What Explains the Rapid Growth in Transition Economies?
Author:
Mr. Garbis Iradian
Search for other papers by Mr. Garbis Iradian in
Current site
Google Scholar Close
Publication Date:
04 Nov 2009
eISBN:
9781589069107
Language:
English
Keywords:
SP; transition economy; market reform; inflation rate; TFP gain; estimation result; Terms of trade; Estimation techniques; Growth accounting; Global; East Asia; Western Europe
Download PDF (449.4 KB)

This paper analytically explores and empirically tests a number of hypotheses to explain the rapid growth in transition economies. Using the latest panel data, the paper finds that growth in transition economies has been higher because of the recovery of lost output, progress in market reforms, and favorable external conditions. These results are consistent with estimates from the global sample that includes 123 countries, and are robust to instrumental variable estimations and other robustness tests. A general implication of the findings is that some of the factors behind the rapid growth are unlikely to continue for a very long time and that the challenge would be to further improve the investment climate, which will require broadening the scope of macroeconomic reform into a second generation of reforms encompassing structural and institutional areas.

Abstract

This paper analytically explores and empirically tests a number of hypotheses to explain the rapid growth in transition economies. Using the latest panel data, the paper finds that growth in transition economies has been higher because of the recovery of lost output, progress in market reforms, and favorable external conditions. These results are consistent with estimates from the global sample that includes 123 countries, and are robust to instrumental variable estimations and other robustness tests. A general implication of the findings is that some of the factors behind the rapid growth are unlikely to continue for a very long time and that the challenge would be to further improve the investment climate, which will require broadening the scope of macroeconomic reform into a second generation of reforms encompassing structural and institutional areas.

Transition economies are in a resurgent phase. From a 17-year perspective—that is including the sharp fall in the early 1990s—the record is not better than the average for developing countries. In the past decade, however, growth in most of the transition economies compares very favorably with the fastest growing economies in East Asia. In particular, the average unweighted average annual growth of the Commonwealth of Independent States (CIS) was about 7 percent in 1996–2006. This strong performance is in welcome contrast to the first half of the 1990s, when cumulative output contracted from 30 to 60 percent in these economies. Therefore, determining whether the strong rebound from the posttransition setbacks is more the result of conditions, including progress in market reforms that will support continuing growth requires an examination of the underlying influences.

Transition countries have been typically excluded from cross-country studies of the long term because of the short historical span, and because earlier data collection methods were unreliable. However, significant improvements have been made in data quality in recent years. In the economic literature there are mainly two approaches with regard to sources of growth, namely the cross-section growth accounting approach and the panel regression approach. This paper uses both approaches to analyze the sources of the recent rapid growth in transition economies, with particular emphasis on the CIS. The research questions include the following:

  • Is either investment or total factor productivity (TFP) growth responsible for the major shifts in economic growth?1

  • Is the recent strong growth explained by a bounce back from the initial posttransition setbacks (recovery of lost output)?

  • Do market reforms explain the variance in relative output performance?

  • To what extent is the recovery of growth driven by favorable external conditions?

The growth-accounting exercise in this paper suggests that the recent strong growth has been driven largely by growth in TFP. On average, capital accumulation made a modest contribution and employment rates continued to drop through 2004 in most transition countries. Assuming that TFP growth slows down, other sources of growth will be essential to sustain a rapid catchup. In this connection, the recent trend of faster capital accumulation and the recent improvement in employment in some of the transition countries are expected to play a more important role in future growth.

The results of the panel regressions suggest the following: (1) transition countries that experienced larger declines in output during the early 1990s tended to grow at faster rates; (2) improvements in macroeconomic policies and market reforms explain about half of the total growth; and (3) the growth acceleration payoff to reforms in 2001–06 was enhanced by the favorable external environment (positive terms-of-trade shock, and global technological innovation). These external factors have accounted for about one-fourth of the average annual growth in transition economies.

I. Growth Accounting Approach

This section uses the growth accounting approach to determine the sources of growth in transition economies. Given the relatively short time span since the breakdown of the former Soviet Union, there are only few papers that have analyzed the determinants of growth in transition economies. Broeck and Koen (2000) analyzed the determinants of the sharp fall in output in transition economies in the 1990s. Loukoianova and Uigovskaya (2004) extended the work of Broek and Koen through data covering the period 1991–2000 for the low-income CIS. Schadler and others (2006) focused on the Baltic and Central European countries (Box 1).

Review of Literature on Growth, Reforms, and Institutions in Transition Economies

Although there is agreement that stabilization policies are important, no consensus has yet been reached on the role of reforms in the recent strong recovery of transition economies. Below are the main papers and their findings relevant to the current study.

  • Abed and Davoodi (2000) find that progress on structural reforms is both statistically more significant and economically more important than corruption in explaining differences in economic performance in transition economies.

  • Broeck and Koen (2000) and Loukoianova and Uigovskaya (2004), using the growth accounting approach, found that most of the decline in 1991–97 and the subsequent recovery in output are explained by the movements in total factor productivity growth. In the absence of factor prices they assumed shares of 0.3 for capital and 0.7 for labor.

  • Campos and Coricelli (2002) assess the implications of Broeck and Koen’s growth accounting estimates; emphasize the role of reforms and institutions in dictating the path of transition process; and note that isolating reallocation from accumulation and technological progress remains a major challenge in transition economies.

  • Fidrmuc (2003) cast doubts on the benefits of reform and Lawson and Wang (2004) failed to find a strong and positive effect of reforms on growth.

  • Falcetti, Lysenko, and Sanfey (2005) found a positive and strong link between progress in market-oriented reforms and economic growth.

  • Campos and Horváth (2006) constructed objective measures for three main reform areas (internal liberalization, external liberalization, and privatization) in transition economies for the period from 1989 to 2000.

  • Beck and Laeven (2006), using natural resource reliance and the years under socialism to extract the exogenous component of institution building, showed the importance of institutions in explaining the variation in economic development and growth across transition economies during the first decade of transition.

  • Schadler and others (2006) examined the progress toward income convergence achieved by the five Central European and the three Baltics countries and the policy challenges that these countries will face in facilitating the catch-up process. In the panel regression approach, the main variables used to explain growth were population growth, partner country growth, the relative price of investment goods, schooling, openness, government taxation, and institutional quality.

  • Babetskii and Campos (2007) reviewed 43 econometric studies and found that the existing subjective measures of reform, controlling for institutions, and initial conditions were the main factors in decreasing the probability of reporting a significant and positive effect of reform on growth.

Methodology and Data Issues

The data set suffers from various serious weaknesses due to underreporting by private enterprises to avoid taxes and regulations, particularly in the early years of transition. The decline in output during the first half of the 1990s could be overstated because the statistical system was designed to collect information only on publicly owned enterprises. Beyond the mid-1990s, the information on the emerging private sector gradually became available and was incorporated into the statistical system.

One major concern about the measurement of the capital stock is that a significant portion of the communist capital stock may have been permanently scrapped. If so, this would cause the contribution of capital accumulation to be underestimated during the subsequent recovery. In order to address this concern, a one-time adjustment for the permanent scrapping of a significant portion of the capital stock during the communist era is applied; that is, the capital stock is reduced by the same rate as output between 1990 and 1995 so that the capital-output ratio is not allowed to rise during the course of the sharp contraction in output. Also, the quality of capital is treated the same over time and across countries.

Obviously, workers in different countries have different levels of skills. Typically, education and number of hours worked are emphasized as key components of effective labor. Such data, however, are not available for most transition countries. Few growth-accounting studies on nontransition economies made adjustments to labor quality by including education, age, or gender (Boseworth and Collins, 2003). Such information is available only for selected years and is limited for industrial countries and some emerging and developing countries. More importantly, the education level in transition economies, as measured by secondary school attainment or average years of study, is relatively high as compared with other developing and emerging economies, and there is little variation among them. Thus, the correlation between the education level and growth is expected to be weak in this case. In the absence of adequate indicators that reflect changes in the quality of labor over time and across countries, the growth in TFP will therefore be overestimated.

Another major challenge in using the growth accounting approach is the proper estimate of the shares of capital and labor in output. There are several approaches in the growth literature to estimating the shares of capital and labor in output. The first approach assumes that factor markets are perfectly competitive so that earnings of the factors are proportional to their productivities. However, this approach cannot be used for some transition countries due to lack of detailed national accounts statistics. The second approach uses a priori measure of capital share in the range of 0.3 to 0.4. Aiyer and Dalgaard (2005) establish that the standard Cobb-Douglas methodology of assuming a constant capital share of one-third for all countries is a very good approximation to a more general formulation under which countries have different aggregate production functions that do not require a constant elasticity of substitution among factors. The third approach estimates the coefficients of the production function by regressing the growth rate of output on the growth rates in capital and labor. The intercept then measures the growth in TFP, and the coefficients on the factor growth rates measure the shares of capital and labor, respectively. The main advantage of this process is that it dispenses with the assumption that factor social marginal products coincide with the observable factor process. The disadvantage of the regression approach, however, is that the growth of capital and labor cannot usually be regarded as exogenous with respect to variations in TFP (Barro, 1999). The fourth approach, used by Hsieh (2002), is the dual exercise. The advantage of using the dual is that factor prices, primarily wages and interest rates, are observed as an equilibrium outcome in a marketplace. However, a number of assumptions and estimates have to be made in order to construct the data on quantities of output and capital needed for a primal growth accounting exercise (Hsieh, 2002, p. 519). Shigeru, Khan, and Murao (2003) proposed a fifth approach that does not need the assumption of perfectly competitive factor markets nor assumes any particular functional form of the aggregate production function. Their approach is based on nonparametric kernel derivative estimation techniques developed in the statistics and econometrics literature (Pagan and Ullah, 1999). This approach estimates much lower elasticity of output with respect to capital (around 0.20) for several East Asian countries, thus emphasizing even more the role of the residual (growth in TFP) in explaining growth.

