Abstract

In the face of a global economic slowdown after 1980, countries that maintained or increased growth rates typically expanded trade and deepened their financial systems as well. Growth was set back when countries were unable to resolve multiple claims on fiscal resources and experienced high fiscal volatility, inflation, and rising levels of debt. Where all these factors were favorable, as in East Asia, growth was exceptionally rapid. Turkey benefited from the entrepreneurship made possible through trade and financial sector liberalization in 1980, but fiscal policy instability and inflation constrained its growth. Looking ahead, as the growth potential of Turkey’s trade reforms begins to reach its maximum levels, the nation will need continued structural reforms to spur increased entrepreneurship in a competitive environment, while maintaining a commitment to fiscal policy discipline.

In the face of a global economic slowdown after 1980, countries that maintained or increased growth rates typically expanded trade and deepened their financial systems as well. Growth was set back when countries were unable to resolve multiple claims on fiscal resources and experienced high fiscal volatility, inflation, and rising levels of debt. Where all these factors were favorable, as in East Asia, growth was exceptionally rapid. Turkey benefited from the entrepreneurship made possible through trade and financial sector liberalization in 1980, but fiscal policy instability and inflation constrained its growth. Looking ahead, as the growth potential of Turkey’s trade reforms begins to reach its maximum levels, the nation will need continued structural reforms to spur increased entrepreneurship in a competitive environment, while maintaining a commitment to fiscal policy discipline.

Turkey’s per capita income grew between 1960 and 1980 at about 2.3 percent per year. Although world growth slowed after 1980 from 2.5 to 1 percent annually, Turkey continued to grow at an annual rate of 2.3 percent, thereby improving its performance relative to the world average. An important turning point in Turkish economic history occurred in 1980, as noted by Krueger (2004, p. 2):

The ambitious program that began on January 26, 1980 was intended to mark the start of serious economic reform efforts after years of false starts. A big shake-up in both policymaking and policy signaled a completely fresh approach. The reforms had three main aims: economic stabilization, including a reduction in the rate of inflation; a deliberate shift away from import substitution and towards an export-oriented economy; and a move towards a more market-oriented economy.

Early results from the reforms were favorable, but sources of policy instability and macro uncertainty, which had long been endemic, quickly reasserted themselves and rose to a new level of virulence. Thus, Turkey achieved solid growth but also suffered increasing volatility, giving up hard-won output gains in frequent and severe crises.

Growth was helped especially by trade and financial sector liberalization, which marked an important break with the past and allowed for the vibrant expression of Turkish entrepreneurship. Early introduction of export subsidies was followed by reduction of tariff and nontariff barriers, culminating in the customs union with the European Community in 1996. In the financial sector, deregulation of interest rates was followed by enhancement of accounting standards and the opening up of the capital account in 1989. Trade and credit expansion made possible by the liberalization have been Turkey’s principal sources of growth since 1980, though the growth dividend hit diminishing returns in the 1990s. Turkey’s increasing reliance on manufactured exports brought the added benefit of reducing exposure to external (terms of trade) volatility; in contrast, vulnerabilities in the financial sector have remained a source of concern.

The main source of instability, however, was the stop-go pattern of economic management, reflected in a high level of fiscal discretionary spending, high and rising inflation, and an increasing external debt position. Though inflation fell briefly in the early 1980s, it rose to new highs thereafter, as did volatility in fiscal expenditures. A sharp deterioration in the fiscal position in the 1990s and a continued rise in debt levels eventually ended in the economic crisis and contraction in 2001. In such an environment of uncertainty, domestic investment and productivity growth were held back.

This chapter places Turkey’s growth experience in an international context, comparing the growth achievements of a cross-section of countries between 1980 and 2000 with their growth in the preceding two decades. The regression results do not prove causation but should be viewed as an attempt to approximately quantify the contribution to growth of the two opposing forces of liberalization and macro instability. The analysis suggests that the key challenge for Turkey is to continue to increase the scope for private entrepreneurship through structural reforms, while instituting fiscal discipline to reduce macro uncertainty and hence spur further increases in productivity and investment.

The chapter first highlights key features of Turkey’s economic performance before examining the correlates of growth, placing Turkey in a comparative perspective. The main quantitative results based on cross-country growth analyses are then presented, including the decomposition of growth into total factor productivity growth and investment growth. Finally, the chapter briefly examines the sources of macro instability and asks, in particular, whether economic liberalization generates domestic conflicts over limited resources that result in volatility, ultimately hurting growth.

Recent Growth and Volatility Trends

Over the past several decades, Turkey has struggled to raise—and then keep—its per capita income above the world average. Turkey came close to the global average as early as 1976 at the peak of a short-lived growth spurt, but the effort collapsed during the debt crisis of 1979–80 (Figure 2.1). Adopting a series of major reforms, the country then embarked on an apparently more sustained trend toward closing the gap with the global economy. In 1988, Turkey’s level of per capita GNP surpassed the global average for the first time in the twentieth century, ultimately exceeding global average GNP by 10 percent in 2000. Unfortunately, this trend ended with yet another growth collapse in 2001, reducing output by about 9 percent and bringing per capita income back to almost precisely the global average.1

Figure 2.1.
Figure 2.1.

