The Export Performance of Turkey Following Trade Liberalisation

Selen Sarisoy - MLitt.

A radical change in Turkey's export policy began in 1980, necessitated by a poor economic performance in the 1970s. In this paper, Selen Sarisoy discusses this change in export policy since Turkey's economic liberalisation in 1980.

Introduction

This study will focus on Turkey's export performance after the 1980 foreign trade liberalisation.

The export performance will be analysed under four sections:

A special emphasis will be given to changes in Turkey's export performance after 1980 vis-à-vis the 1970s.

Background

Turkey is situated between West and East bringing many different cultures together. Although 99% of the population is Muslim, Turkey is, and has always been, a Western-oriented secular country. Turkey is a typical middle income developing country with a population of 60 million and a GDP of 169.3 billion US$ in 1995.

Although trade liberalisation attempts started in the 1970s, these attempts were generally unsuccessful as they were only short-term solutions to balance-of-payments and foreign exchange problems (Baysan and Blitzer, 1991). By the end of the 1970s, Turkey faced severe economic problems. Despite the unfavourable external environment, policy makers attempted to maintain or even accelerate aggregate economic growth through inflationary policies, heavy foreign borrowing and postponement of structural adjustments. These unsustainable expansionary policies led to increasing inflation and unemployment, coupled with shortages and labour unrest (Baysan and Blitzer, 1990).

This political and economic turmoil finally led to the realisation that changes were required. Starting in January 1980, a series of comprehensive economic policy changes were introduced. In the long-term, reform policies aimed to change the structure of the economy from etatism to a market structure with the objective to render a more efficient allocation of resources through the price mechanism. In the short-term, the target was to ameliorate the balance-of-payments situation through export growth and to achieve international creditworthiness. The steps followed to achieve the short-term objectives were the adoption of a flexible exchange rate policy, more effective export promotion measures to promote export growth, and gradual import liberalisation (Baysan and Blitzer, 1990). This move towards greater freedom of trade created the necessary external environment to enable Turkey to benefit from trade gains. Unlike the 1970s, the 1980 reforms changed the economy and liberalised trade permanently by reducing or eliminating protectionist measures (Baysan and Blitzer, 1991).

By liberalising imports, Turkey became more open to foreign trade. The openness of any economy involves the absence of restrictions on the movement of goods and factors of production, such as tariffs and quotas to protect domestic industry. Openness, by definition, is more than the existence of trade (Geary, 1991). The extent of openness can be calculated by the ratio of exports and imports to GDP (Greenaway and Sapsford, 1996). Turkey's ratio of exports and imports to its GDP was 17% in 1970. The ratio rose to 20% in 1980 and 34% in 1995 (Figure 1). Reduction or elimination of tariffs and quotas, and the movement of resources from import-substituting activity to the growth of export industries were the main contributing factors in the liberalisation of imports.

Figure 1: The ratio of exports and imports to GDP in Turkey, 1970-1995.

Apologies - graph awaiting insertion.

Source: Author's calculations, OECD Economic Survey of Turkey (various issues).

Export Performance Analysis

Commodity Composition of Exports

Before 1980, the commodity composition of Turkish exports was concentrated on agricultural products. In 1970, agricultural products constituted 75% of exports. Although the share of agricultural products declined steadily towards the end of the 1970-79 period, Turkey remained mainly an agricultural products exporter. The share of processed and manufactured products rose steadily from 19% in 1970 to 35% in 1979. Mining and quarrying remained at around 6% during this period. Within the processed and manufactured products category, processed agricultural products (such as olive oil, sugar, food and beverages) took the highest share, followed by textiles. In the second half of the 1970-79 period, the share of textiles increased faster than that of agricultural processed products, an increase from 4% in 1970 to 17% in 1979.

The analysis of commodity composition of exports indicate that Turkey, as a land and labour abundant country, specialised in resource and labour-intensive industries, where it had a comparative advantage. Therefore, Turkey was a net exporter of agricultural products and a net importer of industrial products in the 1970s.