Results for Transition Economies

Given the data limitations for transition economies, particularly the low income CIS economies, this section uses the third approach mentioned above. Table 1 shows the regional estimates of the shares of capital and labor. The estimated TFP growth by regions, which here is the residual, was the highest for the Baltics, followed by the CIS.2 Ideally, separate production function for each country should be used. But given the short historical span (1991–2006), it is assumed that the production functions are similar for each region. The countries included in each region share some common characteristics and are likely to have similar production functions. The estimated sum of the capital and labor elasticities are not far away from unity for the three Baltics and the five Central and Eastern European countries, but slightly higher than 1 for the CIS. The endogeneity problem is partially addressed by using two-stage least squares (2SLS), although finding good instruments remains a challenge. The estimates of the shares of capital and labor are found to be robust to different estimation techniques.

Table 1.

Estimates of Input Shares, 1996–2006

article image
Source: Author’s calculations.

Using the estimated shares of capital and labor from Table 1, the sources of growth are derived for the 26 transition countries (Figure 1). The average annual TFP growth in the CIS was higher than in Central European and six Southeast European economies, but lower than in the Baltics. A natural question is then, what were the factors that led to this high TFP growth? It is most likely that the inefficiencies inherited from central planning left much scope for managerial improvements, labor shedding, and gains from interindustry resource reallocation. Higher TFP growth could also be explained by the scale of some of the transition economies, which are relatively poor economies with very low endowment of technology. Hence for a given technological innovation, the smaller the initial endowment, the higher is the growth of TFP. When capital is scarce, its marginal productivity is considerable. Therefore, for similar investment rates, the contribution of capital deepening should be larger in economies with less capital. Increases in capacity utilization could also raise TFP growth. The sharp fall of economic growth in most transition economies before 1995 is largely attributed to steep decline in TFP. In the second phase (1996–2000) there was improvement in TFP growth in almost half of these economies.

Figure 1.
Figure 1.

Contributions to Average Real GDP Growth, 1996–2006

(In percentage points of GDP)

Citation: IMF Staff Papers 2009, 004; 10.5089/9781589069107.024.A005

Sources: Author’s own calculations based on the following assumptions: Initial capital stock to GDP ratio of 2 and annual depreciation rate of capital stock of 5 percent. The elasticity of output with respect to capital and labor are derived from the regional production functions in Table 1.Note: TFP = total factor productivity. First panel: ARM = Armenia; AZE = Azerbaijan; BEL = Belarus; GEO = Georgia; KYR = Kyrgyzstan; MDA = Moldova; MON = Mongolia; KAZ = Kazakhstan; RUS = Russia; TJK = Tajikistan; UKR = Ukraine; UZB = Uzbekistan. Second panel: EST = Estonia; LAT = Latvia; LTU = Lithuania; CZE = Czech Republic; HUN = Hungary; POL = Poland; SVK = Slovak Republic; SVN = Slovenia; ALB = Albania; BIH = Bosnia and Herzegovina; BGR = Bulgaria; HRV = Croatia; MAC = Macedonia; ROM = Romania.

The sensitivity of the TFP growth estimates were also examined under different assumptions of initial capital to output ratio (k) and elasticities of output with respect to capital. An increase in k from 2.0 to 2.5 raises the estimated TFP growth for the 15 countries of the former Soviet Union from 2.4 to 3.2 percentage points, and for the six Southeast European economies from 1.0 to 1.7 percentage points. A decrease in the capital share from 0.6 to 0.4 (close to the share of capital used in the literature) increases average annual TFP growth from 2.4 to 3.5 percentage points for the former Soviet Union countries (equivalent to 60 percent of output growth in 1996–2006). In general, countries with higher capital shares will tend to have lower TFP growth; a higher elasticity of output with respect to capital would result in a rise of the contribution of physical capital and a decline in the contribution of TFP growth. It should be noted that the methodology used here does not adjust factor inputs for quality changes. The implication is that the incremental effect on growth of embodied technological advancement is not attributed to capital but is rather measured as a higher level of TFP. The same measurement problem can also arise in the case of labor. Education and on-the-job training would improve the quality of labor. This would be reflected in higher TFP. This “mismeasurement” of TFP may well be significant in the case of transition economies, following the move from central planning to market economies.

Historical Perspective from Industrial and East Asian Economies

During the “Golden Age” (postwar period) in Western Europe and Japan, there were strong contributions to growth from TFP gains. Catching up, scale effects, and improvements in resource allocation made strong contributions to TFP during 1950–60 in the major industrial countries (Maddison, 1996).3 These improvements stemmed from adjusting to trade liberalization, exploiting opportunities for mass production as larger and better integrated markets emerged, and from moving resources out of relatively low-productivity agriculture. As catch-up growth weakened, the magnitude of TFP growth fell markedly after 1973. The estimated TFP growth shown in Figure 2, using a simple growth accounting approach, is close to other more sophisticated techniques in the growth literature. Consistent with the results found in this paper, Jorgenson and Yip (2001) found that between 1960 and 1973 TFP growth accounted for more than half of the growth in output for France, Germany, and Japan, but somewhat less than half of output in the United States. The relative importance of TFP growth declined substantially after 1973. Jorgenson (2005) shows TFP growth of 0.5 for the United States, −0.1 for Germany, and 0.85 for Japan for the period 1995–2001. Amador and Coimbra (2007), using a dynamic translog stochastic production frontier (computed through Bayesian Statistical methods), found that France, Germany, and the United Kingdom moved to a new lower floor of TFP contribution in the last two decades. In contrast, the United States and Canada recorded slightly higher TFP acceleration after the mid-1980s.

Figure 2.
Figure 2.

Total Factor Productivity Growth, Historical Perspective

(In percentage points)

Citation: IMF Staff Papers 2009, 004; 10.5089/9781589069107.024.A005

Sources: Author’s estimates with the exception for the period 1950–60, which is based on Christiansen, Cummings, and Jorgenson (1980).Note: The assumptions are that the initial capital stock to GDP ratio was 2; annual depreciation of 5 percent; and elasticities of output with respect to capital and labor of 0.3 and 0.7, respectively.

East Asian growth has relied much more heavily on factor inputs, both labor and capital, and less on TFP growth than that of “Golden Age” Europe and the current rapid growth in transition countries. The estimates in this section also show that TFP growth has accounted for about 40 percent of the output for South Korea and for slightly more than half for China in the past two decades. There are very few countries around the world that were able to sustain rapid growth for more than 15 years with relatively low shares of investment in GDP. In Chile, factor accumulation accounted for two-thirds of the growth in 1986–95, and about 90 percent in 1996–2006. Ireland’s impressive economic performance over the past two decades was also driven largely by factor inputs. On the other hand, India achieved its growth with relatively little emphasis on capital accumulation and more substantial gains in TFP. In that mix of gains, India differs from the East Asian economies.

A key question for prospective growth is whether the TFP gains achieved thus far have already eliminated most of the inefficiencies of central planning—and will therefore soon fade away. Sustaining productivity growth rates such as those experienced recently in some of the transition countries is difficult. Underutilized labor combined with the recent trend of faster capital accumulation is expected to play a more important role in the medium-term growth.

II. Panel Regression Approach

This section reports estimates of spare specifications of growth equation including new variables that are of particular importance to transition economies, such as recovery of lost output, and improvements in market reforms. The aim is to use up-to-date data and experiences of a large number of countries over long periods to identify key determinants of growth for transition economies and to form a view of growth prospects that complements the growth accounting exercise in Section I.

This section differs from previous empirical studies on the determinants of growth in the following aspects. First, the main focus is on transition economies using the latest information (1991–2006). Second, it analyzes a new set of explanatory variables, including output recovery index, and different measures of market reforms and institutional qualities that may have influenced variation in output and TFP over time and across countries. Third, it tests for endogeneity of some of the explanatory variables so that appropriate econometric methods can be chosen. Fourth, it assesses the importance of period-specific effects (in the form of world economic conditions) based on a large sample of countries that includes developing and developed countries.

Methodology and Data Issues

Although the main focus of the paper is the transition countries, a heterogeneous data set is also used including most advanced and developing economies. Such an approach would improve the statistical reliability of the results. The transition sample includes 12 CIS countries, three Baltic countries, five Central European countries, and six Southeastern European countries covering the period 1991–2006. The global sample consists of 123 countries with data spanning the period 1980–2006 (data for the transition economies have a shorter span). The use of a large panel of countries over an extended period of time allows sufficient freedom to enrich the menu of variables used on the determinants of growth, and improves the statistical reliability of the results. Pure cross-section, four-year nonoverlapping averages, and annual data are used to estimate the regressions.

