Sources of Growth and the Output Gap for the
Turkish Economy
Kivilcim Metin O¨zcan, U¨mit O¨zlaleand C- ag˘ri Sarikaya
Abstract
In this study, we analyze the growth dynamics of the Turkish economy. After investigating the sources of growth and comparing both the output and the inflation performance with the other MENA countries, we find that the Turkish economy has not achieved a sustainable growth path and has a much worse inflation per-formance. Then, we take a closer look at the volatile output performance and discuss the role of the wrong macroeconomic policies on the unsustainable output performance. Utilizing an output gap measure, we find that the short-term capital flows seem to be the main driving force in generating excessive fluctuations in the output. Based on these findings, we can say that the high economic growth rates, which the Turkish economy has achieved in the recent years owe much to massive short-term capital inflows. At this stage, it is too early to evaluate the effects of the recent structural reforms on the long-term output dynamics.
JEL classifications: C1, E3, O4
Keywords: Turkish economy, emerging markets, output gap, economic growth
1. Introduction
Despite several stabilization and structural adjustment programs, all of which were backed by the World Bank and the International Monetary Fund (IMF), the Turkish economy can be characterized as having an unsustainable output growth and an inflationary environment over the last four decades. Similar to other emerging market economies, a boom–bust cycle on the output dynamics can be observed. After the collapse of the exchange rate-based stabilization program in 2001, which was followed by the deepest financial crisis that Turkey has ever experienced, the economy has started to recover with very high eco-nomic growth rates. Moreover, unlike the previous recovery periods, the price stability objective of the monetary authorities seems to have been achieved, with inflation rates being under 10% as of mid-2005.
However, it is an open question whether these recent positive developments after the crisis stem from a major structural change in the macroeconomic dy-namics or simply reflect the expansionary phase of the boom–bust cycle. On the one hand, with the IMF’s supervision, the government and the central bank took important steps to achieve fiscal discipline and low levels of inflation, which are the two necessary preconditions for a sustained long-run growth performance. While the government followed a tight fiscal policy to put the debt stock on a sustainable path, the central bank started to implement a form of inflation targeting, where it is clearly stated that the main objective of the monetary policy is price stability.
On the other hand, it can be argued that the still-high real interest rates and the increasing appetite of the investors to invest in the emerging market econ-omies have induced massive short-term capital inflows, thus helping the Turkish economy to perform at high economic growth rates without any pressure on price dynamics. As the recent experiences in the emerging markets suggest, high growth rates that are mainly driven by short-term capital flows are not sustain-able and generally lead to excessive fluctuations in the macroeconomic dynamics in the long run. Moreover, with short-term capital inflows, pressure on exchange rates is relieved, which clearly affects inflation performance positively both through the lower cost of imported intermediate inputs and the prices of the imported final goods. However, as soon as the concerns regarding the financing of the current account deficit increase, a possible adjustment in the exchange rates has the potential to disrupt the price dynamics and generate lower eco-nomic growth. As a result, it is not clear whether the recent positive develop-ments in the Turkish economy imply a sustainable long-run growth performance or simply signal a temporary improvement.
To address the above-mentioned issue, this study analyzes the output dy-namics of the Turkish economy. First, beginning with the 1960s, we present an overview of the economy from a growth perspective. Then, we provide a brief comparison of the Turkish economy with the other MENA countries. The main finding in these two sections is that the Turkish economy has been far from having a sustainable growth path especially during the last two decades.
Therefore, following the comparison of the Turkish economy with the other MENA countries, we discuss the negative outcomes of the macroeconomic pol-icies implemented over the last two decades, with special reference to the finan-cial crises in 1994 and 2001. One main result of this discussion is that the short-term capital flows are becoming increasingly important in shaping macroeco-nomic dynamics. Thus, we continue our analysis with an empirical investigation of the influence of short-term capital flows on the excessive fluctuations in the output. In this respect, we utilize an output gap measure and show its comove-ment with the capital inflows. Using this gap measure, we also display the busi-ness cycle characteristics of the Turkish economy. Such an exercise will not only enable us to develop an idea about the recent macroeconomic framework, but it will also reveal which macroeconomic variables are procyclical or countercy-clical as well as whether these variables lead or lag behind the fluctuations in the output. Section 6 concludes the discussion.
2. An overview of the Turkish economy from a growth perspective
This section provides a brief historical overview of the Turkish economy from a growth perspective by identifying periods by stabilization policies and the
growth path followed.1In this section, we will mainly focus on capital formation
and the sources of growth. The evaluation of the Turkish economy is analyzed by dividing the entire period into three phases: 1963–1980, 1980–1999 and from the millennium onward.
2.1. Inward-oriented strategy: 1963– 1980 period
While there was no conscious planning or coordination of economic policies in the 1950s, a new constitution was formulated under the revolutionary govern-ment in 1960–1961. One of its key changes was social and economic develop-ment planning. The role of the state gained importance by following the import substitution policies together with economy-wide planning. During the First and Second Five-Year Plans, the traditional strategy of modernization and indus-trialization in a mixed economy was attempted systematically. In both plans, the main aim was to achieve a higher rate of growth and investment in a balanced economy. Some social issues addressed in the plans were more equitable income distribution, better regional balance, land reform, economic organization and the efficiency of the State Economic Enterprises (SEEs) (for more detail, see
Celasun and Rodrik (1989), S-enses (1991) and Metin (1995, pp. 17–18)).
The pace of economic growth accelerated during the 1960s and most of the 1970s was also good. In fact, the GDP increased at an average annual rate of
1See, for example, Aricanli and Rodrik (1990), Celasun and Rodrik (1989), Ekinci (1990, 2000), Metin-Ozcan et al., (2001) and O¨zatay (1999) among others for more detailed analyses and discus-sions of these and related issues of the Turkish economy.
6.4% during the First Plan period (1963–1967), 6.7% during the Second Plan period (1968–1972) and 7.2% during the Third Plan period (1973–1977). In the 1960s, development occurred with moderate reliance on foreign aid. In the 1970s, it was supported by large increases in workers’ remittances, but after 1974 by excessive reliance on foreign borrowing.
The long-term growth of industry has been particularly rapid, averaging 9.2% between 1963 and 1977. This led to a substantial structural change in the economy. Agricultural output growth averaged 3.3% per annum during this period, somewhat above the annual population growth rate of 2.5%. Population growth remained fairly high (2.5% per annum). Since GDP growth was fairly high, this allowed a relatively fast growth in per capita income. However, the rising income level does not appear to have been accompanied by a better in-come distribution. In fact, the 1973 household inin-come survey showed that the income was unequally distributed. The top 10% of the population received at least 40% of the income while the bottom 40% received about 11.5%. Con-siderable inequality within the agricultural sector (owing largely to extremely unequal land distribution) was the main cause of the overall income inequality (World Bank Country Study (1980, p. 11))
During the 1963–1977 period, both real private investment and real public investment increased at almost the same pace, averaging 10.8% and 11.4% per annum, respectively. This revealed the crowding-in effect in both sectors, which is in line with inward-oriented strategy (see I˙smihan (2003, p. 144). From 1973 to 1979, public sector investment increased tremendously with a 22.1% annual rise, while private investment grew by 10.2%.
During the 1960s, the macroeconomic environment was stable compared to that in the 1970s. In fact, the inflation rate, which is regarded as an important indicator of macroeconomic instability, was quite low during the 1960s (5.2% per annum) compared to that in the 1970s (27% per annum). Starting in the mid-1970s, the economy become even more unsteady since Turkey had delayed its internal adjustment to the external shocks of this period via reserve decu-mulation initially and excessive short-term borrowing later on (Celasun and Rodrik, 1989). In addition to macroeconomic instability, from the mid-1977 onward, the political environment became more unstable owing to successive weak coalition governments. In 1978–1979, the crisis in the economy deepened on account of inflation and balance of payment difficulties, so the government needed to implement more effective policies. On January 24, 1980, the govern-ment announced a major stabilization program, which included a series of new economic measures that were intended to solve the high inflation, economic stagnation and unmanagable balance of payment problems.
