0.22 1985 n.a. n.a. n.a. 66000 184 -0.25 51500 141 -0.24 1986 1006000 1028 0.51 70440 96 0.12 62980 65 0.47 1987 1842000 1741 0.27 86360 85 0.24 79670 57 0.71 1988 2684000 2112 -0.07 209710 122 0.22 139560 103 -0.03 1989 4462000 2089 0.08 310170 125 0.27 269750 124 0.18 1990 7646000 2002 0.54 442350 121 0.52 388850 103 0.55 1991 13606000 1453 1.28 797770 117 0.68 588500 89 0.63 1992 24081000 2104 0.68 1185080 110 0.63 987820 97 0.55 1993 39172000 2277 0.61 1599740 110 0.40 1614840 79 0.92 1994 72175000 2447 0.14 3133000 119 0.03 2727920 79 0.32 1995 179552000 2730 0.62 6769000 138 0.23 5422000 121 0.09 1996 249470000 2482 0.50 16871000 177 0.34 15114000 159 0.32 1997 492009000 1719 1.13 28503000 128 0.58 25488000 126 0.43 1998 1080318000 1672 1.98 46969000 110 0.94 39110000 74 1.38 1999 1612139000 1568 1.79 65189000 103 0.64 60064000 83 0.91 2000 2153528000 1490 1.45 90000000 103 0.46 81600000 100 0.37 2001 2670246000 1214 0.80 139840000 108 0.07 128976000 101 0.05 2002 4077918000 1244 1.18 212834000 120 0.18 151161000 95 0.06 2003 6502353000 1473 1.95 305652000 126 0.61 231830000 119 0.29 2004 6667345000 2197 1.25 318705000 132 0.69 261507000 134 0.37 2005 6830000000 2594 1.02 260000000 115 0.68 250000000 123 0.51 19 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Potatoes Poultry Sheepmeat domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1961 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1962 n.a. 125 n.a. n.a. n.a. n.a. n.a. 1250 n.a. 1963 n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1500 n.a. 1964 n.a. 60 n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1965 n.a. 47 n.a. n.a. n.a. n.a. n.a. 740 n.a. 1966 694 47 0.27 9300 n.a. n.a. n.a. n.a. n.a. 1967 872 69 0.07 11000 n.a. n.a. n.a. n.a. n.a. 1968 877 77 -0.10 12000 n.a. n.a. n.a. 835 n.a. 1969 839 63 0.06 12700 n.a. n.a. n.a. 807 n.a. 1970 920 82 -0.05 14100 n.a. n.a. n.a. 834 n.a. 1971 986 83 -0.25 19000 n.a. n.a. n.a. 1027 n.a. 1972 1240 76 0.11 23900 n.a. n.a. n.a. 1139 n.a. 1973 1530 94 0.11 26000 n.a. n.a. n.a. 1450 n.a. 1974 1810 102 0.22 30000 n.a. n.a. n.a. 2158 n.a. 1975 2130 113 0.20 34000 n.a. n.a. n.a. 2136 n.a. 1976 3030 193 -0.08 43000 n.a. n.a. n.a. 2106 n.a. 1977 3270 175 -0.18 60000 n.a. n.a. n.a. 2396 n.a. 1978 5410 122 0.32 101000 n.a. n.a. n.a. 2890 n.a. 1979 9910 171 0.33 259000 n.a. n.a. n.a. 3054 n.a. 1980 16520 189 0.03 506000 n.a. n.a. n.a. 3105 n.a. 1981 21250 233 -0.29 354000 n.a. n.a. n.a. 3083 n.a. 1982 22880 148 -0.15 412000 n.a. n.a. n.a. 2641 n.a. 1983 29810 130 -0.06 514000 n.a. n.a. n.a. 2347 n.a. 1984 62190 149 0.13 700000 n.a. n.a. n.a. 2109 n.a. 1985 97000 154 0.31 1067000 n.a. n.a. n.a. 2009 n.a. 1986 85000 127 -0.06 650000 927 0.31 1305000 1698 0.09 1987 131000 103 0.40 883000 867 0.42 2133000 1950 0.22 1988 162000 94 0.14 1613000 