Averaging the data over time eliminates short-term disturbances as well as business cycle effects from the data, while allowing one to test for long-run market reform dynamics. Failure to eliminate short-run dynamics typically leads to highly correlated time series and to gross overestimation of coefficients. The choice of four-year periods is dictated by the data time span for transition economies (1991–2006), which gives four observations for each country. For the global sample, each country is represented by seven observations. Unlike a pure cross-country regression using long-period average data, a panel regression (with each observation representing four- or five-year averages) provides additional information because it captures both time-series and cross-sectional information. The definition and sources of data are described in the data appendix.

To examine the relationship between market reforms, recovery of lost output, and growth, this section follows the standard empirical growth literature,4 and uses the following linear growth regression model:

g i ⁢ t = β ⁢ Z i ⁢ t + λ ⁢ X i ⁢ t + μ i + v t + ε i ⁢ t , ( 1 )

where git, the dependent variable, is the per capita real GDP growth rate or TFP growth rate in country i during the period t, Z is the vector of “core explanatory variables” that are believed to have contributed to the rapid growth in transition countries (including recovery of lost output index, market reforms, and institutional quality). X comprises a set of control variables that are often used in the growth literature, including terms of trade, level of development as proxied by initial GDP per capita, inflation rate, fiscal balance, government size in the economy (as captured by government consumption to GDP ratio), and investment. µi is a country-specific unobservable effect, vt is a time-specific factor, and εit is the disturbance term. The paper also controls for time-specific growth effects emanating from changes in the external economic environment by including world cycle dummies.

The panel regressions for annual and four-year average are estimated using three different methodologies: (1) pooled ordinary least squares (OLS) (fixed, random effects, or seemingly unrelated regressions); (2) 2SLS; and (3) the generalized method of moments (GMM).5 For the pure cross-section, the OLS could be considered as an efficient estimation technique.6

In equation (1) the core explanatory variables (vector Z, which includes measures of market reforms and quality of institutions) may not be entirely exogenous. If the causality runs mainly from these variables to growth then the problem may be benign, but if it runs from growth to these variables then the problem is more severe. The Durbin-Wu-Hausman test suggests that endogeneity is present, albeit not very strong. This problem is addressed, to a certain extent, by using the 2SLS and the GMM techniques.

We need good instruments for market reforms, quality of institutions, and investment. The following variables are used as instruments: (1) lagged values of the exogenous explanatory variables; (2) commodity exports as share of total exports; (3) distance from the capital of the respective country to Brussels; (4) an index measure of ethnic fractionalization as measured by Alesina and others (2003); and (5) a period trend. Using more appropriate instruments would yield more efficient instrumental variable (IV) estimates. But given the large number of explanatory variables and the relatively small cross-sectional dimension (particularly in the “transition” sample) by the standards of common panel data, overfitting should be avoided by working with a reduced number of IVs.

Weak instruments correspond to weak identification of some or all of the unknown parameters. Weak identification leads to GMM statistics with nonnormal distributions, even in large samples, so that conventional IV or GMM inferences are misleading. A rule of thumb for checking for weak instruments is the F-statistics, which is used to test the hypothesis that the coefficients on the instruments equal zero in the first stage of the 2SLS (see Stock, Wright, and Yogo, 2002). When there is a single endogenous regressor, a first-stage F-statistics of less than 10 indicates that the instruments are weak, in which case the 2SLS estimator is biased (even in large samples), and 2SLS t-statistics and confidence intervals are unreliable.

Determinants of Growth

The explanatory variables used in this paper can be divided into four groups: (1) recovery of lost output or catch-up process (in the case of the “transition” sample); (2) market reforms and institutions; (3) other measures of macroeconomic and investment; and (4) external conditions.

Recovery of Lost Output

For the sample that includes only transition economies, the real GDP index (1990 = 100) of the previous period is used to test whether the amplitude of output recovery is influenced by the magnitude of the fall in output before recovery. The experiences of many countries show that usually sharp contractions in output due to crisis, wars, or other major shocks to the economy may be followed by strong growth that offsets the initial decline. This combined with corrective policies and structural reforms to reduce inefficiencies could spur strong economic recovery above the original trend line.

There are two models in the literature on the recovery of lost output following recessions, financial crisis, or a change in economic regime (Cerra and Saxena, 2005b). The first model (based on Friedman, 1993) suggests that recession can be characterized as a temporary fall in output. After the negative large “temporary” shock dissipates, output returns back to trend in a rapid-growth recovery phase (Figure 3). The second model (based on Hamilton, 1989) suggests that the stochastic trend in output undergoes regime switching between positive and negative states, resulting in a permanent output loss.

Figure 3.
Figure 3.

Two Views of Lost Output Recovery

Citation: IMF Staff Papers 2009, 004; 10.5089/9781589069107.024.A005

Negative shocks in theory impose only a temporary restraint on output, but may lead to rapid future growth that offsets the initial decline. First, negative shocks could stimulate political and economic reforms. Corrective policies could prompt an economic recovery above the original trend line if they reduce inefficiencies. Second, following Schumpeter’s idea of “creative destruction,” a sharp fall in output may cleanse the economy of inefficient firms, leading to higher productivity and economic growth (Caballero and Hammour, 1994). The core process of change comprises two elements: reallocation of resources from old to new activities (via closures and bankruptcies, combined with the establishment of new enterprises), and restructuring within surviving firms (via labor rationalization, product line change, and new investment). These can be thought of as the dynamic movements resulting from the establishment of new incentives and are reminiscent of the Schumpeterian concept of “creative destruction” by entrepreneurial activity, only with a much larger impact than what Schumpeter’s model envisioned.

However, Cerra and Saxena (2005a) found that recessions or large contractions in output due to crisis, wars, or other reasons are in general not followed by high-growth recovery phases. They conclude that when output drops, it tends to remain well below its previous trend. The data used by Cerra and Saxena consisted of annual observations spanning 192 countries from 1960 to 2001, and thus their sample did not capture the recent strong growth in transition economies.

The recovery of lost output in the case of the transition sample is proxied by the real GDP index, with 1990 = 100. The index series are constructed using the real GDP growth rate estimates. For the global sample, the following indices are used: (1) 0 if initial real GDP was greater than 95 percent of its value in 1990; (2) 0.5 if initial real GDP was between 70 and 95 percent; and (3) 1 if initial real GDP was less than 70 percent.

Market Reforms and Institutional Development

Measures of market reforms for transition economies have been constructed by the World Bank (De Melo, Denizer, and Gelb, 1996), the European Bank for Reconstruction and Development (EBRD), and most recently by Campos and Horváth (2006). Critics have noted that proxies market reforms as measured by the EBRD are outcome-oriented rather than measuring inputs. In particular, there is a concern about the reliability of the EBRD scores, particularly during the early years of transition. In 2000, EBRD made an effort to backdate the indicators to 1990. This implies that the ratings for the early 1990s have to be treated cautiously, especially as these were the years in which information flows were limited. For robustness test, therefore, this paper also uses other measures of market reforms. In this regard, Table 2 presents the development of market reforms in transition economies as measured by the EBRD and Campos and Horváth.

Table 2.

Evolution of Selected Growth Determinants in Transition Economies

article image
article image
Sources: IMF, World Economic Outlook database; EBRD, Transition Reports; Campos and Horváth (2006); and International Country Risk Guide (ICGR).

Average of eight EBRD transition reform indicators (price liberalization, competition policy, banking reform, trade and foreign exchange system, large-scale privatization, small-scale privatization, governance and enterprise reforms, and infrastructure). The transition indicators range from 1 to 4.3, with 1 representing little or no change from a rigid centrally planned economy and 4.3 representing the standards of an industrialized market economy.

Based on Campos and Horváth (2006). Simple average of internal liberalization, external liberalization, and privatization indices.

The ICRG ratings reflect risk of the following components: government stability, socioeconomic conditions, investment profile, internal conflict, external conflict, corruption, law and order, ethnic tensions, democratic accountability, and bureaucratic quality.

The overall EBRD score in Table 2 is the unweighted average of eight structural reform indicators: price liberalization, small-scale privatization, large-scale privatization, competition policy, trade liberalization, financial sector reform, governance and enterprise reforms, and infrastructure reform. The EBRD indicators range from 1 to 4.3, where 4.3 indicates that the country’s structural characteristics are comparable to those prevailing on average in market economies, and 1 represents conditions before reform in a centrally planned economy with dominant state ownership of the means of production. However, it could be argued that the EBRD market reform scores are output-oriented rather than measuring inputs.

Campos and Horváth (2006) argue that the market reform indicators of the World Bank and EBRD are subjective and may be influenced by ex post reports of favorable or unfavorable performance raising the problem of possible endogeneity. But, in recent years, the EBRD Transition Reports have highlighted the main factors behind the change in their scores for each category of reform. They have constructed three categories of reform. The first captures internal liberalization reflecting the extent of price and wage liberalization. The second captures external liberalization efforts as reflected in the severity of trade barriers and capital controls. The third index captures the extent of privatization efforts. These three indices were constructed for 25 transition economies for all years between 1989 and 2000. Using Campos and Horváth’s methodology this paper extended the market reform indices through 2006.