2.2. The 1980s and the 1990s
Turkey began taking serious steps to liberalize and strengthen its economy a full 25 years ago. In 1980, Turkey made a fundamental decision to switch its
economic strategy from an inward-oriented growth to an outward-oriented one. Prior to the reform program, tariff barriers were high, state ownership prevailed in key sectors and competition was strangled by regulations. The 1980 program had stabilization as well as structural aspects with efforts to liberalize trade and finance. During the outward-oriented period of 1980–1999, the real GNP grew at an average annual rate of 4.2% (see Table 1 which provides summary
in-formation on the Turkish economy for the 1980-1999 period).2
The economic growth rate was higher during the 1980s (5.2% per year) compared to that in the 1990s (3.2% per year). Capital formation performance in the private sector was better compared to that of the public sector during this period. Real private (public) investment grew at an average annual rate of 6.1% (1.6%) from 1980 to 1999. The most important development in this period was the changing role of the state in the investment process. The share of core public infrastructural investment in total public investment rose from 37.3% in the inward-oriented period to 50.5% in the outward-oriented one. While the private investment to GNP ratio (in current prices) rose from 12.8% in the 1980s to 18.1% in the 1990s, the public investment/GNP ratio dropped from 8.8% in the
1980s to 6.2% in the 1990s owing to budgetary pressures.3
Turkey experienced a very severe financial crisis in early 1994 mainly owing to unsustainable fiscal balance and monetization of the domestic debt. GNP con-tracted by 6.1% during 1994, which is the peak rate of contraction of the Turkish economy during the 1963–1999 period. Similarly, real public investment fell
Table 1. Selected indicators of the Turkish economy (1980–1999)
Output and capital formation Annual average growth rate
Real GNP 4.2
Real private fixed investment 6.1
Real public fixed investment 1.6
Real public fixed core infrastructural investment 2.7 Real public fixed non-core infrastructural investment 0.8 Composition of public investmenta
Core infrastructural investment (as % of total) 50.5 Non-core infrastructural investment (as % of total) 49.5 Macroeconomic instabilitya
Inflation rateb(%) 61.8
Source: I˙smihan et al. (2005), Table 1, p. 241. aSimple period average.
bPercentage change in the GNP deflator.
2For further discussion, see I˙smihan et al. (2005, p. 241).
3See Conway (1990), Uygur (1995) and I˙smihan et al. (2005, p. 242) for a detailed discussion of private and public investment behavior in Turkey.
dramatically by about 40% from 1993 to 1994. This is evidence of the negative effect of macroeconomic instability on the fiscal ‘‘ability’’ of governments to make investments. Real private investment, in contrast, contracted only moderately (about 5%). Inflation peaked in 1994 with 107.3% per annum and the Turkish lira depreciated by>70% against the US$ in 1994 (I˙smihan et al., 2005, p. 242). Following this major crisis in mid-1994, Turkey adopted an IMF-based stand-by agreement to eliminate the severe financial crisis. The inflation rate decreased from 107.3% in 1994 to 87.2% in 1995. However, macroeconomic instability continued until the late 1990s, and public sector balances were un-sustainable owing to heavy reliance on domestic borrowing. In addition, the East Asian and Russian crises of 1997–1998 and the two devastating earth-quakes of 1999 had a negative impact on the Turkish economy. In December
1999, Turkey signed a 3-year IMF-based stand-by agreement.4The objectives of
the program were tight fiscal and monetary policies, ambitious structural re-forms and the use of a predetermined exchange rate path as a nominal anchor. The disinflation program had a major impact on banks’ balance sheets. At the end of 2000, the Turkish banking system was seriously affected by increasing interest, exchange rate and credit risks because of the short-term maturity structure of financial sources used in the banking sector and the open positions of the Turkish banks. Owing to a major banking crisis, unfortunately, this program failed in the early 2001 and the real GNP contracted by 9.4% in 2001. Then, backed by the IMF and the World Bank, Turkey signed another program, called ‘‘Transition to a Strong Economy,’’ to eliminate the confidence crisis and financial instability. The strategy was strongly based on market orientation and openness to the world economy. An important pillar of the program consisted of a renewed effort to eliminate structural weaknesses that had not been fully tackled by the 2000 program, particularly by strengthening governance and improving economic management.
2.3. The millennium onward
The trend in economic growth during the period from 1998, the year prior to the economic crisis that affected the early 2000s, to the end of 2004 may be analyzed in two subperiods: 1999–2000 and 2001–2004. In the first subperiod, the average annual increase in the GDP was around 1.1% and in the second subperiod 3.7%. If the 1999–2004 period is taken as a whole, the rates of growth of the GNP and GDP are 2.9% and 2.3%, respectively. In spite of the improvement throughout the 2002–2004 period, the negative effects of the two banking crises experienced in 1999 and 2001 were still being felt (see Figure 1 to observe the GDP growth rates). A high GDP growth rate of 7% was achieved in 2002, followed by 4.8% in 2003 and 9% in 2004. The Turkish economy displayed an especially high growth
performance in 2004. After the 2001 banking crisis, inflation had been on a steady downward trend and came close to the single-digit level as of 2004. Starting from the first quarter of 2002, high growth figures were achieved without interruption over a period of 10 quarters. Therefore, there are indications that the Turkish economy has been free of the structure, which is characterized by boom–bust cycles since the 1990s and has embarked on a high-growth process (Pre-Accession Economic Programme, State Planning Organization (SPO) (2004, p. 27)).
At this point, it is necessary to discuss whether these positive developments caused expectations to change permanently. The course of Turkish economy has clearly seen improvement in expectations since 2001 (see Figure 2 below).
The upper-left panel of the figure indicates that business confidence has been stable and dominated by optimists since the beginning of 2002. Moreover, the consecutive achievements in reaching the year-end inflation targets since 2002 have increased the credibility of macroeconomic policies and have helped in-flation expectations maintain a downward trend since then. In line with the robust economic growth and disinflation process, improved economic funda-mentals have led to a more conducive environment for employment and invest-ment decisions (right panel of the figure). However, although business tendency survey results point to a significant improvement in the expectations, it is too early at this stage to claim that the changes in these expectations are permanent. Private fixed capital investment emerged as the other element that drove growth in 2004 (see Independent Social Scientists’ Alliance (ISSA) Report, 2005 and Pre-Accession Economic Programme, State Planning Organization (SPO) (2004) for further discussion).
When the total savings–investment balance, which is given in Table 2 is considered, the first point to be noted is that some revival in private sector fixed capital investment has been observed (13.7%) in 2004. However, this figure is still far behind the averages of the private fixed investment to GDP ratios of
GDP Growth Rate -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1990–1994 (17.9%) and 1995–1999 (17.4%) periods. It is also observed that this revival is largely financed by increases in external resources (4.6% in 2004) and that the domestic savings/GDP ratio, which was showing an increasing trend since the 2000s remained at the same level as during 1990–1994.
The second point to be raised is that the ratio of private investment to GDP was realized at 11.3% in 2002 and maintained at this level in 2003 along with a significant increase recorded in private investment in 2004 (13.7%). The third point is that the ratio of public investments to GDP, which was around 5%–6% during the 1990s (see Table 2), decreased to as low as to 4.0% in 2004 as a result of the tight fiscal policy implemented. In addition, stocks are rising and invest-ments, especially public investments are steadily declining. In fact, there is no possible economic explanation of the observation that stock increases have be-come a considerable determining factor in economic growth. However, Inde-pendent Social Scientists’ Alliance (ISSA) Report (2005, p. 11) put forth two reasons, which shed light on the phenomenon. First, it seems that the fact that the economy has recently become more dependent on imported inputs and consequent decreases in value added/gross production value ratios in all sectors including industry in the first place is not duly considered by the SIS, which still uses the technical coefficients of the 1990s. Hence, when high estimations are
-70 -50 -30 -10 10 30 50
Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05
Business Confidence (Optimists-Pessimists) % % % % -50 -40 -30 -20 -10 0 10 20
Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05
Total Employment Tendency (More-Less)
0 10 20 30 40 50 60 70
Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05
PPI Inflation Expectations for the Next Twelve Months
-80 -70 -60 -50 -40 -30 -20 -10 0 10
Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05
Investment Tendency for the Next Twelve Months (More-Less)
made for GDP growth (based on gross production values), final demands, which cannot be explained by changes in investment, consumption and net exports turn into virtual demand elements and reflect themselves in stock changes as a
residual (see also Somc-ag˘ (2004)). A less likely second possibility is the
under-estimation of private consumption expenditures, which are not as closely fol-lowed as investments by the State Planning Organization (SPO), and carrying the impact of this faulty estimation to stock changes (see Independent Social Scientists’ Alliance (ISSA) Report (2005, p. 11)).