After 1980, the commodity composition of exports changed dramatically. Manufactured products replaced agricultural products in export commodity structure. The share of agricultural products in total exports fell from 57% in 1980 to 16% in 1995. Despite the fall in percentage share, the value of agricultural products in US$ doubled between 1980 and 1995. The share of manufactured products (excluding agricultural manufactured products) increased steadily from 29% in 1980 to 69% in 1991. Among manufactured products, the traditional export sector, textiles, constituted the highest share. Textiles increased from 15% in 1980 to 32% in 1991. Among other traditional export industries, 'hides and leather' was among the fastest growing industries following the non-traditional 'iron and steel' industry. Post-1980 commodity composition was altered by the emergence of non-traditional export industries such as iron and steel, rubber manufactures, non-ferrous metals, chemicals, glass, ceramics and the brick-tile industry. Iron and steel exports grew from 1% in 1980 to 11% in 1991.

The reason for the change in the commodity composition of Turkish exports was the new development strategy, which targeted industrialisation through export growth (Taskin and Yeldan, 1996). Resources were thus reallocated from the agricultural sector to industry. As the above analysis indicates, Turkey was exporting mainly resource-intensive agricultural products in the 1970s, labour-intensive textiles, together with leather and hide in the 1980s, while rapidly growing capital-intensive export industries emerged in the 1990s.

Comparative Advantage

According to the Heckscher-Ohlin theory, the determinants of traded good industries are identified by their factor contents. Countries that are rich in certain factors will export goods that make use of the abundant factors intensively. While Ricardian comparative advantage is confined to labour productivity, where only one factor of production is considered, the Heckscher-Ohlin theory develops comparative advantage from a unidimensional theory to a multidimensional one.

The result of the preceding analysis on the composition of export commodities indicates a change in the factor composition of the export products. In the 1970s, on average 65.5% of export products were resource-intensive (agricultural products and non-ferrous metals), while 13% were labour-intensive (textiles and leather and hides) and 2.8% (iron and steel, rubber and plastics, industrial chemicals, glass and ceramics) were capital-intensive. After 1980, the share of resource-intensive industries declined to 21% , while the labour-intensive industries' share increased to 37%, and the non-traditional, newly emerging capital-intensive industries accounted for 19% in 1991. Although government policy targeted industrialisation through export growth, the share of labour-intensive sectors still accounted for most of the exports in the 1990s and the capital accumulation has not been satisfactory (Taskin and Yeldan, 1996).

The question arises as to whether the change in factor contents was an indication of a change in the comparative advantage of export industries for Turkey. One method to analyse a country's comparative advantage is to calculate the Revealed Comparative Advantage (RCA). RCA indexes were calculated for the period 1980-1990 by Togan (1994). According to the calculated indexes, Turkey's comparative advantage was in explosives, inorganic chemicals, clothing, iron and steel, non-ferrous metals and rubber manufactures, to cite a few. Therefore, for this period, some of the areas where Turkey's comparative advantage lay were labour-intensive sectors (such as clothing), but most were in capital-intensive sectors.

The commodity composition of Turkish exports for 1980-1990 period indicates that Turkey's exports in the above-mentioned sectors increased rapidly. However, the industries excluding iron and steel constituted only a very small percentage in total exports. By the end of 1980s, the textiles sector was still the major export item among manufactured products, which had a high, but falling RCA index value. According to the RCA index values, for the 1980-1990 period, Turkey should have exported more capital-intensive products but like many developed and developing countries, Turkey protected its labour-intensive industries by giving them preferential treatment in terms of export subsidies (Togan, 1994).

Two important conclusions can be drawn from the analysis of comparative advantage. The first conclusion is that comparative advantage is dynamic. As RCA indicates, Turkey has a traditional comparative advantage in textiles, but this advantage is decreasing. The newly emerging export sectors such as iron and steel, chemicals, non-ferrous metals and rubber manufacture had increasing RCA values that indicated increasing comparative advantage. The second conclusion is that comparative advantage is not given, it is made. As stated earlier, government policy reallocated resources from agriculture to the manufactured products industry. This, in return, altered Turkey's comparative advantage from resource-intensive to labour-intensive, and further to capital-intensive sectors.