Institutional development is particularly important to investigate for the transition economies, where the institutions of central planning were a key constraint on growth in the early 1990s. Easterly (2003), and Rodrik, Subramanian, and Trebbi (2004) show that institutions are more robustly associated with faster growth than policies. Major institutional changes have taken place since the breakdown of the Former Soviet Union in 1991. Assessments of public institutions are mainly based on two indices: the political risk rating, International Country Risk Guide (ICRG), and the World Bank’s governance indicators. The ICRG comprises 22 variables in three subcategories of risk: political, financial, and economic, for 143 countries and is available since 1984. The political risk rating, which is used in this paper, includes five indices of perceptions of government stability, democratic accountability, law and order, quality of bureaucracy, and corruption in government. The World Bank’s governance indicators comprise six measures of institutional development: voice and accountability, political instability, government effectiveness, regulatory burden, rule of law, and control of corruption. Unfortunately, the World Bank’s governance data do not exist prior to 1996, and therefore are not used in this paper. It should be noted that the market reform as measured by the EBRD and Campos and Horváth and the ICGR are highly correlated with each other (Figure 4).

Figure 4.
Figure 4.

Different Measures of Market Reforms and Institutional Quality are Correlated

(Four-year nonoverlapping averages, 1991–2006)

Citation: IMF Staff Papers 2009, 004; 10.5089/9781589069107.024.A005

Based on both the EBRD and Campos and Horváth scores, Table 2 suggests that progress has been achieved in market reforms in most transition economies. But the CIS still remain far behind the five Central European and the three Baltic countries. In general, reform is most advanced in the privatization of small-scale enterprises, the liberalization of foreign trade and exchange, and the elimination of price controls. Structural reforms are least advanced in the regulation and supervision of the banking and financial sector, the development and enforcement of competition, and the reform of governance in both the private and the public sectors. Among the CIS countries Armenia and Georgia so far achieved an average market reform index as measured by the EBRD score of more than three.7 Progress in market reforms has been particularly slow in Belarus and Uzbekistan. Turkmenistan virtually did not reform its economy with the exception of some small-scale privatization and price liberalization. An important question is whether the ICRG index picks some of the dimensions specifically related to transition. In this regard, Figure 4 compares the ICRG index with the EBRD or Campos and Horváth indices. It shows that institutional quality as measured by the ICRG is highly correlated with the measure of EBRD’s market reforms, which is also highly correlated with market reforms measures based on Campos and Horváth.

Macroeconomic Stabilization, Investment, and Convergence

There is an array of policy determinants of growth (see Barro and Sala-i-Martin, 2004). In this paper the impact of macroeconomic stabilization is measured by the logarithm of the inflation rate,8 and the overall fiscal balance as a ratio of GDP. Inflation is a policy result, but the fiscal balance refers more to the policy itself. It should be noted that the improvement in the overall fiscal position in Azerbaijan, Kazakhstan, Russia, and Uzbekistan was largely due to the substantial increase in government revenues from oil, gas, and other major commodities. Fiscal policy may influence growth through the size of government in the economy, as measured by the ratio of government consumption to GDP. Higher government consumption is believed to reduce growth prospects. This effect is normally associated with the crowding out of private sector investment, higher rent-seeking behavior, and distorted market incentives including higher taxation.

There is little disagreement in the general growth literature that investment is a major engine of growth. In the transition economies, with a history of excessive capital accumulation and inefficient use, the role of investment in the initial recovery phase (perhaps through the late 1990s) was relatively less important. In recent years, however, there has been some increase in the investment ratio, albeit from a very low level. Most of the increase in investment has been in the hydrocarbon and metallurgy sectors.

In the case of the full sample, the paper also considers convergence as one of the determinants of growth. Most neoclassical growth models have shown that the potential for economic growth rate also depends on a country’s level of development as proxied by the initial per capita income (Barro and Sala-i-Martin, 2004). The coefficient for this variable is expected to be negative, implying that poor countries tend to grow faster than richer countries as each country converges toward its steady state.

External Conditions

Economic activity in a country is also affected by external conditions. The literature provides ample evidence of the transmission via international trade and external financial flows (see Mendoza, 1997). The change in the terms-of-trade index is included to account for possible exogenous shocks in international commodity prices that may have an impact on per capita growth. This index is derived from export prices relative to import prices. Terms-of-trade shocks capture changes in both the international demand for a country’s exports and the cost of production and consumption inputs (Barro and Sala-i-Martin, 2004). This variable may also be included in the list of IVs because its movement depends primarily on world conditions and therefore is largely exogenous with respect to per capita growth for an individual country.

Several empirical studies have found a positive and significant link between improvement in the terms of trade and economic growth (Fisher, 1993; Mendoza, 1997). Barro (1997) notes that if the quantities of domestically produced goods do not change, then an improvement in the terms of trade raises real gross domestic income, but does not affect real GDP. Movements in real GDP occur only if shifts in the terms of trade bring about a change in domestic employment and output.

Estimation Results

The correlation matrix (Table 3) of the explanatory variables indicates high correlation (of more than 0.50) in the following explanatory variables: the market reform index, institutional quality, inflation rate, and per capita income. This implies that if these variables are included in the same regression, the estimated coefficients may not be individually reliable due to high multicollinearity.

Table 3.

Correlation Coefficient Matrix, Transition Sample

article image
article image
Source: Author’s calculations based on 26 transition economies covering the period 1996–2006. Note: TFP = total factor productivity.

Simple average of eight EBRD transition reform indicators (price liberalization, competition policy, banking reform, trade and foreign exchange system, large-scale privatization, small-scale privatization, governance and enterprise reforms, and infrastructure). The transition indicators range from 1 to 4.3, with 1 representing little or no change from a rigid centrally planned economy and 4.3 representing the standards of an industrialized market economy.

Based on Campos and Horváth (2006). Simple average of international liberalization index, external liberalization index, and privatization index.

Average of six institutional concepts: voice and accountability, political stability, government effectiveness, regulatory burden, rule of law, and control of corruption. Each of these indicators is distributed normally, with a mean of zero and a standard deviation of one. The scores lie between −2.5 and 2.5, with higher scores corresponding to better outcome.

ICRG = International Country Risk Guide. The ICGR ratings reflect risk ratings of the following components: government stability, socioeconomic conditions, investment profile, internal conflict, external conflict, corruption, law and order, ethnic tensions, democratic accountability, and bureaucratic quality.

First-Stage Estimation of Endogenous Variables

Table 4 shows the first-stage estimation results. Based on Stock and Yogo (2005) criteria, the IVs used to estimate the exogenous component of market reforms and institutions are strong (exceed the threshold values by a large margin).9 The estimated coefficients for raw material exports, distances from Brussels, civil conflict or war dummy, and period trend have the expected sign and are highly significant in the regressions where measures of market reform are used as the dependent variables, columns (a) and (b). Together the exogenous variables explain about 70 percent of the variation in the scores for market reforms. In column (c), the exogenous variables explain about 50 percent of the variation in institutional development as measured by the ICGR scores. Also, specification tests confirm the validity of the IVs. The test of the overidentifying restrictions cannot be rejected, while the null hypothesis that the excluded exogenous variables do no explain market reform or institutional quality scores is strongly rejected.

Table 4.

First-Stage Estimation Results

(Exogenous determinants of reforms, institutions, and investment)

article image
Source: Author’s estimates. Note: ICRG = International Country Risk Guide. * and ** indicate significant at the 5 and 1 percent confidence levels, respectively.

Share of fuel, ores, metal, and agricultural raw exports in total exports.

Measured in 1,000 km and refers to distance from capital of respective country to Brussels.

Probability that two randomly selected individuals in a country are not from the same ethnic group.

Values of 1 for 1991–94; 2 for 1995–98; 3 for 1999–2002; and 4 for 2003–06.

Resource endowments and entrenchment of communist elite have together influenced the degree of market reforms or institutional building during the transition period. Hoff and Stiglitz (2004) suggest that a higher ratio of natural resources to GDP or total exports decreases the political constituency for institutional development. In transition economies with less dependence on natural resources (Albania, Armenia, Moldova, Georgia, and the Baltics), the elites had fewer incentives to clinch to power and thus were more likely to allow the introduction of market reforms and good governance. Beck and Laeven (2006) show that transition countries that rely more on natural resources (Russia, Azerbaijan, Kazakhstan, Tajikistan, and Uzbekistan) experienced less market reform or institutional development during the transition process.

Also, distance from the capitals of the transition economies to the center of Western Europe (say, Brussels) is widely acknowledged to be a primary force shaping the opportunity for interaction among states including eagerness for market reforms and institutional development. Distance from Brussels is also highly and positively correlated with the years under socialism. In contrast to the Baltics and Central European economies, most of the CIS economies had long periods of communist rule and their capitals are relatively far away from the center of Western Europe. Old elites of the Central and Eastern European and Baltic economies with closer proximity to the capital of Western European countries had fewer possibilities to maintain their power. The prospects of future European Union (EU) membership have fostered institutional building, both through political incentives and through assistance from the major EU members (Roland and Verdier, 2003). Easterly and Levine (1997) show that ethnic diversity tends to reduce the provision of public goods, including the institutions that support business transactions. Also, Alesina and others (2003) conclude that ethnic fractionalization is likely to be an important determinant of economic success, both in terms of output and quality of institutions. Ethnic fractionalization fosters rent-seeking and might not be conducive to the building of strong market institutions.