Finally, the total domestic savings had an upward trend after 2000. The large increases of domestic demand that were observed in 2003 and 2004 have in-creased the economy’s need for foreign savings.
The most unpleasant result of the high growth rate was its failure to generate employment. Throughout 2004, unemployment remained a problem of the economy awaiting solution (see State Institute of Statistics (SIS), 2004). An important observation relating to labor market data for the period following 2002 is the presence of sharp increases in the national income together with overly slow expansion in employment, which is known as jobless growth in the
literature (see UNCTAD, 2002, 2003).5 Therefore, while the contribution of
employment to growth was negative owing to the decline in employment, the contribution of capital stock and total factor productivity (TFP) to growth increased considerably, reaching 52.2% and 58.5%, respectively, in 2003 and 2004 (Pre-Accession Economic Programme, State Planning Organization (SPO) (2004, p. 31).
Table 2. Total savings–investment balance (as percentage of GDP)
Year 1990–1994 1995–1999 2001 2002 2003 2004
Total investment 23.0 23.2 15.9 21.3 22.8 25.9
Fixed capital investment 24.2 22.9 17.2 16.6 15.5 17.7
Public investment 6.3 5.5 5.3 5.3 4.2 4.0 Private investment 17.9 17.4 11.9 11.3 11.3 13.7 Stock changes –1.2 0.3 –1.3 4.7 7.3 8.2 Total savings 23.1 23.3 15.9 21.3 22.8 25.9 Domestic savings 21.3 19.2 18.2 19.8 19.5 21.3 Foreign savings 1.8 4.1 –2.3 1.5 3.3 4.6
Note: The savings–investment balance is based on State Institute of Statistics (SIS) figures of the GDP by expenditures. Public investments in the table include the investments of the SEEs in addition to those of the general government.Source: Pre-Accession Economic Programme (2004, p. 31, Table 2.1).
5Turkey’s jobless growth starting from the late 1990s and the early 2000s is believed partly to have been due to tightening labor market regulations so as to be more like the EU. The labor market will be one of the key factors in ensuring a sustainable growth environment and enhancing the com-petitiveness of the economy during the process of EU accession (see further discussion, Pre-Accession Economic Programme, State Planning Organization (SPO) (2004, p. 80).
After these characteristics of the Turkish economy are presented, it will be insightful to compare the overall performance of the Turkish economy with the MENA countries. The next section provides such a comparison.
3. Comparison with the MENA countries
In this section, first, the economic growth and inflation performance of the Turkish economy will be compared to those of some MENA countries. Second, using the results of the previous studies that employed a growth accounting method decomposing aggregate GDP growth into labor, capital and technical progress for the MENA countries, these decomposed variables are compared with the ones obtained from the Turkish economy.
Interesting results are obtained when the inflation and output performance of the Turkish economy are compared with those of the non-oil MENA countries (see Table 3). Excluding Turkey, the countries in the non-oil MENA region have performed remarkably well in terms of inflation. As Table 3 shows, the average annual inflation rate for the region is 3.4% when the Turkish economy is not taken into consideration. The fact that the Turkish economy had a 56% average inflation rate for the sample period clearly shows why it should be treated as an outlier in its inflation performance.
When the standard deviation of the inflation, which can be regarded as a source of volatility, is calculated, it is seen that the inflation rate is also volatile. Turkish output growth was also very volatile compared to the other non-oil MENA countries during this period of>25 years.
However, despite the good inflation performance over the last two decades, the economic growth of the MENA region has been very slow, as Table 4 indicates. Although the region needed to grow to at least 5%–6% to reduce high-level unemployment, the average growth rate was only 2% for 1980–1990 and 3% for 1990–1998. Moreover, during the 1981–90, 1991–98 and 1980–2000
Table 3. Inflation and output performances for the period 1980–2003: Turkey
and non-oil MENA countries QA :1
Country Inflation rate (%) Standard Deviation Output (%) Standard Deviation
Turkey 56.9 22.7 4.08 4.78 Non-oil MENA 3.40 0.029 4.56 0.025 Egypt 4.40 0.028 6.2 0.024 Jordan 3.40 0.023 a Morocco 1.97 0.020 3.3 0.024 Tunisia 2.50 0.016 4.1 0.015 Syria 4.6 0.054 5.8 0.039
Note: Non-oil MENA countries comprises Egypt, Jordan, Morocco, Syria and Tunisia. Turkey is not included in the calculations.Source: Our calculations are based on the IFS CDROM data. aWe could not find the data in the IFS CDROM.
periods, the real GDP per capita growth in the region was 0.4%, 0.2% and -0.07%, respectively (see Sekkat (2003) Table 1, p. 5). The Turkish economy, in contrast, grew to an average of 5.4% during the 1980–1990 period. However, mostly owing to the financial crisis in 1994 and the Russian crisis in 1998, the economy grew at a slower pace between 1990 and 1998. Finally, the major economic crisis in 2001 led to even less economic growth rates between 2001 and 2004.
These findings clearly show that the Turkish economy suffered high and volatile inflation. The high economic growth rates were mostly achieved by stimulating the aggregate demand, which did not put the economy on a sus-tainable growth track and resulted in an inflationary environment.
In Table 5, we investigate the growth rates in the factors of production across subperiods for the non-oil MENA countries and Turkey. It is evident from Table 5 that the Turkish economy suffered a sharper decrease in growth rate during 2001–2003 and exhibited more volatility compared to the non-oil MENA countries. Another important finding is that the growth rate of labor turned out to be negative after the major economic crisis in February 2001. Unemployment rose and the contribution of labor to growth was negative.
In Table 6, by utilizing a growth accounting framework6we further elaborate
the differences in growth dynamics across the MENA7 countries and Turkey.
Table 4. GDP and population growth in Turkey and the MENA countries
Country Periods GDP Growth Population Growth
Turkey 1980–1990a 4.5a 2.6a 1990–1998a 5.0a 1.7a 1980–2004a 4.0a 2.4a MENAb 1980–1990a 2.0c 3.2c 1990–1998a 3.0c 2.3c 1980–2003c 4.5a 1.1a
Source: World Bank Development Indicators, CDROM, 1999, World Development Report, 1983, 1989, 1999/2000.
a
Own calculations.
bIt includes non-oil MENA countries, which are Egypt, Jordan, Morocco, Tunisia and Syria. cSekkat (2003), Table 1.
6The growth accounting method measures the contributions of the growth in capital per-worker and TFP to the growth of the real output per-worker. The measurement of TFP is the simplest econo-metric method and it is commonly used to evaluate past and future performance of growth in an economy. During the calculation of TFP growth, a two-factor growth accounting framework is usually used with labor and quantity of capital.
7For the growth accounting exercise for the MENA countries, see Makdisi et al. (2000), Senhadji (2000), Sekkat (2003), Bosworth and Collins (2003) and Hakura (2004).