Real Per Capita GDP

(In 1990 international Geary-Khamis dollars)

Turkey, therefore, can be characterized over the past quarter century or so as having had above-average growth performance but also a high degree of volatility. At about 2.3 percent per year, Turkey’s per capita income grew somewhat faster than world per capita income between 1960 and 2000. Turkish performance far surpassed that of sub-Saharan Africa, where per capita incomes grew at less than 0.6 percent a year over the same period, but Turkey lagged behind East Asian economies, where annual per capita growth was more than one percentage point higher (Table 2.1).

Table 2.1.

Real Per Capita GDP Growth, 1960–2000

(Annual percent changes)

article image
Source: Penn World Table Version 6.1 (Heston, Summers, and Aten, 2002).

Notably, Turkey grew at an average rate of 2.3 percent annually not only from 1960–2000, but also during each of the 20-year subperiods (1960–80 and 1981–2000). While Turkey’s growth rate thus remained unchanged in absolute terms over the two subperiods, its relative growth performance actually improved, as world growth slowed in the second period (Figures 2.1 and 2.2). The 90 countries in the sample grew at an average rate of 2.6 percent annually over 1960–80 but at only 1.3 percent annually from 1981–2000. Because Turkish per capita output growth was relatively high after 1980, its per capita output exceeded the global average in 1988 (though that gain was subsequently lost in the 2001 crisis, which is not considered in the analysis below).

Figure 2.2.
Figure 2.2.

Real Per Capita GDP Growth

(Backward-looking 10-year average; in percent)

Though Turkey maintained respectably high growth rates, an important feature of the nation’s growth from 1980 to 2000 was a sharp increase in volatility (Figure 2.3). An increase in volatility by itself is not a sign of concern—many East Asian countries have grown rapidly and experienced brief periods of high volatility. Indeed, Tornell, Westermann, and Martinez (2004) point out that the link between volatility and growth may be positive or negative. In this view, the deregulation of the financial sector is necessary to support higher growth, but financial sector fragilities may also lead to crises. The volatility may thus be a necessary price for achieving high long-term growth. Other research, such as that by Ramey and Ramey (1995), suggests that volatility may have no redeeming feature. An important question, therefore, is whether Turkey could have done better if volatility had been contained. As shown below, the increased volatility in Turkey was related to instability in fiscal and monetary policies that are associated with a growth-reducing effect.

Figure 2.3.
Figure 2.3.

Per Capita GDP Growth Volatility

(Backward-looking five-year standard deviation; in percent)

Growth Accounting

A growth accounting exercise provides insights into the channels through which growth has occurred. Growth accounting decomposes observed output growth into components that can be attributed to the increased use of factor inputs and those due to growth in total factor productivity (TFP). As is conventional, we assume a constant returns-to-scale, Cobb-Douglas production function:

Yt=AtKtα(htLt)1α

where Y is aggregate output, A is the level of technology, or TFP, K is the physical capital stock, h is the human capital stock per worker, L is the number of workers, and α is the elasticity of output with respect to physical capital. Expressing all variables in per-worker terms (denoted by small caps), taking log differences across time and omitting time indices, we obtain:

Δlog(y) = Δlog(A) + αΔlog(k) + (1–α)Δlog(h)

or

Δlog(A) = Δlog(y)–αΔlog(k)–(1–α)Δlog(h)

In Turkey and all comparator groups, the worldwide growth slowdown after 1980 was accompanied by a significant slowdown in TFP growth (Table 2.2),2 although Turkey’s TFP decline was less pronounced. Physical capital growth, which decreased in all of the comparator groups, increased in Turkey, consistent with the observed increase in Turkish investment rates. Overall, then, Turkey’s relative (and absolute) growth performance improved due to a more moderate slowdown in overall factor productivity, and to a rise in physical capital accumulation. Human capital appears to have grown in line with cross-country experience.

Table 2.2.

Growth Accounting, 1961–2000

(Annual percent changes)

article image
Note: The table covers 73 countries from the sample for which Bosworth and Collins (2003) provide physical and human capital stock data. GDP data are from Penn World Table Version 6.1; total factor productivity (TFP) is calculated as the residual, as described in the text.

The remainder of this chapter considers the factors that were correlated with growth, then returns to the growth accounting exercise to examine whether these correlates worked through the productivity or investment channels.

Correlates of Growth

This section describes shifts in the overall macro and policy environment between the periods 1960–80 and 1980–2000, which are then the basis of a more formal cross-country regression analysis in the next section. Of particular interest is the identification of variables that proxy policy changes that have sustained Turkey’s growth and those that have been unfavorable to growth.3

Turkey experienced a substantial increase in the degree of trade openness and, consequently, in its trade-to-GNP ratio. Though the initial focus was on increased subsidies for exports, import barriers were gradually reduced starting in 1984, resulting in a substantial decline in both tariff and nontariff barriers by the mid-1990s (Krueger and Aktan, 1992; and Ozler and Yilmaz, 2004). In January 1996, Turkey reinforced its commitment to open trade by forming a customs union with the European Community (Erzan, Filiztekin, and Zenginobuz, 2003). Turkey adopted the European Union’s common external tariff (about 4 percent for manufactured goods) against third country imports in 1996 and the preferential agreements the EU has concluded with third countries by 2001. At the same time, the EU, which had eliminated tariffs on most Turkish manufactures, committed to abolish quotas, particularly on textiles and clothing, and gradually on iron and steel (although use of antidumping legislation continues to be a source of nontariff barriers).