Results of the Transition Sample

The estimation results for the transition sample are reported in Table 5. The F-tests demonstrate that the instruments are related to the endogenous variables that are instrumenting for, at high levels of statistical significance. Overall, the fit is good for this type of panel data. In all cases, the variables have the theoretically expected sign, but their magnitude and significance differs depending on the variables included, frequency of the data used (annual or period averages), and the estimation techniques. There are several interesting findings:

Table 5.

Estimation Results for the Transition Sample, 1991–2006

article image
article image
Source: Author’s estimates. Note: TFP = total factor productivity; GMM = generalized method of moments; 2SLS = two-stage least squares. * and ** indicate significant at the 5 and 1 percent confidence levels, respectively.

The choice of a fixed effects versus a random effects specification is justified by a Hausman test.

Initial real GDP index (1990 = 100), constructed from the real GDP growth rates.

Average of eight EBRD transition reform indicators (price liberalization, competition policy, banking reform, trade and foreign exchange system, large-scale privatization, small-scale privatization, governance and enterprise reforms, and infrastructure). The transition indicators range from 1 to 4.3, with 1 representing little or no change from a rigid centrally planned economy and 4.3 representing the standards of an industrialized market economy.

Average of internal liberalization index, external liberalization index, and privatization index.

ICRG = International Country Risk Guide. The ICGR ratings reflect risk ratings of the following components: government stability, socioeconomic conditions, investment profile, internal conflict, external conflict, corruption, law and order, ethnic tensions, democratic accountability, and bureaucratic quality.

The null-hypothesis of the F-test is that the exogenous excluded variables do not explain market reforms or institutional development in the first stage. In addition to the exogenous instrumental variables, lagged values of the policy variables are also used as instruments in the annual data regressions.

First, the estimated coefficient on the recovery of lost output is negative, as expected, and highly significant both in the per capita and TFP growth regression equations. The recovery of lost output effect is sizable: according to the point estimate, given that the average real GDP index in 1996 was about 50 for the CIS (1990 = 100) as compared with 100 in the Central European economies, the difference in per capita growth is expected to be about 3 percentage points in favor of the CIS, assuming other things are equal.

Second, there is a strong link between progress in market reforms as measured both by the EBRD and Campos and Horváth (2006) reform indices on the one hand, and growth in per capita real GDP or TFP on the other hand. Unlike Fidrmuc (2003) and Lawson and Wang (2004) but in agreement with Falcetti, Lysenko, and Sanfey (2005), the estimated coefficients for the EBRD reform index in this study are always positive and highly significant. The magnitude of the estimated coefficient implies that if the average EBRD score for the CIS countries in 2006 were close to the three Baltics then the average growth would have been about 3 percentage points higher than the outcome for 2001–06. The estimated coefficient of market reform (EBRD or Campos and Horváth scores) is larger in magnitude and highly significant when other stabilization policy variables (inflation rate or fiscal balance) are excluded from the estimated equation (see column 1 of Table 5).

Third, unlike previous studies on transition economies, the results suggest that investment is one of the variables that have contributed to the recent rapid growth. The regressions are also estimated without the investment variable. The reason is that the interpretation of the role of this variable is problematic even after the endogeneity problem is addressed. Investment could be capturing the effects of structural reforms that are difficult to quantify, or are already included in the EBRD market reform index. Investment could also change for reasons other than those related to reforms (for example, the large investment in the oil and gas sectors in Azerbaijan and other resource-rich countries).

The estimated coefficient for the terms-of-trade index is also positive and highly significant. Its magnitude is larger when the fiscal balance variable is excluded from the estimated equation, due to the relatively strong correlation between terms of trade and fiscal balance (see Table 3). Favorable terms of trade in the commodity exporter transition countries have shifted their fiscal deficits to significant surpluses (particularly in Azerbaijan, Kazakhstan, and Russia).

As for other explanatory variables, sound macroeconomic policies (including smaller fiscal deficits and government size in the economy) are associated with higher growth in per capita and in TFP growth. It should be noted that the fiscal coefficient is quite large and robust to changes in the specification of the equation and the estimation technique.

Results of the Global Sample

Table 6 shows that all included variables have the right sign and are significant at the 1 percent level, except for the terms of trade and government which is marginally significant. The regional dummies were used to test the hypothesis that different regimes may have characteristics that affect growth differently. This is confirmed with respect to Southeast Asia, which, on average, performed better than did other regions in the period under consideration. The coefficients of the African and Latin American dummies are negative, but insignificant at the 5 percent level.

Table 6.

Estimation Results for the Global Sample, 1980–2006

article image
Source: Author’s estimates. Note: OLS = ordinary least squares; 2SLS = two-stage least squares; GMM = generalized method of moments. * and ** indicate significant at the 5 and 1 percent confidence levels, respectively. The following indices are used: 0 for countries with initial real GDP of more than 95 percent of the real GDP in 1990; 0.5 if real GDP was between 70 and 95 percent; and 1 if it was less than 70 percent. When per capita income and institutional quality are both low, the ability to take advantage of growth opportunities is limited. This effect is captured by the composite convergence term. A negative coefficient for the composite term implies that institutional quality improves, convergence accelerates.

The null-hypothesis of the F-test is that the exogenous excluded variables do not explain institutional development in the first stage. In addition to the exogenous instrumental variables, lagged values of the policy variables were used as instruments in the annual data regressions.

The coefficient of the output recovery index is highly significant and robust to different estimation techniques. The coefficient of 2.77 from the four-year nonoverlapping averages using the GMM technique implies that on average growth in the CIS economies was between 2 to 2.5 percentage points higher than in other countries, depending on the period under consideration. To recall, the output recovery index takes the value of 1 for countries whose initial level of real GDP was less than 70 percent of the real GDP of 1990; 0.5 if initial real GDP was between 70 and 95 percent; and 0 if initial real GDP was greater than 95 percent.

The estimated coefficient on the quality of institutions as measured by the ICRG political index is robust in magnitude and continues to be significant under different estimation techniques. Its significance increases when some of the macroeconomic policy variables are excluded from the right-hand side of the regression equation. This may reflect the impact of institutions on policy sustainability variables and indicate that institutions play a dominant role in explaining cross-country differences in growth. More efficient institutions allow an economy to produce the same output with fewer inputs; bad institutions lower incentives to invest, to work, and to save. The magnitude of the estimated coefficient implies that if the average ICGR for the CIS countries was close to the average for Southeast Asian economies, then the average growth would have been about 1.5 percentage points higher than the outcome for the last decade. The interaction term of ICRG with the log of per capita income is negative and significant, suggesting that lower income countries would benefit more than higher income countries for the same improvement in institutional quality.

The estimated negative coefficient for the population growth implies that faster population growth is associated with slower per capita real GDP growth. Lower population growth means more capital accumulation per worker and hence higher productivity growth. However, this should be interpreted with caution. Continued slow growth in the CIS, particularly in Belarus, Russia, and Ukraine, will be accompanied by changes in age structures, which in the long term could adversely impact the growth prospects.

More importantly, changes in the external environment over the past few years—as captured by the world cycle dummy for 2003–06—show clear favorable growth effect that was substantial and statistically significant. The recent favorable environment explains on average at least 1 percentage point increase in per capita output in most countries. This latter could reflect the rapid progress in technological innovation worldwide, lower interest rate, and easier access to capital markets for most developing and transition economies.

Robustness of the Results

The findings in this paper are robust to different econometric specifications and estimation techniques, using different measures of market reforms, and different samples.

A first estimation problem faced in this study is the decision of which explanatory variables to include in the growth equation. Variables could be significantly correlated with growth depending on which other variables are held constant. This is because economic theories are still not precise enough to decide on the determinants of growth. The high cross-correlation among some of the explanatory variables is also a problem (Table 3). For example, combining different sets of variables one finds that x1 is significant when the regression includes x2 and x3, but becomes insignificant when x4 is included or x2 excluded. In particular, when the inflation rate is included in the regressions, the estimated coefficients on the market reform measures become much smaller. In general, however, the conclusion that recovery of lost output, macroeconomic stabilization, and market reforms significantly contributed to the rapid growth in 2001–06 appears robust to the alternative estimation methodologies and the choice of control variables.

A second concern with the estimated coefficients is the possible sensitivity of the results to the assumption about the form of the growth regression. In particular, the explanatory variables in equation (2) enter the growth equation regression linearly and independently. This reflects an ad hoc assumption that the marginal effect of a change in explanatory variable is constant, both across different levels of the variable and across different economies. In this regard, this paper tested for the robustness of the results by allowing for two types of nonlinearities for the explanatory variables of interest in the panel regression equation. Thus, the paper includes a squared term for the EBRD or the Campos and Horváth measure of market reforms and the proxy for the recovery of lost output variables in the regression specification:

g = β 1 ⁢ Z + β 2 ⁢ Z 2 + λ ⁢ X + μ i + v t + ε i ⁢ t ⁢ . ( 2 )

The question of interest is whether the coefficient estimate β1 remains robust when a squared term is included (a secondary question is whether β2 is itself robust). Allowing for the inclusion of a squared term, the results show that the EBRD or the Campos and Horváth market reform and the recovery of lost output variables remain robust. In addition, the coefficients of the squared terms of these two variable (β2) are significant and have opposite signs as compared with β1.