Geographical Distribution of Exports

For the pre-1980 period, OECD countries were the most important markets for Turkish exports. Exports to OECD countries increased from 430 million US$ in 1970 to 1.4 billion US$ in 1979. The share of exports to the OECD countries amounted to 73% in 1970, 70% in 1975, and 64% in 1979. In 1979, the percentage share of exports to EEC countries was 49%. In the second half of the 1970s, the exports to Middle Eastern and North African countries increased. In 1974, the share of the Middle East was 4% and by 1979 their share reached 14% (Table 1).

Table 1. Geographic distribution of exports, 1970-79.

Geographic Region

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

EEC

60%

62%

61%

67%

47%

44%

49%

50%

50%

49%

other OECD

13%

11%

12%

11%

24%

26%

26%

21%

18%

15%

OECD (total)

73%

73%

73%

78%

71%

70%

75%

71%

68%

64%

Eastern Europe

14%

12%

10%

8%

6%

7%

9%

10%

16%

13%

Middle East

-

-

-

-

4%

12%

8%

14%

12%

14%

Other

13%

15%

17%

14%

19%

11%

8%

5%

4%

9%

TOTAL

100%

100%

100%

100%

100%

100%

100%

100%

100%

100%

Source: Own calculations, OECD Economic Survey of Turkey (various issues).

In the post-1980 period, the share of exports to the Middle East and North African countries continued to increase. However, the OECD countries still accounted for most of the export destinations (Table 2).

Table 2. Geographic distribution of exports, 1980-1995.

Geographic Region

1980

1985

1990

1995

EEC

43%

40%

53%

51%

other OECD

15%

11%

15%

13%

OECD (total)

58%

52%

68%

64%

Eastern Europe

17%

4%

8%

3%

Middle East

22%

41%

19%

15%

Other

3%

3%

5%

18%

TOTAL

100%

100%

100%

100%

Source: Author's calculations, OECD Economic Survey of Turkey, various issues

The number of trading partners is a sign of diversification for a country. As the number of trading partners increases, the country's export portfolio becomes diversified. If a country is trading with only one partner, it is vulnerable to changes in the partner's economy and trade policies. As the geographic destinations of exports widen, the country becomes less vulnerable. It is possible to measure the degree of vulnerability of a country by a 'geographic concentration index', first developed by Hirschman (1945) (Togan, 1994). The Gini-Hirschman coefficient, the index of the country's geographic concentration is calculated as:

Ug = 100 Ö å m (Xk / X)2

Where:

Ug = The Gini-Hirschman coefficient

Xk = The value of exports to country k

X = Total value of the country's exports

m = number of countries

Note: The higher the value of l00Ö (1/m), the more concentrated the country's exports are.

As can be seen in Table 3, Turkey diversified its exports geographically towards the end of the 1970s compared to 1969 and 1970. However, although the number of export markets increased by 32% from 1979 to 1994/1995, exports became less diversified in 1994/1995. This is due to the increased value of exports to the OECD countries.

Table 3. The values of geographic concentration index, 1969-70, 1978-79, 1994-95.

YEAR

Number of Export Markets

Index

1969

65

29.727

1970

66

29.016

1978

69

28.003

1979

69

28.633

1994

91

29.888

1995

91

31.332

Source: Author's calculations using OECD monthly averages, series A.

Intra-Industry Trade

Intra-industry trade is the 'exchange of similar products of a given industry' and can be measured by the Grubel-Lloyd index (Williamson and Milner, 1991). The OECD data for the period 1986-1995 classifies exports and imports into the 10 single-digit categories of the Standard International Trade Classification (SITC) which can be used to calculate the Grubel-Lloyd index using the following equation:

Grubel-Lloyd index = 1- (ç X-Mç / (X+M))

where;

X = The value of exports

M = The value of imports

Table 4 illustrates how the manufactured goods classified as SITC 6 consistently had a high G-L index, an indication of intra-industry trade. Another SITC group indicating high intra-industry trade was beverages and tobacco (SITC 1). Calculating the Grubel-Lloyd index using single digit SITC classification is probably misleading since each classification is aggregated, with many industries classified under one category. Therefore, this table should not be used to prove where intra-industry trade lies. Rather, it should be used to indicate where more detailed analysis should be based. Manufactured goods classified chiefly by material (SITC 6) include leather, leather manufactures, rubber manufactures, textiles yarn, fabrics, glass, glassware, pottery, cement, iron and steel and non-ferrous metals.