Despite the considerable literature that has employed such a framework, there
are only a limited number of studies, which concentrate on the Turkish case.8
Although Turkey has been characterized by rapid growth in human capital, to the best of our knowledge there is no TFP calculation, which considers human
capital accumulation.9
Recently, the SPO, which is essentially based on the work of Saygili et al. (2005), calculated growth accounts for Turkey (1990–2003) in the Pre-Accession Economic Programme, State Planning Organization (SPO) (2004). However, the role of human capital accumulation is neglected in this analysis. Recently, in their growth accounting exercise, Ismihan and Metin-O¨zcan, (2005, p. 16) cal-culated a human capital-augmented labor series by using the average years of schooling of adult population (aged 15 years and over) and the rate of return to schooling, which is approximately 10%.
Table 5. Growth rates of factors of production: Turkey and the non-Oil MENA
countries
Growth Rate
Country Period Output Capital Labor
Turkey 1960–1980a 5.18 7.60 1.42 1981–1990a 5.35 4.63 2.59 1991–1997a 3.96 5.48 3.14 1990–2000b 4.10 7.60 1.50 2001–2003b 1.80 4.30 –0.7 2003b 5.80 4.00 –1.0 Egypta 1960–1980 6.77 7.95 2.20 1981–1990 4.93 8.75 2.52 1991–1997 4.68 2.28 2.84 Jordana 1960–1980 5.76 10.95 3.22 1981–1990 3.19 6.99 5.04 1991–1997 4.18 8.20 6.00 Moroccoa 1960–1980 5.87 8.25 2.71 1981–1990 3.57 4.95 2.46 1991–1997 2.36 3.59 2.61 Tunisiaa 1960–1980 6.61 6.48 2.84 1981–1990 3.72 4.60 2.78 1991–1997 4.30 3.92 2.96
aSekkat (2003) and FEMISE Report on the Euro Mediterranean Partnership (2003 p. 20, Box 3). bPre-Accession Economic Programme State Planning Organization (SPO) (2004), p. 31, Table 2.2.
8See Maras
-liog˘lu and Tiktik (1991), Temel and Saygili (1995), Saygili (1999), Temel et al. (2000), Bosworth and Collins (2003), Saygili et al. (2005), Ismihan and Metin-O¨zcan, (2005, p. 12–15) on growth, capital accumulation and the employment level of the Turkish economy.
9TFP growth is calculated as a ‘‘residual’’ from growth accounting formula and TFP growth can be thought of as a measure of gains (or losses) in economic efficiency.
Table 6 presents the contributions of the various factors to economic growth
for Turkey and different groups of MENA countries.10 In Turkey, both TFP
and capital accumulation were important sources of growth since the 1960s. However, starting from the 1970s macroeconomic instability has increased and become an important problem of the Turkish economy during the late 1980s and the 1990s. During this period political environment was unstable owing to sev-eral coalition governments. As a result, TFP growth exhibited significant var-iability over time. In this decade, a rise in TFP growth is remarkable. We believe that the following factors have contributed to the speedup in TFP: (i) successful lowering of inflation, (ii) structural reforms and stable political environment, (iii) accession to EU membership (see Ismihan and Metin-O¨zcan, 2005, p. 25).
Table 6. Contributions to economic growth by factors of production QA :2
Contribution to Growth
Country Period Capital (%) Labor (%) TFP (%)
Turkey 1960–1980a 4.11c 0.65c 0.42c 1981–1990a 2.50c 1.19c 1.65c 1991–1997a 2.96c 1.44c –0.45c Early 2000sb 0.8b 0.3b 2.3b Egypta,c 1960–1980 4.29 1.01 1.46 1981–1990 4.73 1.16 –0.95 1991–1997 1.23 1.31 2.32 Jordana,c 1960–1980 5.91 1.48 –1.63 1981–1990 3.78 2.32 –2.90 1991–1997 4.43 2.76 –3.00 Moroccoa,c 1960–1980 4.45 1.25 0.17 1981–1990 2.67 1.13 –0.23 1991–1997 1.94 1.20 –0.78 Tunisiaa,c 1960–1980 3.50 1.31 1.81 1981–1990 2.48 1.28 –0.04 1991–1997 2.12 1.36 0.82 MENA countries 1980–2000f 0.99 –0.06 –1.00
Non-oil MENA countriesd 1980–2000f 1.12 0.98 –0.15 Oil MENA countriese 1980–2000f 0.85 –1.09 –1.86 aekkat (2003) and FEMISE Report on the EURO Mediterranean Partnership (2003, p. 20, Box 3). bIsmihan and Metin-O¨zcan (2005, Table 4, p. 21 alpha ¼ 0.50).
cp ¼0.54 d
Non-oil MENA countries comprise Egypt, Jordan, Morocco, Syria and Tunisia. e
Oil- MENA countries comprise Algeria, Islamic Republic of Iran, and Kuwait. f
Hakura (2004, p. 24, Table 4).
10To our knowledge, a comparison between Turkey and the MENA countries has not been made before. Sekkat (2003) compares the growth performance of Morocco and the MENA countries.
The basic conclusion to be drawn regarding the Turkish economy is that, from 1990 onward, while the contributions of capital and labor to the economic growth rate are below the average of the MENA as well as the non-oil MENA countries, the Turkish economy performs remarkably well in terms of TFP. The important contributions of both capital and labor to the economic growth rates between 1960 and 1990 were replaced by TFP in the following years.
Consequently, our message from this section is clear: despite recent im-provements, when the MENA countries are taken as a benchmark, the Turkish economy suffered a high and volatile inflation for the whole sample. In addition, in terms of growth performance, it should be noted that the contribution of TFP to the output growth rate is much higher for the Turkish economy, despite the fact that the contributions of labor and capital are below the average. Finally, in terms of output, it should also be added that the growth performance of Turkey changes significantly over time, which leads to excessive fluctuations in output around its potential level. Such a finding necessitates further focusing on these fluctuations. For this purpose, in the following sections, we first elaborate on the effects of the macroeconomic policies implemented in generating these negative outcomes. Then, we try to build a link between short-term capital flows and the output gap.
4. Macroeconomic policies and the financial crises
Several studies point out that macroeconomic problems intensified during the 1990s. Thus, it is not surprising to witness two financial crises for this particular
period, which resulted mainly from incorrect macroeconomic policy decisions.11
In this section, we investigate these policy mistakes. In addition, we focus on two important negative characteristics of the economy for this particular period: specifically, the lack of fiscal discipline and the untimely liberalization of the financial markets. As will be clarified below, these two interrelated factors are significant, but not solely responsible for the two financial crises, which could have been avoided with the right macroeconomic policy decisions. We start our discussion with the currency crisis of 1994.
4.1. The 1994 currency crisis
As mentioned in the second section, in April 1994, the Turkish economy suffered a major currency crisis, during which the domestic currency depreciated by almost 70% against the US dollar, interest jumped overnight to 700%, and as a result, economic growth decreased by 6%. As discussed in O¨zatay (2000), de-spite the worsening macroeconomic fundamentals in the economy such as the
11For a detailed analysis of the Turkish economy for the sample period, see Ertug˘rul and Selc-uk (2001), O¨zatay (1999) and S-enses (1991).
fiscal discipline, there was no major sign of a macroeconomic-wide crisis. The risk premium in the economy was quite steady and there was no increase in inflation volatility. However, as the Treasury attempted to change the financing mechanism of the deficit, canceled the debt auctions and relied heavily upon
Central Bank resources, a currency crisis became inevitable.12 It is also
inter-esting to note that the macroeconomic fundamentals in the precrisis and post-crisis periods remained almost the same, implying that such a major post-crisis could actually have been avoided if these important policy mistakes in the debt fi-nancing mechanism had not been made. As a result, it can be claimed that the currency crisis in 1994 should not be attributed to the weak macroeconomic performance of the economy. It stems rather from the policy mistakes of the Treasury in financing the debt. As discussed further in O¨zatay (2000), there is a policy lesson that can be deduced from the 1994 crisis. In the absence of an effective stabilization program, the government should have done its best to continue debt finance. Given that the Turkish economy ran very high public deficits during the 1990s, any failure of the Treasury to finance the debt would naturally result in a currency crisis.