Another possibility is that the partial effect of a variable on growth varies over different levels of development. For example, the marginal effect of market reforms (as measured by the EBRD score) could be quite different in Armenia than in Slovenia. One way to capture such linearity is to include an interaction term between the variable of interest and a measure of the country’s level of development (such as per capita income) in the regression specification. That is:

g = β 1 ⁢ Z + β 2 ⁢ Z * log ( Y 0 ) + λ ⁢ X + μ i + v t + ε i ⁢ t , ( 3 )

where Y0 measures the initial GDP per capita in purchasing power parity (PPP) U.S. dollars. Again the key question is whether β1 becomes robust when the interaction term is included. Again the core explanatory variables remain robust. Combined with the results from equation (2), this suggests some important nonlinearities in the correlation between market reforms and growth. The negative interaction term indicates that market reform has less of an effect at higher levels of development.

As an additional robustness check, which also sheds light on differences across country groups, the regressions were run on different subsamples—developed countries, transition, and low-income developing countries. The resulting estimated coefficients do show variations across the groups using a core set of explanatory variables (market reforms, institutions, and terms of trade). Although the variations behave directionally the same across country groups, the size and significance of coefficients are modestly different. Thus, there is support both for the unity of growth drivers and for the variation of impact in terms of their magnitude.

Contribution to Growth Changes

Changes in Growth Rates over Time

With the exception of Kyrgyzstan, all transition economies grew faster in 2001–06 than in 1996–2000. On average, the region grew by 5.3 percentage points a year faster in the latter period. In this regard, the estimated regressions can provide a useful decomposition of the importance of the various factors in explaining differences in growth between the two periods. For this, both the estimated coefficients from the main regression (based on the results in Table 5 using the GMM) and the actual values of the explanatory variables for the two periods under consideration are used. The objective here is to assess the contribution of the change in each category of explanatory variables to a country’s fitted growth equation. The difference between the average country growth performance in 2001–06, denoted by g1, and average growth performance in the same country in the previous period (1996–2000), denoted by g0, can be expressed as follows:

g 1 − g 0 = β ⁢ [ Z 1 − Z 0 ] + λ ⁢ [ X 1 − X 0 ] + γ ⁢ [ W 1 − W 0 ] + ε i ⁢ t ⁢ . ( 4 )

Z is the vector of “core explanatory variables” (recovery of lost output index, EBRD measure of market reforms, investment, and terms of trade). X comprises a set of control variables including fiscal balance, inflation rate, and government consumption. W can be interpreted as an exogenous world environment. It captures the extent to which unaccounted international exogenous factors related to growth (such as productivity of new inventions in 2001–06). εit is the residual (the difference between the actual and the predicted change in growth).

The results of this approach are reported in Table 7. The first two columns show the actual and the fitted changes in the growth rates between 2001–06 and 1996–2000. The effects of changes in external factors such as trade and other global favorable environment factors are shown in columns (3) to (4). The combined impact of macroeconomic stabilization and reforms (specifically lower inflation, improvement in the fiscal position, progress made in market reforms as measured by the EBRD, and smaller size of government in the economy as measured by the government consumption to GDP ratio) are reported in column (5).

Table 7.

Decomposition of Growth Increase between Periods

(2001–06 compared with 1996–2000, in percentage points]

article image
article image
Source: Author’s estimates.

The estimates predict changes in the per capita growth rates quite well for the CIS, and the Central European economies but less well for the Southeastern European economies. For the CIS as a whole, 2.9 percentage points of the 4.9 predicted increase in growth is explained by improvement in macroeconomic stabilization and reforms. Had the market reform been deeper in the CIS, its impact on growth would have been correspondingly larger when multiplied by the estimated marginal growth effect. For example, if reforms in the CIS had attained the levels observed in the Baltics or in the Central European economies, the resulting aggregate growth acceleration impact would have been about 2 percentage points higher.

Changes in Growth Rates Across Regions

Another key advantage of the panel sample in this paper is that it permits the employment of an alternative standard of comparison, relying on cross-regional comparative analysis, to supplement the country-by-country time-series dimension. In this case, the unit of analysis is a comparison of regional aggregates. Specifically, the focus is on explaining the sources of the growth difference between the 12 CIS economies on the one hand, and the eight Central European and Baltic countries, and 11 Southeast Asian economies on the other hand. The comparison here is based on the growth performance over eight years (1999–2006). The difference between the average growth for the CIS region, denoted by gCIS, and average growth for the Central European and Baltic countries, denoted by gCEB, is expressed in equation (5) using the estimated coefficients from the transition sample, but the difference between the CIS and the Southeast Asian (gSEA) economies is expressed in equation (6), using the relevant estimated coefficients from the global sample.

g C ⁢ I ⁢ S − g C ⁢ E ⁢ B = β ⁢ [ Z C ⁢ I ⁢ S − Z C ⁢ E ⁢ B ] + λ ⁢ [ X C ⁢ I ⁢ S − X C ⁢ E ⁢ B ] + ε i ⁢ t , ( 5 )
g C ⁢ I ⁢ S − g S ⁢ E ⁢ A = β ⁢ [ Z C ⁢ I ⁢ S − Z S ⁢ E ⁢ A ] + λ ⁢ [ X C ⁢ I ⁢ S − X S ⁢ E ⁢ A ] + ε i ⁢ t , ( 6 )

where the core and the control explanatory variables are defined as before for equation (5), in equation (6) instead of the market reform measure the institutional quality measure is used.

The first half of Table 8 compares CIS with the Central European and Baltic countries. The factors in favor of the Central European and Baltic countries were: (1) reforms were more advanced; (2) investment was higher; and (3) inflation was significantly lower. However, the positive impacts of these factors were more than offset by other factors in favor of the CIS, including recovery of lost output (initial, 1998, real GDP in the CIS was 59 percent of its level in 1990 as compared with 95 percent in the Central European and Baltic countries), and the terms-of-trade shocks were in favor of the CIS. The second half of Table 7 shows that despite weaker institutions, lower investment, and higher inflation in the CIS as compared with the Southeast Asian economies, recovery of lost output, slower population growth, and more favorable terms-of-trade gave the CIS an edge of 2.3 percentage points each year. The average score for the institutional quality in the Southeast Asian economies was 75 as compared with 64 for the CIS. Assuming hypothetically that the institutional scores of the CIS countries were raised to Southeast Asian economies’ level, then the resulting change in the fitted value for growth using the parameter estimates would be a gain of another 1.2 percentage points per year. These results indicate that institutional reforms could play a major role in sustaining the recent rapid growth in the CIS, particularly when the temporary factors (recovery of lost output and terms of trade) behind the recent rapid growth disappear.

Table 8.

Sources of Regional Differences in Growth, 1999–2006

article image
Source: Author’s estimates. Note: ICRG=International Country Risk Guide.

Southeast Asian countries include China, Hong Kong, India, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Thailand, and Vietnam.

III. Conclusions and Policy Challenges

This paper investigates the main factors behind the rapid growth in transition economies. The central conclusion from the growth accounting exercise is that most of the improvement came from higher TFP, which averaged about 3 percentage points of GDP per year in 2001–06, close to the TFP gains in Western Europe in 1950–60. During the initial years of transition, the disorganization or chaos resulting from the removal of central controls and coordination produced negative TFP growth rates as output fell, and a large part of the capital stock lay idle. Subsequently, as the economies achieved macroeconomic stability and market reforms, the reallocation of resources to more productive activities allowed the economies to generate rapid growth, especially in the CIS, with low rates of investment so that TFP growth rates increased.

Using panel data regression, the results suggest that the rapid per capita real GDP and TFP growth rates since 1999 is explained by the extent of how much transition economies have contracted in terms of real GDP in the 1990s and the degree of progress made in market reforms. Other factors, such as the terms of trade and macroeconomic stabilization, have also contributed to rapid growth. The results are robust to different panel econometric techniques, and different specifications. The main findings are as follows:

  • Transition countries that experienced larger declines in output during the early 1990s tended to grow at much faster rates. On average, of the 8 percent annual average growth rate for transition economies in 2001–06, about 2 percentage points are attributable to the recovery of lost output.

  • The growth impetus associated with macroeconomic stabilization and market reforms has been substantial because of their effect on the overall productivity. Had the market reform been deeper, its impact on TFP growth would have been correspondingly larger.

  • The growth acceleration payoff to reforms in 2001–06 was enhanced by the favorable external environment (positive terms-of-trade shock and global technological innovation). These factors have accounted for about two percentage points of the annual growth in transition economies. The global environment alone in recent years explains 1 percentage point of the annual average growth.

A key question for prospective growth is whether the gains achieved thus far have already eliminated most of the inefficiencies of central planning—and will therefore soon fade away. As the transition countries approach the world technology frontier, thereby exhausting the opportunity for further TFP growth from this source, alternative channels to improve TFP growth will need to be sought. Further improvement in policy and institutions would need to play a role in this endeavor. Also, greater labor use and the recent trend of faster capital accumulation are expected to play a more important role in the medium-term growth.

Appendix I. Sample, Data Definition, and Sources

List of Countries

The set of countries covered in this paper was determined by the availability of key variables; small countries (with population less than one million) were also excluded. The global sample consists of the following 123 countries during 1983–2006:

Transition sample: Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, the Kyrgyz Republic, Moldova, Mongolia, Russia, Tajikistan, Ukraine, Uzbekistan, Estonia, Latvia, Lithuania, Czech Republic, Hungary, Poland, Slovak Republic, Slovenia, Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Macedonia, and Romania.