These measures resulted in a threefold increase in the trade ratio, from under 12 percent to about 38 percent between the first and second period considered in this chapter (Table 2.3). Figure 2.4 shows that the trade ratio had begun to level off following the initial round of reforms and was given a fillip in the lead up to and the period following entry into the EU customs union. Krueger (2004, p. 2) has concluded that “… the reforms introduced succeeded in shifting Turkey permanently towards a more export-oriented economy.”

Table 2.3.

Summary Statistics, Selected Variables

article image
Figure 2.4.
Figure 2.4.

External Trade

(In percent of GDP)

Most countries became more outward-oriented after 1980, as highlighted by the averages in Table 2.3. Turkey’s experience was, therefore, consistent with a more generally observed increase in trade flows following liberalization of trade policy (Wacziarg and Welch, 2003). However, Turkey’s performance was nevertheless remarkable, with the increase in its trade ratio representing the biggest relative increase in the sample.

In theory, more openness also makes a country more vulnerable to external shocks (Rodrik, 1999). However, Turkey’s terms of trade volatility declined in the 1980s and 1990s—much of the export growth was from the manufacturing sector, which increased as a share of Turkish exports from about 36 percent in 1980 to over 90 percent in 2000—reducing the scope of macroeconomic instability that may stem from a volatile trading environment. Instead, investment as a share of GNP increased. Once again, the cross-country empirical evidence presented in Wacziarg and Welch (2003) shows that as countries have expanded trade flows, they also have stepped up their investment. Note, though, that despite these increases, investment shares in Turkey remained relatively low, especially compared with East Asian and Organization for Economic Cooperation and Development (OECD) economies, suggesting the scope for a further increase in investment rates.

As will be shown below, financial development also has had a generally positive influence on growth, spurred by liberalization initiated in 1980. In July of that year, ceilings were lifted on interest rates that banks paid and could charge. In 1985, protection of deposits through deposit insurance, standardization of accounting, and norms for the treatment of nonperforming loans were introduced (Sancak, 2002). The M2/GDP ratio, which had been flat to trending down in the 1970s, rose rapidly following the liberalization (Figure 2.5).4

Figure 2.5.
Figure 2.5.

M2/GDP Ratio

(In percent)

On the down side, macro instability increased, reflected in different forms. In general terms, the source of the instability was an untenable demand on fiscal resources. The most notable development after 1980 was a marked increase in fiscal discretionary volatility (Figure 2.6), a constructed measure reflecting unexpected changes in fiscal expenditure that are inconsistent with past policies or changes in the economic environment.5 Specifically, using annual data over 1960–2000 for each country in the sample,6 we estimate the equation:

Figure 2.6.
Figure 2.6.

Fiscal Discretionary Volatility

(Backward-looking five-year standard deviations; in percent)

ΔGi,t=αi+βiΔYi,t1+γiΔYi,t2+δiΔGi,t1+ηiWi,t+εi,t

where G is the logarithm of real government expenditures, Y is the logarithm of real GNP in constant local currency units, and W is a vector of controls, including terms of trade growth, inflation, inflation squared, and a time trend. The measure σif of fiscal volatility is defined as the standard deviation of the residual of the above regression equation. The intention is to abstract from fiscal responses to the economic cycle and other macroeconomic conditions, such as inflation and external developments. By also including lagged government expenditure growth, the measure of fiscal policy discretion can be interpreted as representing unexpected, cyclically adjusted deviations from a country’s past fiscal policy stance that are unrelated to macroeconomic conditions.

Turkey’s fiscal volatility increased notably at a time when such volatility was declining in other economically important regions. That this is not a purely statistical artifact is supported by analysis of the country’s budgetary processes, which shows not only the inability to restrain aggregate deficits but also lack of transparency and the use of extrabudgetary devices to permit a high rate of discretionary expenditures (Celasun, 1990; and World Bank, 2001). Fiscal volatility has been shown to be a key contributor to overall growth volatility and might thus explain why Turkey’s growth has become more volatile while growth has become more stable in most other countries (as noted above in the discussion of Figure 2.2). Moreover, several studies (Fatás and Mihov, 2003; and Mody and Schindler, 2004) have shown fiscal discretionary volatility to be “harmful” volatility, with a robust, negative effect on growth. That negative relationship is also confirmed in the econometric analyses presented below.

The period from 1981–2000 was characterized by a large average budget deficit and significantly higher inflation—over 46 percent annually. While moderate inflation is typically found to have no significant correlation with growth, high inflation tends to be associated with lower growth, and inflation in Turkey rose sharply over the past several decades (Figure 2.7). Lastly, Turkey’s external debt increased rapidly, much more so than in the average country (Figure 2.8).7

Figure 2.7.
Figure 2.8.
Figure 2.8.

External Debt

(In percent of GDP)

Note that Turkey managed to maintain the size of government at a constant level, not only lower than the sample average, but also avoiding the increased volume of government spending most other countries exhibited. Thus, it was not changes in the size of the government per se that were harmful for growth, but rather a decreased predictability of expenditure growth as well as the deficits that were reflected in inflation.