1026 0.26 2934000 1899 0.03 1989 373000 126 0.38 2612000 991 0.36 4739000 1825 0.22 1990 780000 185 0.60 4545000 1139 0.57 7722000 2367 0.25 1991 795000 175 0.04 6714000 1130 0.43 14681000 2827 0.20 1992 1215000 123 0.37 11527000 1030 0.63 24081000 3099 0.10 1993 2534000 177 0.26 18918000 950 0.83 43015000 3099 0.24 1994 4732000 146 0.10 37329000 978 0.40 85906000 2317 0.29 1995 13937000 207 0.47 79662000 1044 0.82 198809000 3392 0.39 1996 15422000 111 0.75 120890000 1152 0.43 316636000 4169 0.08 1997 29537000 191 -0.02 210192000 1126 0.26 546501000 3593 0.08 1998 73405000 241 0.25 443722000 983 0.95 929015000 2837 0.52 1999 96801000 203 0.14 585502000 908 0.63 1540411000 2766 0.49 2000 132444000 160 0.32 748350000 785 0.56 2244433000 2819 0.33 2001 197265000 110 0.46 1251729000 857 0.19 2789041000 2679 -0.15 2002 264485000 64 1.74 1648397000 742 0.47 5124001000 3113 0.09 2003 307788000 84 1.45 1949975000 848 0.53 6331887000 3499 0.21 2004 355000000 85 1.94 2095185000 839 0.75 6359349000 4180 0.07 2005 400000000 72 3.15 2180000000 980 0.66 6388571000 4188 0.14 20 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Eggs Grapes Apples domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1961 n.a. 401 n.a. n.a. 126 n.a. n.a. n.a. n.a. 1962 n.a. 403 n.a. n.a. 125 n.a. n.a. n.a. n.a. 1963 n.a. 396 n.a. n.a. 117 n.a. n.a. n.a. n.a. 1964 n.a. 235 n.a. n.a. 97 n.a. n.a. n.a. n.a. 1965 n.a. n.a. n.a. n.a. 106 n.a. n.a. n.a. n.a. 1966 5418 n.a. n.a. 1130 109 -0.10 1537 n.a. n.a. 1967 5946 n.a. n.a. 1012 107 -0.20 1466 145 -0.15 1968 6436 n.a. n.a. 1091 102 -0.15 1186 142 -0.34 1969 6709 n.a. n.a. 1258 99 0.00 1301 155 -0.34 1970 7818 n.a. n.a. 1266 88 0.22 1587 188 -0.29 1971 9345 n.a. n.a. 1568 107 -0.08 1652 177 -0.42 1972 9964 n.a. n.a. 1633 123 -0.09 1586 192 -0.44 1973 12000 n.a. n.a. 2540 170 0.02 2060 240 -0.41 1974 17818 n.a. n.a. 3470 226 0.06 2610 328 -0.45 1975 20546 n.a. n.a. 3730 289 -0.17 3140 421 -0.52 1976 22909 n.a. n.a. 4840 327 -0.13 3760 467 -0.53 1977 26364 n.a. n.a. 7870 360 -0.04 4780 459 -0.54 1978 39637 n.a. n.a. 11690 338 0.03 8110 530 -0.54 1979 60000 n.a. n.a. 15730 371 -0.02 12570 568 -0.49 1980 87818 n.a. n.a. 36130 435 -0.02 18430 486 -0.55 1981 157640 1511 -0.18 42910 389 -0.14 23820 463 -0.60 1982 201280 1622 -0.31 47670 289 -0.09 33090 386 -0.53 1983 225090 1205 -0.24 66940 304 -0.10 36540 289 -0.48 1984 327270 1110 -0.20 98990 332 -0.19 53610 253 -0.43 1985 472730 1067 -0.08 144000 321 -0.07 92250 239 -0.20 1986 482400 600 0.27 209000 274 0.07 120000 196 -0.14 1987 825660 856 0.17 277000 321 -0.05 165000 186 -0.02 1988 1168380 751 0.18 419000 308 -0.10 334000 213 0.03 1989 2142180 687 0.53 878000 284 0.44 406000 269 -0.30 1990 3168000 942 0.30 1223000 414 0.12 698000 303 -0.13 1991 4183560 922 0.06 1948000 456 -0.02 1211000 375 -0.26 1992 10241460 846 0.72 3480000 368 0.31 2252000 328 -0.04 1993 12654000 849 0.34 5365000 406 0.16 3292000 347 -0.16 1994 30798000 889 0.20 15583000 389 0.35 8231000 361 -0.23 1995 72396000 805 1.01 19968000 463 -0.06 17569000 341 0.12 1996 125910000 1077 0.51 33299000 495 -0.15 26605000 308 0.10 1997 213732000 906 0.53 50759000 290 0.12 44957000 287 0.00 1998 320184000 725 0.84 121043000 373 0.33 91368000 349 0.08 1999 443952000 608 0.78 177347000 422 0.00 140821000 333 0.01 2000 848556000 765 0.79 280139000 355 0.26 236610000 367 0.03 2001 1248804000 775 0.32 376040000 332 -0.08 302265000 304 -0.19 2002 1696147000 827 0.36 656044000 335 0.30 512562000 345 -0.01 2003 2112000000 1178 0.19 810479000 415 0.30 641282000 432 -0.01 2004 2610000000 977 0.87 979000000 411 0.67 700000000 423 0.16 2005 2700000000 908 1.21 1000000000 517 0.44 750000000 291 0.92 21 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Cotton Tobacco Tomatoes domestic border domestic border domestic border Year price (TL) price ($) NRA price (TL) price ($) NRA price (TL) price ($) NRA 1960 5100 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1961 5140 623 0 4703 986 -0.62 n.a. 29 n.a. 1962 5330 597 0 7936 1060 -0.37 n.a. 333 n.a. 1963 5260 581 0 11651 1497 -0.33 n.a. 51 n.a. 1964 5240 582 0 8896 1581 -0.52 n.a. 59 n.a. 1965 5210 565 0 7431 1308 -0.51 n.a. 100 n.a. 1966 5117 536 -0.17 8216 1262 -0.44 460 77 -0.48 1967 5434 576 -0.20 8588 1285 -0.44 460 34 0.14 1968 5517 608 -0.28 7969 1167 -0.46 465 n.a. n.a. 1969 5646 547 -0.18 6973 1156 -0.52 495 250 -0.84 1970 6108 551 -0.06 7814 1061 -0.38 470 167 -0.76 1971 7730 672 -0.28 9039 1018 -0.44 670 n.a. n.a. 1972 9352 744 -0.14 8142 1051 -0.47 830 222 -0.74 1973 13910 934 0.01 12470 1225 -0.31 1340 125 -0.27 1974 21180 1621 -0.10 18240 1820 -0.31 1460 89 0.13 1975 18160 932 0.24 29630 2791 -0.32 1760 170 -0.34 1976 24180 1141 0.24 36610 3134 -0.32 1790 271 -0.61 1977 28490 1397 -0.10 43440 2843 -0.33 3910 521 -0.67 1978 20050 1254 -0.52 48000 2913 -0.51 7020 262 -0.20 1979 50280 1513 -0.23 62230 2544 -0.44 7710 181 -0.02 1980 92410 1779 -0.39 77570 2792 -0.67 14760 268 -0.35 1981 147680 1691 -0.32 137030 3016 -0.64 21580 255 -0.34 1982 184240 1368 -0.26 191290 3320 -0.68 17150 180 -0.47 1983 275840 1546 -0.27 227510 3420 -0.73 28810 199 -0.41 1984 426480 1672 296950 3103 58000 208 -0.25 1985 535000 1361 781400 3214 98000 191 0.07 1986 319000 927 -0.52 2268000 2783 0.15 140000 169 0.16 1987 525000 1129 -0.49 3402000 3243 0.16 181000 147 0.35 1988 696000 1190 -0.61 4695000 3892 -0.21 310000 139 0.46 1989 1267000 1325 -0.55 8079000 4192 -0.10 481000 138 0.63 1990 1823000 1583 -0.56 11071000 3916 0.07 1001000 318 0.20 1991 2710000 1571 -0.60 26192000 3842 0.57 1285000 234 0.26 1992 4430000 1307 -0.53 33221000 3893 0.19 1854000 234 