Global sample: Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, the Kyrgyz Republic, Moldova, Mongolia, Russia, Tajikistan, Ukraine, Uzbekistan, Estonia, Latvia, Lithuania, Czech Republic, Hungary, Poland, Slovak Republic, Slovenia, Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Macedonia, Romania, Australia, Austria, Canada, Cyprus, Denmark, Finland, France, Germany, Greece, Ireland, Island, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, United Kingdom, the United States, Bangladesh, Cambodia, China, Hong Kong, India, Indonesia, Korea, Malaysia, the Philippines, Singapore, Sri-Lanka, Thailand, Vietnam, Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El-Salvador, Guatemala, Guyana, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay, Venezuela, Algeria, Egypt, Iran, Israel, Jordan, Lebanon, Morocco, Oman, Pakistan, Tunisia, Turkey, Yemen, Kuwait, Saudi Arabia, Syria, the UAE, Angola, Botswana, Burkina Faso, Cameroon, Central African Republic, Ethiopia, Ghana, Gabon, Guinea, Gambia, Guinea-Bissau, Ivory Coast, Kenya, Madagascar, Malawi, Mali, Mauritius, Mozambique, Namibia, Niger, Nigeria, Senegal, South Africa, Sudan, Togo, and Tanzania.

Sample Period

Annual and period average data were used. The transition sample is divided into four subperiods: 1991–94, 1995–98, 1999–2002, and 2003–06. The global sample is divided into seven subperiods: 1980–82, 1983–86, 1987–90, 1991–94, 1995–1998, 1999–2002, and 2003–06. The resulting information was unbalanced because of data limitations for some countries.

Definition and Sources of Data

The data sources used are the databases of the IMF, World Economic Outlook (WEO), United Nations Economic Commission for Europe, and the World Bank.

  • Per capita growth (dependent variable): Per capita real GDP growth rate calculated from national currencies in constant prices. Source: WEO database.

  • Convergence as measured by the initial income per capita in PPP-adjusted (U.S. dollars). Source: World Bank.

  • Investment: Fixed capital formation as percent of GDP. Source: WEO database.

  • Recovery of lost output for transition sample: An index of real GDP for all transition countries is constructed with real GDP for 1990 = 100. Source: Author’s calculations based on the annual real GDP growth rates.

  • Recovery of lost output for global sample: The following indices are used: (1) 0 if initial real GDP was greater than 95 percent of its value in 1990; (2) 0.5 if initial real GDP was between 70 and 95 percent; and (3) 1 if initial real GDP was less than 70 percent.

  • EBRD market reform index (only for transition countries): The unweighted average of eight EBRD structural reform indicators—price liberalization, small-scale privatization, large-scale privatization, competition policy, trade liberalization, financial sector reform, governance and enterprise reforms, and infrastructure reform. The EBRD indicators range from 1 to 4.3, where 4.3 indicates that the country’s structural characteristics are comparable to those prevailing on average in market economies, and 1 represents conditions before reform in a centrally planned economy with dominant state ownership of the means of production. The reform indices are not perfect and their assessment is sometimes influenced by the observed macroeconomic performance, which raises the problem of possible endogeneity. Source: EBRD, Transition Reports (various years).

  • Campos and Horvath (2006) market reform index is based on three categories of reform. The first captures internal liberalization reflecting the extent of price and wage liberalization. The second captures external liberalization efforts as reflected in the severity of trade barriers and capital controls. The third category captures the extent of privatization efforts. These three indices were constructed for 25 transition economies for all years between 1989 and 2000. Using Campos and Horvath’s methodology this paper extended the market reform indices through 2006.

  • The measure on institutional quality is taken from the ICRG, compiled by the private consultancy firm Political Risk Services. This data set covers 143 countries, from 1984 to the present. The composite index is an aggregation of various subcomponents that measure factors such as government stability, democratic accountability, law and order, quality of bureaucracy, and corruption in government. Source: Political Risk Services Group (2007), Syracuse University.

  • Ratio of government consumption to GDP: Source: IMF, International Financial Statistics (IFS) and WEO database.

  • Macroeconomic stabilization as measured by overall fiscal balance as a ratio of GDP, and the logarithm of the inflation rate. Source: IMF, WEO database.

  • External conditions: Terms-of-trade shocks: percentage change in the terms-of-trade index (2000 = 100). Source: IMF, WEO database. World cycle period-specific shifts: Time dummy variables for 1987–90, 1991–94, 1995–1998, 1999–2002, and 2003–06. Source: Author’s construction.

  • Instrumental variables: (1) Commodity exports as percent of total exports: is the share of fuel, ores, metal, and agricultural raw exports in total exports, UNCTAD; (2) distance to Brussels is measured in 1,000 kilometers and refers to the distance from the y capital of the respective country to Brussels; (3) ethnic fractionalization is presented as the probability that two randomly selected individuals in a country are not from the same ethnic group (Alesina and others, 2003); (4) a dummy variable for civil conflict or war; and (5) a trend with values of 1 for 1991–94, 2 for 1995–98, 3 for 1999–2003, and 4 for 2004–06.

Appendix II. Test of Weak Instruments

This appendix is based on Stock, Wright, and Yogo (2001, 2002). They proposed an approach to make interference about weak instruments.

y = Y ⁢ β + u , ( A .1 )
Y = Z ⁢ Π + v , ( A .2 )

where y and Y are T × 1 vectors of observations on endogenous variables, Z is T × K matrix of instruments, and u and v are T × 1 vectors of disturbance terms. The errors [ut vt]’ (t = 1,…,T) are assumed to be iid N (0, Σ), where the elements of Σ are σu2, and σu⁢v2 and σv2 and let ρ=σuv/(σuσv). The reduced equation (A.2) related the endogenous regressor to the instruments.

The concentration parameter, μ2, is a unitless measure of strength of the instruments and is defined as:

μ 2 = Π ′ ⁢ Z ′ ⁢ Z ⁢ Π / σ v 2 ⁢ . ( A .3 )

A useful interpretation of μ2 is in terms of F, the F statistics for testing the hypotheses Π = 0 in (A.2) (that is, the “first-stage F statistics”). Stock, Wright, and Yogo (2002) derive the following equations:

μ ⁢ ( β ^ 2 ⁢ S ⁢ L ⁢ S − β ) = ( σ υ / σ ν ) ⁢ ( Z u + S u ⁢ v / μ ) 1 + 2 ⁢ Z v / μ 2 + S u ⁢ v / μ 2 , ( A .4 )

where

Z u = ( Π ′ ⁢ Z ′ ⁢ u ) / ( σ u ⁢ Π ′ ⁢ Z ′ ⁢ Π ) , Z v = ( Π ′ ⁢ Z ′ ⁢ v ) / ( σ v ⁢ X ⁢ Π ′ ⁢ Z ′ ⁢ Π ) , S u ⁢ v = ( v ′ ⁢ P z ⁢ u ) / ( σ v ⁢ σ u ) , S v ⁢ v = ( v ′ ⁢ P z ⁢ v ) / σ v 2 ⁢ .

Under weak-instrument asymptotics, a threshold value µ2/K is implied. If the actual value of µ2/K exceeds this threshold, then the instruments are strong (for example, 2SLS relative, bias is ＜ 10 percent). Otherwise, the instruments are weak. The first-stage F statistics must be large, typically exceeding 10, for 2SLS inference to be reliable.

References

  • Abed, George, and Hamid Davoodi, 2000, “Corruption, Structural Reforms, and Economic Performance in the Transition Economies,” IMF Working Paper 00/132 (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Aiyer, Shekhar, and Carl-Johan Dalgaard, 2005, “Total Factor Productivity Revisited: A Dual Approach to Development Accounting,” IMF Staff Papers, Vol. 52, No. 1, pp. 82–102.

    • Search Google Scholar
    • Export Citation

  • Alesina, Alberto, W. Easterly, S. Kurlat, and R. Wacziarg, 2003, “Fractionalization,” Journal of Economic Growth, Vol. 8, No. 2 (June), pp. 155–94.

    • Search Google Scholar
    • Export Citation

  • Amador, João, and Carlos Coimbra, 2007, “Total Factor Productivity Growth in the G7 Countries: Different or Alike?” Banca de Portugal Working Paper 9/2007.

    • Search Google Scholar
    • Export Citation

  • Babetskii, I., and N.F. Campos, 2007, “Does Reform Work? An Econometric Examination of the Reform-Growth Puzzle,” CERGE-EI Working Paper No. 322 (Prague, Center for Economic Research and Graduate Education—Economics Institute).

    • Search Google Scholar
    • Export Citation

  • Barro, Robert, 1997, Determinants of Economic Growth: A Cross-Country Empirical Study (Cambridge, Massachusetts, MIT Press).

  • Barro, Robert, 1999, “Notes on Growth Accounting,” Journal of Economic Growth, Vol. 4 (June), pp. 119–37.

  • Barro, Robert, and Xavier Sala-i-Martin, 2004, Economic Growth (London, United Kingdom, MIT Press).

  • Beck, Thorsten, and Luc Laeven, 2006, “Institution Building and Growth in Transition Economies,” Journal of Economic Growth, Vol. 11, pp. 157–86.

    • Search Google Scholar
    • Export Citation

  • Boseworth, Barry P., and Susan M. Collins, 1996, “Economic Growth in East Asia: Accumulation versus Assimilation,” Brooking Papers on Economic Activity, Vol. 27, No. 2, pp. 135–204.