Cross-Country Regression Analysis

The reasons behind the change in Turkey’s relative position between 1960–80 and 1981–2000 can be determined by regressing changes in average growth in the subperiods on changes in key policy variables. The main variables identified are fiscal volatility, inflation, and changes in trade openness. Convergence effects are controlled for by including the difference of the two subperiods’ initial per capita GNP (that is, growth in the first subperiod).

First, however, an alternative growth explanation warrants mention. In an influential study, Rodrik (1999) argues that in recent decades, countries exposed to external volatility were especially likely to suffer from growth slowdowns—more so in divided societies, especially if they had weak internal institutions to manage the conflicts that arise in conditions of economic volatility. If this were the case, Turkey’s drive to open its economy to more trade would have had a harmful effect. Although Turkey has a very high level of inequality, its external volatility was never high and, importantly, declined after 1980 as the country became an exporter of manufactured goods. To pursue these issues in more detail, a composite measure is constructed—referred to as “conflict” in analogy to Rodrik (1999)—as the product of external trade volatility times trade openness (1981–2000 averages) times the 1960–2000 average of income inequality times 100*(1-ICRG1982), divided by the composite measure’s standard deviation.

Column (1) of Table 2.4 shows Rodrik’s finding: the conflict variable is indeed negatively related to growth. However, a substantial and statistically significant fraction of Turkish growth remains unexplained (as indicated by the Turkish “dummy” variable). More importantly, by its construction, interacting four different variables, the conflict measure implicitly assumes that trade openness is harmful. When trade openness is entered by itself in column (2), it is positively related to growth and the conflict measure is insignificant. When the fiscal volatility measure is also included in column (3) and inflation in column (4), the conflict measure further loses statistical significance, with fiscal volatility, inflation, and trade openness all statistically relevant descriptors of growth and the Turkish dummy statistically insignificant.8 The conclusion is that the conflict measure has little importance in the case of Turkey. In the following section, the specification in column (5) is thus taken out as the starting point for examining other potential growth correlates.

Table 2.4.

Cross-Country Growth Difference Regressions

(Dependent variable is GDP growth difference)

article image
Note: Numbers in parentheses are robust p-values. Significance levels are *10%; **5%; ***1%. All regressions include a constant. All variables except conflict are period differences. The conflict variable is constructed as in Rodrik (1999).

Key Growth Results

Table 2.5 presents the main results, especially in column (6). All key variables are highly significant, except the dummy for Turkey, meaning that Turkey’s change in growth is well explained. The partial scatter plots in Figure 2.9, based on the regression specification in column (6), confirm that the results are not being driven by a few outliers. The results lend support to the hypothesis that Turkey underwent conflicting changes in the period after 1980. The trade share increased dramatically and provided the main impetus to growth. At the same time, financial sector development also helped. These positive influences were weakened by deterioration in fiscal and monetary discipline, measured by greater fiscal volatility and high inflation.

Table 2.5.

Additional Cross-Country Growth Difference Regressions

(Dependent variable is GDP growth difference)

article image
Note: Numbers in parentheses are robust p-values. Significance levels are *10%; **5%; ***1%. All regressions include a constant. All variables are period differences.
Figure 2.9.
Figure 2.9.

Partial Scatter Plots of the Growth Differential Against Differentials in Selected Variables

These main results are robust to the inclusion of a number of other variables. In particular, an increase in external volatility did not lower growth. While an increase in the size of government was related to a decrease in growth, the effect was not significant.

From a technical point of view, the regressions are simple ordinary least squares (OLS) and as such present correlations, not causal relationships. However, given its construction, reverse causation is less likely in the case of the fiscal volatility measure, as Fatás and Mihov (2003) also argue. Also, while trade flows could have increased in response to an increase in growth, rather than the other way around, the results of Wacziarg and Welch (2003) suggest that trade flows are influenced by the policy of trade liberalization. For Turkey, the timing of the increase in trade flows coincides well with liberalization. Nevertheless, the results are interpreted here as merely approximations of the quantitative contributions to growth of the two opposing forces of liberalization and macro instability.

Decomposing Growth

Overall, at .0313 percentage point, there was essentially no change in Turkey’s growth over the two sub-periods. But referring to column (6) in Table 2.5, the change in Turkey’s growth can be decomposed into its various sources9 in order to examine how the various factors interacted in keeping aggregate growth unchanged. In order to do this, the regression in column (6) in Table 2.5 is repeated without the Turkey dummy and the resulting coefficients are applied to the actual variable outcomes for Turkey. The results are summarized in Table 2.6. Focusing on the four policy variables, the main growth-reducing factors were (i) the increase in fiscal volatility, associated with a growth reduction of .48 percentage points, and (ii) the rise in inflation, associated with reduced growth of .7 percentage point. The main growth-enhancing factors were (i) the sharp increase in external trade, associated with an increase in growth by 2.35 percentage points, and (ii) financial development, proxied by growth in the M2/GDP ratio, associated with an increase in growth of .46 percentage point.

Table 2.6.

Decomposition of Turkey’s 20-Year Growth Differential

article image

The coefficients are based on the specification (6) in Table 2.5 excluding the Turkey dummy, with virtually identical R-squared and levels of significance.