4.2. The 2001 financial crisis
The financial crisis in February 2001 was the deepest crisis that has ever been experienced in the Turkish economy. To better understand the causes of this crisis, one has to analyze the exchange rate-based stabilization program that was launched in December 1999. The main goal of the program was to bring in-flation down and satisfy fiscal discipline. Based on these two goals, the gov-ernment used the exchange rate as a credible anchor for inflationary expectations. However, similar to the other exchange rate-based stabilization programs that were implemented in other emerging market economies, there was a real appreciation of the domestic currency, and an increasing reliance on capital inflows to finance the current account deficit. At this point, for the program to be successful and to reduce the sensitivity of the economy to capital flows, certain conditions should have been satisfied. First, the government should have taken effective measures to ensure fiscal discipline. More impor-tantly, the banking sector, which was relied on heavily to finance the domestic debt, should not have had structural problems. As mentioned in Akyu¨z and Boratav (2001), since the banking sector was largely dependent on high-yielding T-bills, a rapid disinflation process, which lowered the real rates in the economy, made the banking sector vulnerable to the implemented stabilization program. Economic agents soon recognized that the program was nowhere near satisfying
12There was a strict belief by the government that the debt dynamics could be changed easily and the high real interest rates could be reduced with this attempt. However, the financial markets reacted negatively to this policy change, which resulted in a financial crisis.
the goals, which resulted in massive capital outflows. As a result, the program had to be abandoned and the domestic currency depreciated significantly. The impact of the financial crisis on the real side of the economy was also detri-mental: the economy shrank at the unprecedented rate of 9.4% in 2001 and the unemployment rate increased from 6.50% in the last quarter of 2000 to almost 12% at the beginning of 2002.
When the two economic crises are compared, it can be seen that the 1994 crisis stemmed mainly from the Treasury’s mismanagement of the public debt. For the 2001 crisis, in contrast, the vulnerability of the banking sector and the characteristics of the stabilization program caused the crisis. However, one common feature of these two crises is that the macroeconomic fundamentals were weak and there was no solid macroeconomic structure. Among these macroeconomic fundamentals, fiscal dynamics and the liberalization of the fi-nancial markets emerge as two important characteristics of the Turkish econ-omy. Next, we analyze these two features.
4.3. Fiscal dynamics
Contrary to other emerging market economies, Turkey experienced relatively modest accumulated fiscal debts up to the mid-1990s. Although the government ran budget deficits, these were largely monetized and interest rates remained at steady levels, excluding the crisis period in 1994. As a result, it can easily be claimed that the main source of the high and persistent inflation in the Turkish economy was the lack of fiscal discipline. Since the budget deficits were mainly monetized, the size of the accumulated fiscal debt could remain at moderate
levels. However, as mentioned in Pamukc-u and Yeldan (2005), beginning in the
mid-1990s, there were two additional characteristics, which made the imple-mentation of the fiscal policy much more difficult. First, the monetization of the debt was no longer an option for the Treasury, which gave way to financing the domestic debt through very high real interest rates. Such a strategy not only increased the accumulated fiscal debt but also the risk premium required by
investors to demand government bonds.13 Second, as a result of these
devel-opments, the average maturity of the domestic debt became very short, which made debt financing more difficult for the Treasury. In addition, the banking sector bailouts right after the February 2001 financial crisis further intensified the debt management problem. As a result, the government started a new IMF-led program, which relied heavily on fiscal restraint. To achieve these goals, the fiscal authority aimed to generate a primary budget surplus of 6.5% for the public sector in proportion to the GNP. Although a budgetary discipline seems to be satisfied in recent years, real interest rates are still very high and the debt to
13One reason for the high-risk premium was the probability that the government would renege on the domestic debt.
GDP ratio has not decreased significantly. Pamukc-u and Yeldan (2005) consider these results totally expected and argue that the debt burden in the economy cannot be handled with the present IMF-led program. They add that, to achieve long-term fiscal stability, a detailed restructuring of the terms of the Turkish debt obligations is necessary.
4.4. Liberalization of the financial markets
In the last two decades, an increasing number of developing countries have decided to liberalize their financial markets. The main motive behind these lib-eralization attempts was to restore growth and stability by raising savings. The Turkish economy was no exception in this context, and there were intense lib-eralization attempts, starting at the beginning of the 1980s. However, the most important date for these attempts was the capital account liberalization in Au-gust 1989. When the period following the capital account liberalization is ob-served, it can easily be seen that the foreign direct investments, which promote long-run growth and do not cause any speculative activity, occurred at an almost negligible level. Short-term capital flows were very important and often led to macroeconomic uncertainty through increased financial instability. The econ-omy also observed a divergence of domestic savings away from fixed capital investments toward speculative financial markets. As a result, with a weak fi-nancial structure and a shallow fifi-nancial market, massive short-term capital flows caused volatility in output growth and uncertainty in exchange rates.
It could be argued that financial liberalization could change the structure of the overall economy in a positive manner, by promoting investment in the real side of the economy via extending credits from the financial markets. At the same time, the promising sectors could develop more easily owing to increasing capital flows to these sectors. However, as Boratav and Yeldan (2002) show, the liberalization of the financial markets was also unable to change the structural sectoral composition, the nature of market concentration and the behavior of profit margins in the post-1980 period. In this context, the liberalization at-tempts were effective only on the financial markets and could not restore long-run economic stability.
The liberalization of the financial markets could be considered as untimely. In an economy with a weak financial structure, shallow exchange rate markets and very high real interest rates owing to the lack of fiscal discipline, it was almost inevitable that short-term capital flows would have negative effects on macroeconomic dynamics. The changing risk appetite of foreign investors could easily lead to instability in the economic performance. To make this evident, the next section gives a clearer perspective concerning the relationship between ex-cessive fluctuations in output and short-term capital flows.
5. Output gap and business cycle characteristics
The findings in the previous sections show that there have been excessive fluc-tuations in the output dynamics for the Turkish economy during the last two decades. In this context, it can be stated that there existed a boom–bust cycle, where a sustainable growth path could not be achieved. This characteristic of the Turkish economy differentiates it from its industrialized counterparts, necessi-tating a thorough analysis of these excessive fluctuations. Therefore, in addition to our discussion of the sources of growth, we investigate this issue in brief. For this purpose, we utilize an output gap measure, where the gap is taken as the difference between the actual output and the potential output. While a positive gap implies that the actual output was above its potential level, thus exerting pressure on the price dynamics, a negative gap implies the opposite.
However, measuring the output gap is controversial, especially for emerging market economies. There are two main reasons for this. First, conventional approaches, such as the HP filter and the quadratic time filter, are purely sta-tistical methods, which do not convey any economic information. Second, as stated by Aguiar and Gopinath (2004), emerging markets are prone to frequent shocks that affect the potential output. Therefore, these economies typically do
not have very smooth trends.14
Taking these factors into account, we adopted the output gap measure that was derived exclusively for the Turkish economy in Ece et al. (2005). As can be seen in the appendix, this gap measure is obtained from a semi-structural time-series model, which represents the general characteristics of the Turkish econ-omy. Thus, it is not subject to the criticisms that apply to conventional methods and it can be conveniently used in our analysis.
Figure 3 displays the actual output, the estimated potential output and the output gap. At first sight, it can be seen that the crisis and recovery periods are captured well.
The output gap measure displays the fact that, several times in the sample period, the Turkish economy was subject to severe shocks that created an un-stable environment, ending with rapid contractions in the economic activity. As a consequence, output gap estimates exhibit sharp movements. According to the output gap figure, expansion periods were generally interrupted by economic crises so that the last decade seems to have witnessed three separate periods of recession. Such an observation validates both the excessive boom–bust cycle peculiar to emerging markets and the ‘‘unsustainable growth path’’ that the Turkish economy has followed in the last decade.