0.10 1993 6898000 1027 -0.25 52755000 3613 0.29 4596000 372 0.09 1994 17480000 1029 -0.43 115303000 2829 0.75 7541000 309 -0.18 1995 34261000 1763 -0.58 210930000 3157 0.61 10237000 323 -0.31 1996 51432000 1548 -0.58 440752000 3294 0.82 26793000 299 0.14 1997 98253000 1463 -0.57 729197000 3184 0.58 43482000 268 0.04 1998 189333000 1146 -0.32 1129234000 3187 0.46 71085000 337 -0.13 1999 230000000 1030 -0.42 1411543000 3517 -0.04 89170000 331 -0.36 2000 380000000 1110 -0.39 1734024000 3481 -0.20 165016000 266 -0.01 2001 680000000 769 -0.20 2038807000 3189 -0.31 223320000 218 -0.16 2002 800000000 718 -0.22 3574999000 2816 -0.16 285663000 235 -0.19 2003 1010000000 1210 -0.42 4594231000 2892 0.06 518279000 329 0.05 2004 900000000 1541 -0.57 4800000000 3310 0.02 451000000 396 -0.20 2005 900000000 1344 -0.45 5400000000 3475 0.16 500000000 537 -0.31 22 Appendix Table A1 (cont): Nominal rate of assistance to ag products, Turkey, 1960-2005 Hazelnuts domestic border Year price (TL) price ($) NRA 1960 n.a. n.a. n.a. 1961 n.a. n.a. n.a. 1962 n.a. n.a. n.a. 1963 n.a. n.a. n.a. 1964 n.a. n.a. n.a. 1965 n.a. n.a. n.a. 1966 4926 n.a. n.a. 1967 5030 n.a. n.a. 1968 5191 n.a. n.a. 1969 5661 n.a. n.a. 1970 6494 n.a. n.a. 1971 7395 n.a. n.a. 1972 8082 n.a. n.a. 1973 8630 n.a. n.a. 1974 11720 n.a. n.a. 1975 13470 n.a. n.a. 1976 14380 814 0.03 1977 15730 861 -0.20 1978 20600 1026 -0.40 1979 41430 1395 -0.32 1980 81570 1742 -0.45 1981 116020 1999 -0.55 1982 138450 1363 -0.44 1983 164380 4583 -0.85 1984 203050 1207 -0.55 1985 476000 1415 -0.30 1986 700000 n.a. n.a. 1987 1068000 3652 -0.68 1988 1655000 2348 -0.54 1989 2102000 1534 -0.36 1990 2965000 1641 -0.31 1991 4218000 1543 -0.37 1992 6971500 1606 -0.40 1993 12178000 2251 -0.52 1994 46651500 2641 -0.40 1995 61840000 2241 -0.40 1996 113571000 1948 -0.26 1997 270876000 3116 -0.45 1998 494441984 2727 -0.25 1999 737043008 2223 -0.21 2000 1041550020 1891 -0.12 2001 1336525950 1375 -0.21 2002 1670203010 1032 0.07 2003 1895928960 1158 0.09 2004 n.a. 2406 n.a. 2005 n.a. n.a. n.a. 23 Appendix Table A2: Relative rate of assistance, Turkey, 1961 to 2005 (percent) Year NRA ag NRA nonag RRA 1961 -26 100 -63 1962 -13 24 -30 1963 -12 33 -34 1964 -24 86 -59 1965 -13 101 -57 1966 1 123 -55 1967 -23 267 -79 1968 -35 100 -68 1969 -18 113 -62 1970 10 75 -37 1971 -10 69 -47 1972 -8 41 -35 1973 -20 39 -42 1974 -7 24 -25 1975 -7 17 -20 1976 3 20 -14 1977 -13 25 -30 1978 -13 94 -55 1979 -11 122 -60 1980 -24 52 -50 1981 -27 42 -49 1982 -26 24 -41 1983 -49 23 -59 1984 -21 n.a. n.a. 1985 1 n.a. n.a. 1986 9 23 -11 1987 4 21 -14 1988 -5 19 -20 1989 10 16 -5 1990 17 14 3 1991 25 12 11 1992 23 10 12 1993 26 8 17 1994 8 6 1 1995 4 4 -1 1996 14 2 11 1997 19 2 17 1998 43 1 42 1999 25 1 24 2000 23 1 22 2001 1 1 0 2002 24 1 23 2003 39 1 38 2004 29 1 28 2005 31 0 30