Trade therefore was the main component in keeping Turkey’s growth constant, as opposed to decreasing as the global average did. If Turkey had also kept inflation and fiscal volatility at the 1960–80 levels, it would have grown almost 1.2 percentage points faster, or about 3.5 percent annually during 1981–2000, about the same rate as the rapidly growing East Asian economies. In cumulative terms, Turkey’s GNP would have then been about 27 percent higher in 2000 than it actually turned out to be. In contrast, if macro instability had increased as it did, but trade and financial sector reforms had not occurred, Turkish growth could have been below that of sub-Saharan Africa.

To consider the possibility that the dynamics of growth changed in the 1990s, the sample was divided into decadal averages and the same regression as above was run for decadal growth differences. The results are consistent with those from the 20–year difference regressions, but allow for decomposing the sources of growth on a decadal basis, thereby providing a richer picture of growth developments in Turkey. With three observations per country (for three changes in decadal growth rates between 1960 and 2000), a panel regression was run that included time dummies for the 1980s and the 1990s. The coefficients obtained were then applied to Turkey’s actual outcomes of the variables under consideration. Table 2.7 summarizes the calculations.

Table 2.7.

Decomposition of Turkey’s Decadal Growth Differentials

article image

The coefficients were estimated by regressing changes in decadal growth on decadal changes in the explanatory variables plus time dummies. The adjusted R-squared was .425, and all coefficients except the dummy for the 1990s were significant at the 5 percent level or better.

Growth in trade had overall positive effects on growth, but those effects were most pronounced in the 1980s, when the trend in trade growth appears to have permanently increased. The regress in financial development in the 1970s reduced growth, but its recovery, mostly in the 1980s but also the 1990s, accounted for a significant increase in growth.

While fiscal volatility has a negative effect on the level of growth, the reduction in Turkey’s fiscal volatility in the 1990s helped improve growth outcomes relative to the 1980s. However, as Table 2.6 shows, that improvement was not sufficient to overcome the negative effects over 1981–2000 as a whole. Ever-increasing inflation during the three time periods continuously reduced growth. Overall, the results appear to support the assumption that the 1980s were a structural break period, with big growth effects during the decade but smaller growth effects afterwards.

Channels of Growth

Linking the cross-country regressions back to the growth accounting exercise requires examining the channels through which the growth determinants acted. Table 2.8 shows that total factor productivity growth was hurt mainly by high inflation and helped by financial development. Investment was reduced in countries with high fiscal volatility. Thus, while the determinants of foreign direct investment (FDI) are not explicitly considered, these results offer a perspective on the low levels of FDI in Turkey.

Table 2.8.

Cross-Country Difference Regressions with Assorted Dependent Variables

article image
Note: Numbers in parentheses are robust p-values. Significance levels are *10%; **5%; ***1%. All regressions include a constant. GDP and total factor productivity are differences in growth rates, INV is the difference in the log of the investment share.

Presumably, macro instability hurt both domestic and foreign investment.10 Investment increased, however, in the context of more open trade flows, consistent with evidence reported by Wacziarg and Welch (2003). The findings here are inconsistent with micro studies (Ozler and Yilmaz, 2004) that find that TFP increases more rapidly during periods of trade liberalization, although the finding here is that trade liberalization has a stronger impact through investment.

Sources of Policy Volatility

The analysis above leads to the conclusion that further liberalization to improve economic incentives for entrepreneurship, when combined with efforts to reduce fiscal indiscipline, should dramatically improve Turkey’s growth outlook. Such an optimistic conclusion would not be warranted if liberalization were accompanied by economic losses for some who could make a political claim on scarce fiscal resources. The consequent effort to make them whole might then have led to bursts of discretionary fiscal expenditures, which would have increased the fiscal deficit while also generating macro volatility. It is possible, for example, that Turkey’s remarkable trade liberalization was possible in the first place only through increased discretionary fiscal policy to compensate those who lost from increased openness (several authors have suggested this, especially Waterbury, 1992). While it is not possible to fully explore this question here, the determinants of fiscal volatility can be briefly addressed.

Table 2.9 shows that fiscal volatility is high principally in countries with weak political checks and balances. Thus, more political constraints on executive authority reduce fiscal volatility. Similarly, a presidential form of government, traditionally thought to be more decisive in actions but less constrained, is associated with higher volatility. Notice, though, that the dummy variable for Turkey in the more volatile second period is positive and significant when these conventional measures of political checks and balances are used. Indeed, when the dummy variable for the presidential form of government is added, the Turkey dummy increases in size. The implication is that, although Turkey has had a parliamentary form of government, it has not benefited from the checks and balances inherent in such a system.

Table 2.9.

Cross-Country Fiscal Volatility Regressions

(Dependent variable is log discretionary fiscal volatility)

article image
Note: Numbers in parentheses are robust p-values. Significance levels are *10%; **5%; ***1%. All regressions include a constant.

Political risk component from the International Country Risk Guide (ICRG) index average from 1984–2000.

For the second period, 1980–2000, there is also a measure of political risk,11 and the finding is that more political instability is associated with more fiscal volatility. Moreover, the Turkish dummy becomes statistically insignificant. Thus, although constitutional arrangements and political institutions create a legal basis for controlling excessive executive discretion, the workings in practice of the relevant institutions limit the scope for exercising restraint. Several factors may be at work here. Önis (2004) argues that in the interest of moving rapidly ahead with much needed reforms, there was a tendency in Turkey to adopt a Latin American presidential style of decision making, disregarding the legal checks and balances and creating ad hoc bureaucratic structures that weakened the traditional bureaucracy and introduced possibilities for political corruption. At the same time, weak governments consisting of multiparty coalitions and facing frequent elections also had the incentive to patronize their electoral supporters and abandon fiscal discipline. In particular, the agricultural sector and wage earners had to be repeatedly compensated for electoral advantage (Waterbury, 1992).