5.1. Role of short-term capital flows
Emphasis can be placed on the significance of short-term capital flows as a major determinant of the above-mentioned output gap, especially after the capital account liberalization of 1989. Capital flows that are accompanied by weak financial structure have led to a number of economic crises with severe social and economic costs in many emerging market economies. A similar argument can be valid for Turkey considering the growth performance in the aftermath of the financial liberalization policy. The experiences of the last decade especially point to the dependence of growth performance on capital flows (Figure 4).
Actual and Potential Output
9.9 10.0 10.1 10.2 10.3 10.4 10.5 90Q1 91Q1 92Q1 93Q1 94Q1 95Q1 96Q1 97Q1 98Q1 99Q1 00Q1 01Q1 02Q1 03Q1 04Q1 Actual Output Potential Output Output Gap -10 -8 -6 -4 -2 0 2 4 6 8 90Q1 91Q1 92Q1 93Q1 94Q1 95Q1 96Q1 97Q1 98Q1 99Q1 00Q1 01Q1 02Q1 03Q1 04Q1 % deviation
In fact, economic conditions providing speculative arbitrage opportunities were a result of unsuccessful disinflationary efforts, which dragged the economy
into ‘‘slowed down depreciation-correction’’ cycles (Ertug˘rul and Selc-uk, 2001).
In this respect, the fluctuations during the 1990s can be characterized as a fi-nancial cycle determined by capital movements. Therefore, periods of capital inflows coincided with high economic growth periods followed by subsequent real depreciation and crisis periods. In an effort to control inflation as well as ease domestic borrowing, the policies focused on slowing down the depreciation of the Turkish lira.
However, policies oriented toward controlling the exchange rate generally lead to an initial decline in interest rates and a real appreciation of the domestic currency owing to the slow catch-up of the inflation rate. As a stylized fact, similar to the general characteristic of exchange rate-based stabilization pro-grams, an initial upsurge in domestic demand and real activity along with real appreciation give rise to an increase in imports. The subsequent deterioration of the current account balance and concerns regarding its sustainability reverse the process and trigger capital outflows. Hence, the economy comes to its initial point again with high interest rates, depreciated domestic currency, contracted domestic demand and an improved current account balance. Such a macroeco-nomic instability results in the rising volatility of the growth rate and thus creates mini business cycles, the duration of which shortens as a result of the huge and rapid changes in the size and direction of short-term capital flows.
Figure 5 demonstrates the tight relationship between net capital inflows and the real exchange rate, except for the years 1998 and 1999. During the 1995–1998
-15 -10 -5 0 5 10 15 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 -12 -8 -4 0 4 8 12
Net Capital In flow, billions dollar, left axis GDP Growth, %, right axis
period, capital flows had been quite stable at positive levels and the real ex-change was appreciating. It was possible to sustain the appreciation of the Turkish lira in spite of two consecutive external shocks, the Asian and Russian economic crises; the latter resulted in a huge capital outflow in the third quarter of 1998. Obviously, such an appreciation leads to a temporary increase in output through two distinct channels. First, a surge in consumption is observed, which increases the aggregate demand. Second, an increase in the imported interme-diate and capital goods is observed, which further increases output in the econ-omy.
In terms of monetary policy, it should be noted that during the 1995–1999 period, the main goal for the central bank was to maintain financial stability through a real exchange rate targeting strategy. Thus, other than pursuing a clear price stability objective, the central bank mainly tried to stabilize the
-8 -6 -4 -2 0 2 4 6 8
Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 80 90 100 110 120 130 140 150 160
Net Capital Inflow, bln.$, left axis Real Exchange Rate, right axis
-10 -8 -6 -4 -2 0 2 4 6 8
Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Net Capital Inflow, bln.$ Output Gap, %
negative effects of the volatile capital flows on the real exchange rate for the sample period.
As a result, it can be claimed that short-term capital flows stand out as a very important factor in generating the excessive fluctuations in output performance. The comovement of the output gap and the capital inflows in Figure 5 provide further support for this claim. In addition, the close relationship between these two series does not seem to have broken down in recent years. Such an ob-servation makes us think that the high economic growth rates and the perform-ance of the Turkish economy above its potential level in recent years owe much to the increasing short-term capital inflows. At this stage, it is still early to state whether the Turkish economy has entered a sustainable growth path. The recent developments more closely resemble an expansionary phase of a boom–bust cycle.
As a final exercise, it can also be claimed that short-term capital flows are attracted by the state of the economy, i.e. the output gap. When the economy performs at high economic growth rates, investors might have incentives to invest in Turkish financial securities, which, in turn, generate short-term capital inflows. To investigate this issue, we first checked the cross correlations of the output gap and the short-term capital flows. Then, we performed a Granger-causality test between these two variables. The results can be seen below.
For the 1992Q1–2004Q4 period, the cross correlations between the output gap (GAP) and short-term capital flows (STCF) show that the highest
corre-lation (0.57) is observed between GAPtand STCFt-1. Therefore, STCF may be a
leading indicator of the output gap (Table 7).
However, since correlation itself cannot be interpreted as a causality between the two variables, we also perform a Granger causality test by estimating a VAR
QA :3 system. According to the test results, past values of STCF can be used to
fore-cast GAP at the 5% significance level, while the opposite is not true. Therefore, we find a one-way causality between the two variables and short-term capital flows are the Granger-cause of the output gap (Table 8).
This finding is an empirical one, which also has an underlying economic reason. Short-term capital inflows lead to a real appreciation of the domestic
currency, which may affect the output gap from different channels.15First of all,
a real appreciation of the Turkish lira may trigger the demand for imported
Table 7. Cross correlations of GAPtwith STCFt7i
(t–5) (t–4) (t–3) (t–2) (t–1) (t) (t+1) (t+2) (t+3) (t+4) (t+5) 0.05 0.05 0.19 0.36 0.58 0.54 0.23 –0.11 –0.30 –0.37 –0.18
15For a more detailed discussion of the effect of the real exchange rate on the output gap, see Ece et al. (2005).
goods by bringing their relative prices down, thereby increasing the output gap. It may also lead to a narrowing of the output gap through an increase in the potential output by inducing firms to import capital goods from abroad. In contrast, in a country like Turkey, which is heavily dependent on intermediate goods’ imports to expand production, a real appreciation may also increase the supply in the economy. Regardless of which effect dominates, short-term capital flows surely have a significant impact on consumption, investment and produc-tion decisions, and thus on the output gap, mainly through the exchange rate channel.
5.2. Business cycle facts for Turkey
After stressing the role of the short-term capital flows in generating the output gap, this subsection illustrates the business cycle facts for Turkey by examining how the cyclical behavior of the selected macroeconomic variables are related to the output gap. Such an exercise reveals which factors lead or lag behind the fluctuations in the output as well as showing whether these variables are pos-itively (procyclical) or negatively (countercyclical) associated with the output gap. Following Kydland and Prescott (1990), we first characterize the business cycle facts for Turkey by using the HP-filter to detrend selected real factors and then we compare their cyclical behavior with our output gap estimate. We em-ployed the HP-filter in decomposing the real factors for two reasons. First, we wanted the business cycle analysis to be consistent with the exercise that Kyd-land and Prescott (1990) provided. Second, it would require a significant amount of additional work to come up with a different structural model for each factor and then employ the Kalman filter to decompose these factors into their trend and cyclical components. Therefore, we decided to adopt the main variable of interest (the output gap) from a structural framework while we used conven-tional techniques for the other variables. The main implications of the analysis did not change even when we derived the output gap from the HP-filter.
The cyclical behavior of the series is evaluated according to the degree of comovement with the output gap and depending on whether there is a phase shift in the movement of a variable relative to the cycle. A positive (negative) contemporaneous correlation coefficient is interpreted as the variable being procyclical (countercyclical). If the largest correlation coefficient appears in
Table 8. Granger causality test between GAP and STCF
Dependent variable: GAP
Excluded w2 P-value
STCF 5.7456 0.0165
Dependent variable: STCF
Excluded w2 P-value
column (t-i), where i>0, then the series tends to move i quarters before the real GDP and leads the cycle. Similarly, the largest correlation coefficient appearing in column (t+j), where j>0, indicates that the series lags behind the gap by j quarters.