Table 4. G-L index for single-digit SITC categories.

SITC classification

1986

1988

1990

1992

1994

1995

0 Food and live animals

0.25

0.14

0.68

0.35

0.25

0.60

1 Beverages and tobacco

0.62

0.81

0.86

0.96

0.49

0.54

2 Crude minerals

0.84

0.73

0.62

0.42

0.42

0.36

3 Mineral fuels

0.16

0.20

0.12

0.12

0.12

0.12

4 Animal and vegetable oil

0.72

0.55

0.64

0.73

0.57

0.68

5 Chemicals

0.41

0.60

0.42

0.34

0.37

0.29

6 Manufactures goods

0.81

0.82

0.93

0.93

0.83

0.97

7 Machinery

0.19

0.30

0.22

0.28

0.37

0.35

8 Miscellaneous manufactures

0.38

0.30

0.40

0.39

0.35

0.41

9 Commodities not classified elsewhere

0.75

0.67

0.48

0.24

0.19

0.57

Source: Own calculation using OECD, Foreign Trade by Commodity, Series C.

Note: Index ranges from 0-1. Higher levels of intra-industry trade are indicated as index approaches 1.

Analysis of more detailed classifications indicates that intra-industry trade is highest in the iron and steel industry (SITC 67), followed by glass and ceramics (SITC 664, 665, 666), non-ferrous metals (SITC 68) and rubber manufactures (SITC 62). Comparatively, intra-industry trade in textiles (SITC 65) remained low over the period. Overall, high intra-industry trade lies in non-traditional and newly emerging exports industries.

The scale economies of iron and steel, rubber manufactures, and glass and ceramics, may be just one cause of intra-industry trade. Krugman (1979) stated that because of scale economies in varieties produced, there was larger output and greater variety with international trade. So, 'economies of scale can be shown to give rise to trade and to gains from trade even when there are no international differences in tastes, technology or factor endowments' (Krugman in Pomfret, 1991). However, given the fact that Turkey's major trading partners are industrialised countries, it is unlikely that the condition of 'same technology of factor endowments' will hold. Turkey does not have the same capital accumulation and production technology as her industrialised trading partners. Therefore, two probable hypotheses remain to explain intra-industry trade: the intermediate goods theory and the neo-factors proportion theory.

Intermediate goods theory relates the Heckscher-Ohlin model to processes rather than to products. The production of most final goods involves a series of intermediate stages of production. Comparative advantage may therefore dictate that the location of the various processes should be in different countries, exploiting different factor endowments (Williamson and Milner, 1991). After import liberalisation, Turkey imported raw materials and new technology, especially in the capital-intensive sectors and exported finished products in the same sector (Baysan and Blitzer, 1991).

The neo-factors proportion theory explains the two-way trade flow between developed countries and developing countries by quality differentiation. Based on this theory, Turkey and the OECD countries may trade the same good. The difference is the proportion of factors of production used to produce the product. Capital-rich OECD countries will use more capital-intensive technology, while Turkey will employ more labour since capital is relatively more expensive to produce this product. Therefore, the same product produced by the OECD countries will be regarded as better quality since its production is more capital-intensive.

Intra-industry trade in such sectors will emerge because developed countries will specialise in a high quality version of the product while the developing nations will specialise in lower qualities of the same product. High-income consumers in the developing country will demand the high quality version of this product. Therefore, developed nations will export their products to developing countries, and in return import a lower quality version of the same product from the developing world.

These hypotheses may explain the intra-industry trade in iron and steel, glass and ceramics, and rubber manufactures. The intra-industry trade in non-ferrous metals could be a mere statistical error due to over-aggregation in the SITC category. The same division (SITC 68) includes silver, platinum, copper, nickel, aluminium, lead, zinc, tin, uranium and its alloys. Exporting silver and importing uranium would erroneously appear as intra-industry trade. The beverages and tobacco (SITC 1) division consists of non-alcoholic and alcoholic beverages as well as manufactured and unmanufactured tobacco. Therefore, a desire for variety, product differentiation and the intermediate goods theory can explain the existence of intra-industry trade for this category.