If trade openness is used as a proxy for the broader liberalization effort undertaken since 1980, it is not associated by itself with higher fiscal volatility. However, once political instability is controlled for, more trade openness since 1980 is associated with greater fiscal discretion and volatility. Thus, controlling for the basic political decision-making structure, reflected in the nature of constraints on executive power, there is some plausibility to the idea that “paying off” constituencies that did not directly benefit from trade was necessary to sustain liberalization. This idea is related to—but different from—one suggested by Rodrik (1999), who argues that more openness creates the need to shield the economy from external shocks and hence requires a bigger government. The suggestion here is instead that more openness may actually bring growth benefits to some but losses to others who need to be compensated.

Looking ahead, then, the lesson is not to restrict the scope for economic liberalization but to strengthen the political institutions that limit harmful discretion. In this regard, Turkey appears to be on a positive trend, including institutional developments as measured, for example, by political constraints. Additional steps in the form of fiscal rules and other measures that impose self-discipline could help in the long term.

Conclusions

Over the past four decades, Turkey experienced several stretches of rapid growth but then relinquished its gains during periodic crises. The result was a solid rather than what could have been a spectacular growth performance. The Turkish story demonstrates the country’s entrepreneurial potential for significant growth. Trade and financial sector liberalization in 1980 opened up new opportunities to which the private sector responded energetically, creating a modern and, in some instances, world-class industrial base. These reforms, and the resulting dramatic increase in external trade, were thus key to helping Turkey avoid the secular decline in growth rates that affected most of the rest of the world. Absent these liberalization initiatives, Turkey would likely have grown at rates similar to those of the worst performers in the sample, the sub-Saharan countries.

However, Turkey could have done better than it did. Without deterioration in fiscal and monetary management, Turkey might well have reached growth rates on the order of those found among the best performers in the sample, the rapidly growing East Asian economies.

For Turkey, the crucial question now is whether the forces generating growth can be retained—and bolstered—while those uncertainties inimical to growth are scaled back. Arguably, while the private sector responded in the manner expected by the architects of the reform, the architects’ vision of the role of the government proved to be misguided. It was expected that competing demands on limited fiscal resources would be resolved by higher growth and, to this end, the government would further contribute to growth through public investments (Önis and Reidel, 1993). Under this benign vision of the operation of the government, it was considered appropriate to incur debt, which would be paid off by higher rates of growth (Anand, Chhibber, and van Wijnbergen, 1988). However, in practice, government expenditures appear to have been ultimately directed in large measure by the short electoral cycles and the need to maintain the constituent base in the context of weak coalition governments.

The lesson is that sustained growth will require continued structural reforms—including those related to the financial system, corporate governance, and labor markets—to create new avenues for private entrepreneurship as the impetus to growth from earlier liberalization tapers off. At the same time, limiting macro uncertainties will remain crucial. The prognosis on this latter score appears to be an optimistic one, reflected in the recent steady decline in inflation rates. Maintaining this commitment will ultimately require political discipline, which may be helped by fiscal and monetary rules that are not easy to reverse.

1

Turkey’s per capita GNP was $6,033 in 2001, measured in 1990 international Geary-Khamis dollars, compared with a world average of $6,049 (Maddison, 2001).

2

The physical and human capital stock data underlying the growth accounting exercise are from Bosworth and Collins (2003). The GNP series is from Penn World Table Version 6.1 (Heston, Summers, and Aten, 2002).

3

All data are from the World Bank World Development Indicators database, except real GDP, which is from Penn World Table Version 6.1 (Heston, Summers, and Aten, 2002), schooling (from Barro-Lee, 2000), political constraints (from Fatás and Mihov, 2003), and the International Country Risk Guide measure (from Bosworth and Collins, 2003).

4

Despite its important contribution to growth, the financial sector still has significant weaknesses. The banking system is highly concentrated. In the past, weak banking practices, poor regulation, and lack of corporate governance made the banking system vulnerable. The operations of state banks posed particular difficulties. Macro and structural reforms are not unrelated. For example, Alper and Önis (2004) argue that endemic budget deficits, government ownership in the banking system, and failure to provide proper banking system regulation are intimately linked.

5

See Fatás and Mihov (2003) for a similar construction of a measure of discretionary fiscal volatility.

6

The cross-country sample consists of 90 countries. However, because country coverage of many variables is not complete, sample sizes differ across regression specifications.

7

Several studies, most notably Pattillo, Poirson, and Ricci (2004), have found external debt levels to have a negative (if nonlinear) impact on growth. Because the available external debt data are fairly limited, however, they are not included in the regressions.

8

The data also show that the correlation coefficient between “conflict” and “external shocks” is .9, suggesting that the conflict variable is mostly picking up external shocks (terms of trade volatility multiplied by the trade share), with little variation stemming from institutions and inequality or their interactions with external shocks.

9

The unexplained part, represented by the Turkey dummy, is very small and insignificant and we therefore ignore it.