The cyclical facts are summarized in Table 9. At first glance, one obvious fact is that most of the selected variables are procyclical and coincide with the busi-ness cycle in Turkey. Moreover, major output components, private consumption and private investment tend to move together with the cycle.
Our output gap estimate is closely tied to private consumption making it procyclical, as their contemporaneous correlation is 0.82. Durable consumption also comoves with the cycle with a high correlation. Private investment, one of the main determinants of potential output, exhibits a strong contemporaneous relationship with the output gap, so private investment is said to be procyclical. Moreover, among the subitems of private investment expenditures, machinery-equipment investment is closely related to the cycle whereas building investment is found to be acyclical. Selected indicators of resource utilization, such as ca-pacity utilization, number of workers and working hours, move with the output gap simultaneously. Finally, the relationship between labor costs and output gap is found to be a lagged one rather than a contemporaneous one. Both real wages and real unit wages move with a one-quarter lag relative to the cycle. Therefore, the cyclical behaviors of real wages and real unit wages turn out to be consistent with the standard adjustment process in the sense that a positive output gap puts upward pressure on real marginal costs through increased demand for labor.
Finally, it can be claimed that the output gap measure that is employed throughout the analysis can be criticized for being purely statistical and lacks the theoretical background. For this purpose, to check the robustness of the results, we used another output gap measure, which is derived from Dees et al. (2005). This measure also takes into account the evolution of trend output of Turkey’s
Table 9. Cross correlation of output gap (Gapt) with selected variables
(1988Q2–2004Q4) QA :4 (t–5) (t–4) (t–3) (t–2) (t–1) (t) (t+1) (t+2) (t+3) (t+4) (t+5) Private consumption –0.13 –0.09 0.09 0.34 0.64 0.82 0.60 0.27 0.02 –0.11 –0.17 Durable consumption –0.18 –0.08 0.12 0.37 0.62 0.74 0.52 0.24 0.05 –0.03 –0.07 Private investment –0.12 –0.08 0.09 0.28 0.53 0.66 0.56 0.32 0.07 –0.09 –0.18 Machinery equipment –0.14 –0.09 0.09 0.31 0.60 0.72 0.60 0.33 0.05 –0.13 –0.22 Construction –0.01 –0.02 –0.01 0.00 0.06 0.11 0.16 0.16 0.12 0.17 0.17 Capacity utilization –0.02 0.03 0.17 0.36 0.53 0.59 0.25 –0.09 –0.29 –0.37 –0.27 Number of workers –0.01 0.00 0.08 0.23 0.37 0.49 0.44 0.22 –0.05 –0.23 –0.32 Working hours 0.01 –0.01 0.09 0.26 0.44 0.54 0.41 0.17 –0.12 –0.30 –0.33 Labor productivity –0.07 –0.06 0.12 0.29 0.48 0.53 0.20 –0.02 –0.12 –0.12 0.02 Real wagea –0.02 0.09 0.22 0.37 0.52 0.61 0.62 0.42 0.20 –0.04 –0.16 Real unit wagea 0.15 0.22 0.25 0.27 0.38 0.51 0.62 0.52 0.32 0.03 –0.17 a
trading partners. As Tables B.1, B.2 and B.3 in Appendix B shows, while we find slightly smaller correlations between output gap and the other relevant variables, the main implications of the business cycle analysis do not change. In addition, we find that the short-term capital flows again Granger-cause the output gap, which supports the previous empirical results.
6. Conclusion
Unsustainable output growth and high inflation are the two prevailing macro-economic characteristics of the Turkish economy. These two factors also indi-cate a boom–bust cycle, which the economy has experienced more severely in the last decade. In this respect, as of mid-2005, after the negative effects of the 2001 crisis were partly eliminated, there has been considerable discussion about whether the recent positive developments in the output and the inflation dy-namics are the result of a new macroeconomic policy framework or whether they simply reflect an expansionary phase of such a business cycle.
Taking this discussion as our reference point, we tried to analyze the growth dynamics in the Turkish economy. Throughout the study, we handled the issue in three ways. First, we presented an overview of the Turkish economy from a growth perspective and compared its performance in terms of output and in-flation with the other MENA countries. The main finding from this approach is that Turkey has not followed the right macroeconomic policies to have a sus-tainable growth path. Moreover, the Turkish economy has suffered from high and volatile inflation. In addition, the growth accounting exercise shows that the TFP is the main cause behind the recent increase in output growth rates for the Turkish economy. Then, we discuss the characteristics of these incorrect mac-roeconomic policies, with a special reference to the two financial crises in 1994 and 2001. This discussion brings to the fore an important link between short-term capital inflows and excessive output fluctuations. Finally, we performed an empirical investigation of the role of the short-term capital flows in generating these fluctuations. It can be seen clearly that there is a very close relationship between the output dynamics and the capital inflows. The periods of high growth rates also coincide with the periods of massive short-term capital inflows. The fact that these capital inflows reached a maximum by mid-2005 makes us think that the recent positive developments owe much to the increases in the capital inflows. Thus, we can say that it is too early to judge the success of the new macroeconomic policy framework that was put into effect after the 2001 crisis. The effects of this new framework will be better understood when the short-term capital inflows decrease to moderate levels.
Lastly, based on the output gap measure, we analyzed the business cycle characteristics of the Turkish economy. As a result, we found that most of the major output components tend to move with the cycle. Moreover, the indicators of resource utilization move with the output gap simultaneously. Finally, a
positive output gap exerts an upward pressure on the marginal costs through the labor demand channel.
Uncited References
State Planning Organization (SPO), 2001. References
Aguiar, M. and G. Gopinath (2004), Emerging market business cycles: the cycle is the trend, Mimeo, University of Chicago.
Akyu¨z, Y. and K. Boratav (2001), ‘‘The making of the Turkish financial crisis’’, prepared for a conference on Financialization of the Global Economy, PERI, December 7–9, University of Massachusetts, Amherst, MA.
Aricanli, T. and D. Rodrik (1990), The Political Economy of Turkey: Debt, Adjustment, and Sustainability, London: Macmillan.
Boone, L., M. Juillard, D. Laxton and P. N’Diaye (2001), ‘‘How well do al-ternative time-varying parameter models of the NAIRU help policy makers forecast unemployment and inflation in the OECD countries?’’, IMF Work-ing Paper, Draft version.
Boratav, K. and E. Yeldan (2002), ‘‘Turkey, 1980–2000: financial liberalization, macroeconomic (in)-stability and patterns of distribution’’, Mimeo, January. Bosworth, B. and S.M. Collins (2003), The empirics of growth: an update,
Brookings Institution.
Butler, L. (1996), ‘‘A semi-structural method to estimate potential output: com-bining economic theory with a time-series filter’’, The Bank of Canada’s New Quarterly Projection Model (QPM), Part 4.
Celasun, M. and D. Rodrik (1989), ‘‘Debt, adjustment and growth: Turkey’’, in: J. Sachs and S.M. Collins, editors, Developing Country Debt And Economic Performance, Country Studies, Chicago: University of Chicago Press. Conway, P. (1990), ‘‘The record on private investment in Turkey’’, in: T.
Aricanli and D. Rodrik, editors, The Political Economy of Turkey: Debt, Adjustment, and Sustainability, London: Macmillan.
Dees, S., F. Di Mauro, M.H. Pesaran and L.V. Smith (2005), ‘‘Exploring the international linkages of the Euro area: a global VAR analysis’’, CESifo Working Paper No. 1425.
Ece, D., H. Kara, F. O¨g˘u¨nc-, U¨. O¨zlale and C-. Sarikaya (2005), ‘‘Estimating
output gap for the Turkish economy’’, Unpublished Manuscript, The Central Bank of the Republic of Turkey, Research Department, Ankara, Turkey. Ekinci, N. (1990), ‘‘Macroeconomic developments in Turkey: 1980–1988’’,
METU Studies in Development, Vol. 17(1–2), pp. 73–114.