Turkey's exports increased from 2.9 billion US$ in 1980 to 21.6 billion US$ in 1995 with an average growth rate of 15 percent. The most important result of the liberalisation program has been Turkey's spectacular export performance, especially the growth of exports of manufactured products. The change from agricultural exports to exports of manufactured products utilised Turkey's dynamic comparative advantage perfectly, but within manufactured products, labour-intensive sectors still have the largest share.

Conclusion

In the 1990s, the structure of Turkish exports experienced a period of transition. As capital-intensive industries emerge as important export industries, and capital accumulation increases, Turkey's exports will move towards capital-intensive sectors. The rise of capital-intensive export industries does not mean that labour-intensive export industries will perish. Turkey can sustain its comparative advantage in labour-intensive industries, such as textiles, by value adding. Higher quality, value added labour-intensive products should target countries with higher income elasticities for such products.

Bibliography

Baysan,Tercan and Blitzer, Charles (1991) 'Turkey' in Liberalising Foreign Trade. Papageorgiou, Demetris, Michaely, Micheal and Choksi, Armeane (eds), Basil Blackwell: Cambridge, vol. 6, p. 263.

Baysan, Tercan and Blitzer, Charles (1990) 'Turkey's Trade Liberalisation in the 1980s and Prospects for its Sustainability' in The Political Economy of Turkey. Aricanli, Tosun and Rodrik, Dani (eds), Macmillian Press Ltd: London, p. 9, p. 60.

Geary, Patrick (1982) 'How much inflation is imported?' in Inflation in the Irish Economy: A Contemporary Perspective. McAleese, Dermot and Ryan, Louden (eds), Helicon (for the Confederation of Irish Industry): Dublin, p. 6.

Greenaway, David and Sapsford, David (1996) 'Trade Reform and Changes in Terms of Trade in Turkey' in The Economy of Turkey Since Liberalisation. Togan, Subidey and Balasubramanyam, V.N. (eds), Macmillan Press Ltd: London, p. 52.

OECD (1974/1975) OECD Economic Surveys: Turkey. Paris.

OECD (1978) OECD Economic Surveys: Turkey. Paris.

OECD (1982) OECD Economic Surveys: Turkey. Paris.

OECD (1984/1985) OECD Economic Surveys: Turkey. Paris.

OECD (1986/1987) OECD Economic Surveys: Turkey. Paris.

OECD (1989/1990) OECD Economic Surveys: Turkey. Paris.

OECD (1990/1991) OECD Economic Surveys: Turkey. Paris.

OECD (1993) OECD Economic Surveys: Turkey. Paris.

OECD (1997) OECD Economic Surveys: Turkey, Paris.

OECD (1970) 'Monthly averages, Series A' in OECD: Trade by Geographical Distribution: Paris.

OECD (1979) 'Monthly averages, Series A' in OECD: Trade by Geographical Distribution. Paris.

OECD (1996) 'Monthly averages, Series A' in OECD: Trade by Geographical Distribution. Paris.

Pomfret, Richard (1991) International Trade: An Introduction to Theory and Policy. MA, Blackwell: Cambridge, p. 74-88.

Senses, Fikret (1990) in An Assessment of the Pattern of Turkish Manufactured Export Growth in the 1980 and its Prospects. Aricanli, Tosun and Rodrik, Dani (eds).

Taskin, Fatma and Yeldan, Erinc (1996) 'Export Expansion, Capital Accumulation and Distribution in Turkish Manufacturing, 1980-92' in The Economy of Turkey since Liberalization. Togan, Subidey and Balasubramanyam, V.N. (eds), Macmillan Press Ltd: London, p. 155.

Togan, Subidey (1994) Foreign Trade Regime and Trade Liberalization in Turkey During 1980s. Avebury Ashgate Publishing Ltd: Vermont.

Togan, Subidey (1996) 'Trade Liberalisation on Competitive Structure in Turkey during the 1980s' in The Economy of Turkey since Liberalisation, Togan, Subidey and Balasubramanyam, V.N. (eds), Macmillan Press Ltd: London, p. 5.

Williamson, John and Milner, Chris (1991) The World Economy. New York University Press: New York, p. 62, 90-91.