10

Dutz, Us, and Yilmaz (2005) argue that while macroeconomic instability was the single most important deterrent to foreign direct investment, the perception by foreigners that they may not face a level playing field also limited FDI. They suggest that strengthening the competition policy regime and increasing transparency in bureaucratic processes might help.

11

The political risk measure can take on values between 0 and 100, whereby higher values are associated with lower political risk.

  • Alesina, Alberto, and Roberto Perotti, 1996, “Fixed Adjustments in OECD Countries: Composition and Macroeconomic Effects,” IMF Working Paper 96/70 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Allen, Mark, Christoph Rosenberg, Christian Keller, Brad Setser, and Nouriel Roubini, 2002, “A Balance Sheet Approach to Financial Crisis,” IMF Working Paper 02/210 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Alper, C. Emre, and Ziya Önis, 2004, “The Turkish Banking System, Financial Crises and the IMF in the Age of Capital Account Liberalization: A Political Economy Perspective,” New Perspectives on Turkey No. 30 (Istanbul: History Foundation of Turkey).

    • Search Google Scholar
    • Export Citation
  • Anand, Ritu, Ajay Chhibber, and Sweder van Wijnbergen, 1988, “External Balance, Fiscal Policy, and Growth in Turkey,” World Bank Policy, Planning, and Research Working Paper 86 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Barro, Robert J., and Jong-Wha Lee, 2000, “International Data on Educational Attainment: Updates and Implications,” CID Working Paper No. 42 (Cambridge, MA: Center for International Development at Harvard University).

    • Search Google Scholar
    • Export Citation
  • Blanchard, Oliver, 2004, “Fiscal Dominance and Inflation Targeting: Lessons from Brazil,” NBER Working Paper No. 10389 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Bohn, Henning, 1998, “The Behavior of U.S. Public Debt and Deficits,” Quarterly Journal of Economics, Vol. 113 (August), pp. 94963.

    • Search Google Scholar
    • Export Citation
  • Bosworth, Barry, and Susan Collins, 2003, “The Empirics of Growth: An Update,” Economics of Developing Countries Papers (September) (Washington: Brookings Institution).

    • Search Google Scholar
    • Export Citation
  • Celasun, Merih, 1990, “Fiscal Aspects of Adjustment in the 1990s,” in The Political Economy of Turkey: Debt, Adjustment, and Sustainability, ed. by Tosun Aricanli and Dani Rodrik (London: Macmillan).

    • Search Google Scholar
    • Export Citation
  • Celasun, Oya, R., Gaston Gelos, and Allesandro Prati, 2003, “Would ‘Cold Turkey’ Work in Turkey?IMF Working Paper 03/49 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Celasun, Oya, R., Gaston Gelos, and Allesandro Prati, 2004, “Obstacles to Disinflation: What is the Role of Fiscal Expectations?Economic Policy, Vol. 19, No. 40 (October), pp. 44181.

    • Search Google Scholar
    • Export Citation
  • Cohen, Daniel, 2000, “The HIPC Initiative: True and False Promises,” Organization for Economic Cooperation and Development Centre Working Paper No. 166 (Paris: OECD Development Centre).

    • Search Google Scholar
    • Export Citation
  • Dutz, Mark, Melek Us, and Kamil Yilmaz, 2005, “Turkey’s Foreign Direct Investment Challenges: Competition, the Rule of Law, and EU Accession,” in Turkey: Towards EU Accession, ed. by Bernard Hoekman and Subidey Togan (Oxford: Oxford University Press for the World Bank).

    • Search Google Scholar
    • Export Citation
  • Easterly, William, 2001, “The Lost Decades: Developing Countries’ Stagnation in Spite of Policy Reform, 1980–1998,” Journal of Economic Growth, Vol. 6, No. 2, pp. 13557.

    • Search Google Scholar
    • Export Citation
  • Erzan, Refik, Alpay Filiztekin, and Unal Zenginobuz, 2003, “Impact of the Customs Union with EU on Turkish Manufacturing Industry,” paper presented at the ninth annual conference of the Economic Research Forum, United Arab Emirates, October 2628.

    • Search Google Scholar
    • Export Citation
  • Fatás, Antonio, and Ilian Mihov, 2003, “The Case for Restricting Fiscal Policy Discretion,” Quarterly Journal of Economics, Vol. 118, No. 4 (November), pp.141947.

    • Search Google Scholar
    • Export Citation
  • Favero, Carlo, 2002, “How Do European Monetary and Fiscal Authorities Behave?Center for Economic Policy Research Discussion Paper No. 3426 (London: Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation
  • Favero, Carlo, Francesco Giavazzi, 2004, “Inflation Targeting and Debt: Lessons from Brazil,” Center for Economic Policy Research Discussion Paper No. 4376 (London: Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation
  • Gali, Jordi, and Mark Gertler, 1999, “Inflation Dynamics: A Structural Econometric Analysis,” Journal of Monetary Economics, Vol. 44, Issue 2, pp. 195222.

    • Search Google Scholar
    • Export Citation
  • Gali, Jordi, and Mark Gertler, Roberto Perotti, 2003, “Fiscal Policy and Monetary Integration in Europe,” Economic Policy, Vol. 18 No. 37 (October), pp. 53572.