Ekinci, N. (2000), ‘‘The IMF and reforming the public sector in Turkey’’, METU E.R.C. Working Papers in Economics, No. 00/04, Ankara, METU.
Ertug˘rul, A. and F. Selc-uk (2001), ‘‘A brief account of the Turkish economy 1980–2000’’, Russian and East European Finance and Trade, Vol. 37(6), pp. 6–28.
Femise Report on the EURO Mediterranean Partnership (2003), ERF, Cairo,
Egypt and Institut de la Mediterranee, France. QA :6
Hakura D. S. (2004), ‘‘Growth in the Middle East and North Africa’’, IMF Working Paper WP/04/56.
Independent Social Scientists’ Alliance (ISSA) Report (2005), ‘‘On economic
and social life in Turkey in early 2005’’,
http://www.bag-imsizsosyalbilimciler.org.
I˙smihan, M. (2003), ‘‘The role of politics and stability on public spending dy-namics and macroeconomic performance: theory and evidence from Tur-key’’, Unpublished PhD Thesis, The Graduate School of Social Sciences of Middle East technical University, Ankara.
Ismihan, M. and K. Metin-O¨zcan (2005), ‘‘Sources of growth in Turkish econ-omy, 1960–2004’’, ERF 12th Annual Conference Paper, Cairo, Egypt. I˙smihan, M., K. Metin-O¨zcan and A. Tansel (2005), ‘‘The role of
macroeco-nomic instability in public and private capital accumulation and growth: the case of Turkey 1963–1999’’, Applied Economics, Vol. 37, pp. 239–251. Kydland, F. and E. Prescott (1990), ‘‘Business cycles: real facts and a monetary
myth’’, Federal Reserve Bank of Minneapolis Quarterly Review, Spring, pp. 3–18.
Makdisi S., Fattah Z. and I. Limam (2000), ‘‘Determinants of growth in the MENA countries’’, The World Bank sponsored project on the Global De-velopment Network.
Maras-liog˘lu, H. and Tiktik, A. (1991), Tu¨rkiye Ekonomisinde Sekto¨rel Gelis-meler
: U¨retim, Sermaye Birikimi veI˙stihdam 1968– 1988, Devlet Planlama Tes-kilati,
I˙ktisadi Planlama Bas-kanlig˘i, DPT: 2271- I˙PB: 428.
Metin, K. (1995), The analysis of inflation: the case of Turkey (1948– 1988), QA :7
Ankara: Nurol Publishing Co., Capital Markets Board Publication No. 20. Metin-Ozcan, K., E. Voyvoda and E. Yeldan (2001), ‘‘Dynamics of macroeco-nomic adjustment in a globalized developing economy: growth, accumula-tion and distribuaccumula-tion, Turkey 1969–1998’’, Canadian Journal of Development Studies, Vol. 22(1), pp. 219–253.
O¨zatay, F. (1999), ‘‘Populist policies and the role of economic institutions in the performance of the Turkish economy’’, Yapi Kredi Economic Review, Vol. 10(1), pp. 13–26.
O¨zatay, F. (2000), ‘‘The 1994 currency crisis in Turkey’’, Journal of Policy Re-form, Vol. 3, pp. 327–352.
Pamukc-u T. and Yeldan, E. (2005), ‘‘Country profile: Turkey, public sector and
fiscal policy issues’’, Mimeo, March 2005.
Saygili, S-. (1999), Technical change efficiency, growth and exports: the case of
Turkish economy, Unpublished PhD thesis, University of Kent at Canter-bury, England.
Saygili, S-., C. Cihan ve H. Yurtog˘g˘lu (2005), ‘‘Tu¨rkiye Ekonomisinde Sermaye Birikimi Verimlilik ve Bu¨yu¨me: 1972–2003, DPT yayin No. 2686, Ankara. Sekkat, K. (2003), ‘‘The sources of growth in Morocco: an empirical analysis in
a regional perspective’’ ERF Working Paper, Cairo.
Senhadji, A. (2000), ‘‘Sources of economic growth: an extensive growth ac-counting exercise’’, IMF Staff Papers, Vol. 47(1), pp. 129–157.
S-enses, F. (1991), ‘‘Turkey’s stabilization and structural adjustment program
retrospect and prospect’’, The Developing Economies, Vol. 29, pp. 210–234.
Somc-ag˘, S. (2004), ‘‘Confession of imaginary stock by the SIS’’,
http://www.se-limsomcag.org.
State Institute of Statistics (SIS) (2004), Household Labor Surveys, Ankara, Turkey.
State Planning Organization (SPO) (2004), Pre-Accession Economic Programme, Ankara: Republic of Turkey: SPO, November 2004.
State Planning Organization (SPO) (2001), Main Economic Indicators, Ankara: Republic of Turkey: May, SPO, 2001.
Temel, A., ve S-. Saygili (1995), ‘‘An estimation of gross fixed capital formation’’,
in: T. Bulutay, editor, Investment and the Labour Market in Turkey:
Pro-ceedings of a Seminar Held in Ankara 7 December 1995, Ankara; SIS. QA :8
Temel, A., E. Boyar and S-. Saygili (2000), Tu¨rkiye Ekonomisinde Bu¨yu¨me ve
Ekonomik Yapi Deg˘is-meleri: 1946-1999, Unpublished Manuscript, State
Planning Organization, Ankara, Turkey.
UNCTAD (2003), Trade and Development Report, 2002 and 2003, New York: UNCTAD.
Uygur, E. (1995), ‘‘Recent estimates of investment and private investment be-haviour in Turkey’’, in T. Bulutay, editor, Investment and The Labour Market in Turkey: Proceedings of a Seminar Held in Ankara 7 December 1995, An-kara; SIS.
World Bank Country Study (1980), Turkey: policies and prospects for growth, Washington, DC, USA: The World Bank.
Appendix A: The output gap estimation for the Turkish economy
This appendix briefly introduces the output gap model and the methodology that is used in Ece et al. (2005). The system of equations, in which the output gap is derived as the difference between the actual output and the potential output, is as follows:
(1) Inflation-output gap dynamics:
pt¼a1;tpt1þa2;tpt2þa3;tgapt1þa4;treertþnt
yt¼y
t þgapt
(3) Potential output equation:
yt ¼yt1þmt1þZt
(4) Potential output growth rate equation:
mt¼ ð1 rtÞm0þrtmt1þt
(5) Output gap dynamics:
gapt¼g1;tgapt1þg2;trtþg3;tDItþg4;treertþBt
where pt is the inflation rate defined as the logarithmic difference of quarterly
seasonally adjusted consumer price index (CPI), gaptthe unobserved output gap,
reert the logarithmic difference of the real effective exchange rate, yt the
log-arithmic seasonally adjusted real gross domestic product, y*t the unobserved
potential output, mtthe potential output growth rate, rtthe ex post real interest
rate based on 3-month Treasury auction rates and DItthe demand index, which
is constructed from the Business Tendency Survey of the Central Bank of the
Republic of Turkey. In contrast, vt, Zt, etand xtrepresent shocks to the system,
which are assumed to be i.i.d. with zero mean and constant variances.
It is important to remind that the parameters of the system are time varying. Therefore, one has to make a time-series specification for the evolution of these parameters. It is assumed that each time-varying parameter follows a random walk. Such a specification can be defended on theoretical grounds: Since any structural change on the dynamics of the model–thus the system parame-ters–cannot be known a priori, it is intuitive to specify a random-walk process for each parameter. As a result, the system includes nine more equations, where each time-varying parameter follows a random-walk process.
Finally, it should be noted that the output gap variable, which needs to be estimated within the system, appears in multiplicative form with the time-var-ying parameters, which, again, are to be estimated. Thus, the model has a non-linear characteristic, where the standard Kalman filter is inappropriate and should be replaced by the extended Kalman filter algorithm, which is designed for estimating non-linear state space models. For details about the extended Kalman filter algorithm, one can either read the appendix in Ece et al. or ‘‘Theory and practice of recursive identification’’ by Ljung and Soderstrom