    • Search Google Scholar
    • Export Citation
  • Garcia, Marcio, and Roberto Rigobon, 2004, “A Risk Management Approach to Emerging Market’s Sovereign Debt Sustainability with an Application to Brazilian Data,” NBER Working Paper No. 10336 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Giavazzi, Francesco, 2003, “Inflation Targeting and Fiscal Policy Regime: The Experience of Brazil,” Bank of England Quarterly Bulletin (Autumn).

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, 2003, “Debt Sustainability, Brazil, and the IMF,” IIE Working Paper 03-1 (Washington: Institute for International Economics).

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, Philip Turner, 2004, Controlling Currency Mismatches in Emerging Markets (Washington: Institute of International Economics).

    • Search Google Scholar
    • Export Citation
  • Gupta, Sanjeev, Emanuele Baldacci, Benedict J. Clements, and Erwin R. Tiongson, 2003, “What Sustains Fiscal Consolidations in Emerging Market Countries?IMF Working Paper 02/224 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Heston, Alan, Robert Summers, and Bettina Aten, 2002, Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), 2003, “Public Debt in Emerging Market Economies: Is It Too High?in World Economic Outlook September, 2005 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Krueger, Anne O., 2004. “Pursuing the Achievable: Macroeconomic Stability and Sustainable Growth in Turkey,” speech to the Economic Congress of Turkey, Izmir, May 5, 2004, www.imf.org/external/np/speeches/2004/050504.htm.

    • Search Google Scholar
    • Export Citation
  • Krueger, Anne O., Okan H., Aktan, 1992, Swimming Against the Tide: Turkish Trade Reforms in the 1980s (San Francisco: International Center for Economic Growth, ICS Press.

    • Search Google Scholar
    • Export Citation
  • Kruger, Mark, and Miguel Messmacher, 2004, “Sovereign Debt Defaults and Financing Needs,” IMF Working Paper 04/53 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Maddison, Angus, 2001, The World Economy: Historical Statistics (Paris: Organization for Economic Cooperation and Development Centre).

  • Manasse, Paolo, Nouriel Roubini, and Axel Schimmelpfennig, 2003, “Predicting Sovereign Debt Crisis,” IMF Working Paper 03/221 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Mody, Ashoka, and Martin Schindler, 2004, “Argentina’s Growth: A Puzzle?” (unpublished; Washington: International Monetary Fund).

  • Önis, Ziya, 2004, “Turgut Ozal and His Economic Legacy: Turkish Neo-Liberalism in Critical Perspective,” Middle Eastern StudiesVol. 40, No. 4.

    • Search Google Scholar
    • Export Citation
  • Önis, Ziya, James Riedel, 1993, Economic Crises and Long-Term Growth in Turkey, World Bank Comparative Macroeconomic Studies, Vol. 1 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Ozler, Sule, and Kamil Yilmaz, 2004, “Does Trade Liberalization Improve Productivity? Plant Level Evidence from Turkish Manufacturing Industry” (unpublished; Istanbul: Koc University).

    • Search Google Scholar
    • Export Citation
  • Pattillo, Catherine, Helene Poirson, and Luca Ricci, 2004, “Through What Channels Does External Debt Affect Growth?in Brookings Trade Forum 2003, ed. by Susan M. Collins and Dani Rodrik (Washington: Brookings Institution), pp. 22977.

    • Search Google Scholar
    • Export Citation
  • Political Risk Services Group, 2003, International Country Risk Guide, www.ICRGOnline.com.

  • Ramey, Garey, and Valerie A. Ramey, 1995, “Cross-Country Evidence on the Link Between Volatility and Growth,” American Economic Review, Vol. 85, No. 5, pp. 113851.

    • Search Google Scholar
    • Export Citation
  • Reinhart, Carmen, Kenneth Rogoff, and Miguel Savastano, 2003, “Debt Intolerance,” NBER Working Paper No. 9908 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Rodrik, Dani, 1999, “Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses,” Journal of Economic Growth, Vol. 4, No. 4, pp. 385412.

    • Search Google Scholar
    • Export Citation
  • Sancak, Cemile, 2002, “Financial Liberalization and Real Investment: Evidence from Turkish Firms,” IMF Working Paper 02/100 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Sargent, Thomas, and Neil Wallace, 1981, “Some Unpleasant Monetarist Arithmetic,” Federal Reserve Bank of Minneapolis Quarterly Review, Vol. 5, No. 3 (Fall).

    • Search Google Scholar
    • Export Citation
  • Shiller, Robert, 1997, “Public Resistance to Indexation: A Puzzle,” Brookings Papers on Economic Activity, No. 1, pp. 159228.

  • Talvi, Ernesto, and Carlos Végh, 2000, “Tax Base Variability and Procyclical Fiscal Policy,” NBER Working Paper No. 7499 (Cambridge, MA: National Bureau for Economic Research).

    • Search Google Scholar
    • Export Citation
  • Tornell, Aaron, and Philip Lane, 1999, “The Voracity Effect,” American Economic Review, Vol. 89, No. 1 (March), pp. 2246.

  • Tornell, Aaron, and Philip Lane, Frank Westermann, and Lorenza Martinez, 2004, “The Positive Link Between Financial Liberalization, Growth and Crises,” NBER Working Paper No. 10293 (Cambridge, MA: National Bureau of Economic Research).