    • Search Google Scholar
    • Export Citation

  • Boseworth, Barry P., and Susan M. Collins, 2003, “The Empirics of Growth: An Update,” Brooking Papers on Economic Activity, No. 2, pp. 113–206.

    • Search Google Scholar
    • Export Citation

  • Broeck, M., and V. Koen, 2000, “The Great Contractions in Russia, the Baltics and the Other Countries of the former Soviet Union: A View from the Supply Side,” IMF Working Paper 2000/32 (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Caballero, Ricardo J., and Mohamad Hammour, 1994, “The Cleansing Effect of Recession,” American Economic Review, Vol. 84, No. 5, pp. 1350–68.

    • Search Google Scholar
    • Export Citation

  • Campos, N.F., and F. Coricelli, 2002, “Growth in Transition: What We Know, What We Don’t Know and What We Should,” Journal of Economic Literature, Vol. 40, pp. 793–836.

    • Search Google Scholar
    • Export Citation

  • Campos, N.F., F. Coricelli, and R. Horvath, 2006, “Reform Redux: Measurement, Determinants and Reversals,” CEPR Discussion Paper Series No. 5673 (London, Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation

  • Cerra, V., and S.C. Saxena, 2005a, “Growth Dynamics: The Myth of Economic Recovery,” IMF Working Paper 05/147 (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Cerra, V., and S.C. Saxena, 2005b, “Did Output Recover from the Asian Crisis?” IMF Staff Papers, Vol. 52, No. 1, pp. 1–23.

  • Christiansen, Laurits R., Dianne Cummings, and Dale W. Jorgenson, 1980, “Economic Growth, 1947–73: An International Comparison,” in New Developments in Productivity Measurement and Analysis, Studies in Income and Wealth, ed. by J.W. Kendrick and B. Vaccara (Chicago, University of Chicago Press).

    • Search Google Scholar
    • Export Citation

  • De Melo, M., C. Denizer, and A. Gelb, 1996, “Patterns of Transition from Plan to Market,” World Bank Economic Review, Vol. 10, No. 3, pp. 397–424.

    • Search Google Scholar
    • Export Citation

  • Easterly, William, 2003, “National Economic Policies and Economic Growth: A Reappraisal,” in Handbook of Economic Growth, ed. by P. Aghion and S. Durlauf (Amsterdam, Elsevier BV).

    • Search Google Scholar
    • Export Citation

  • Easterly, William, and Ross Levine, 1997, “Africa’s Growth Tragedy: Policies and Ethnic Divisions,” Quarterly Journal of Economics, Vol. 112, No. 4 (November), pp. 1203–50.

    • Search Google Scholar
    • Export Citation

  • European Bank for Reconstruction and Development (EBRD), various years, Transition Reports (London, European Bank for Reconstruction and Development).

    • Search Google Scholar
    • Export Citation

  • Falcetti, Elisabetta, Tatiana Lysenko, and Peter Sanfey, 2005, “Reforms and Growth in Transition: Re-examining the Evidence,” EBRD Working Paper No. 90 (London, European Bank for Reconstruction and Development).

    • Search Google Scholar
    • Export Citation

  • Fidrmuc, J., 2003, “Economic Reform, Democracy and Growth during Post-Communist Transition,” European Journal of Political Economy, Vol. 19, No. 3 (September), pp. 583–604.

    • Search Google Scholar
    • Export Citation

  • Fisher, Stanley, 1993, “The Role of Macroeconomic Factors in Growth,” Journal of Monetary Economics, Vol. 32, No. 3 (December), pp. 485–512.

    • Search Google Scholar
    • Export Citation

  • Friedman, Milton, 1993, “The ‘Plucking Model’ of Business Fluctuations Revisited,” Economic Inquiry, Vol. 31 (April), pp. 171–7.

    • Search Google Scholar
    • Export Citation

  • Hamilton, James D., 1989, “A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle,” Econometrica, Vol. 57 (March), pp. 357–84.

    • Search Google Scholar
    • Export Citation

  • Hauk, W.R., and R. Wacziarg, 2004, “A Monte Carlo Study of Growth Regressions” (unpublished; Palo Alto, Stanford University).

  • Hoff, K., and J.E. Stiglitz, 2004, “After the Big Bang? Obstacles to the Emergence of the Rule of Law in Post-Communist Societies,” American Economic Review, Vol. 94, No. 3 (June), pp. 753–63.

    • Search Google Scholar
    • Export Citation

  • Hsieh, Vhang-Tai, 2002, “What Explains the Industrial Revolution in East Asia? Evidence from the Factor Markets,” American Economic Review, Vol. 92, No. 3 (June), pp. 502–22.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund (IMF), 2006, World Economic Outlook, World Economic and Financial Surveys (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Islam, N., 1995, “Growth Empirics: A Panel Data Approach,” Quarterly Journal of Economics, Vol. 110, No. 4, pp. 1127–70.

  • Jorgenson, D.W., 2005, “Accounting for Growth in the Information Age,” in Handbook of Economic Growth, ed. by P. Aghion and S. Durlauf (London, North Holland), Vol. 1.

    • Search Google Scholar
    • Export Citation

  • Jorgenson, D.W., and E. Yip, 2001, “Whatever Happened to Productivity Growth?” in New Development in Productivity Analysis, ed. by C.R. Hulten, E.R. Dean, and M. J. Harper (Chicago, University of Chicago Press).

    • Search Google Scholar
    • Export Citation

  • Lawson, C., and H. Wang, 2004, “Economic Transition in Central and Eastern Europe and the former Soviet Union: Which Policies Worked,” University of Bath, Department of Economics and International Development Working Paper.

    • Search Google Scholar
    • Export Citation

  • Loukoianova, Elena, and Anna Uigovskaya, 2004, “Analysis of Recent Growth in Low-Income CIS Countries,” IMF Working Paper 2000/32 (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Maddison, A., 1996, “Macroeconomic Accounts for European Countries,” in Quantitative Aspects of Postwar European Economic Growth, ed. by B. van Ark and N.F.R. Crafts (Cambridge, United Kingdom, Cambridge University Press).

    • Search Google Scholar
    • Export Citation

  • Mankiw, Gregory, David Romer, and David Weil, 1992, “A Contribution to the Empirics of Economic Growth,” Quarterly Journal of Economics, Vol. 107, No. 2 (May), pp. 407–37.

    • Search Google Scholar
    • Export Citation

  • Mendoza, Enrique, 1997, “Terms of Trade Uncertainty and Economic Growth,” Journal of Development Economics, Vol. 54, No. 2 (December), pp. 323–56.

    • Search Google Scholar
    • Export Citation

  • Pagan, Adrian, and Aman Ullah, 1999, Nonparametric Econometrics (Cambridge, United Kingdom, Cambridge University Press).

  • Political Risk Services, 2007, International Country Risk Guide (East Syracuse, New York, The PRS Group, Inc.). Available via the Internet: www.countrydata.com.

    • Search Google Scholar
    • Export Citation

  • Rodrik, D., A. Subramanian, and F. Trebbi, 2004, “Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development,” Journal of Economic Growth, Vol. 9, No. 2 (June), pp. 131–65.

    • Search Google Scholar
    • Export Citation

  • Roland, G., and T. Verdier, 2003, “Law Enforcement and Transition,” European Economic Review, Vol. 47, No. 4 (August), pp. 669–85.

    • Search Google Scholar
    • Export Citation

  • Schadler, S., A. Mody, A. Abiad, and D. Leigh, 2006, “Growth in the Central and Eastern European Countries of the European Union,” IMF Occasional Paper No. 252 (Washington, International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Shigeru, Iwata, Mohsin S. Khan, and Hiroshi Murao, 2003, “Sources of Economic Growth in East Asia: A Nonparametric Assessment,” IMF Staff Papers, Vol. 50, No. 2, pp. 157–77.

    • Search Google Scholar
    • Export Citation

  • Stock, J.H., J.H. Wright, and M. Yogo, 2001, “Testing for Weak Instruments in Linear IV Regression” (unpublished; Cambridge, Harvard University).

    • Search Google Scholar
    • Export Citation

  • Stock, J.H., J.H. Wright, and M. Yogo, 2002, “A Survey of Weak Instruments and Weak Identification in Generalized Methods of Moments,” Journal of Business and Economic Statistics, Vol. 20, No. 4, pp. 518–29.

    • Search Google Scholar
    • Export Citation

  • Stock, J.H., and M. Yogo, 2005, “Testing for Weak Instruments in Linear Regression,” in Identification and Inference for Econometric Models: A Festschrift in Honor of Thomas J. Rothenberg, ed. by D.W.K. Andrews and J.H. Stock (Cambridge, United Kingdom, Cambridge University Press).

    • Search Google Scholar
    • Export Citation
*

Garbis Iradian is a senior economist with the Middle East and Central Asia Department of the IMF. The author benefited greatly from comments by David Owen, Robert Flood, and Rosalind Oliver (all IMF). Thanks also to Michael Landesmann (Director of Research of the Vienna Institute for International Economic Studies) for useful comments on an earlier version of this paper. In addition, the author would like to thank participants at the IMF’s Middle Eastern and Central Asian Department for discussion and comments.

1

TFP is a measure of elements such as managerial capabilities and organizational competence, research and development, intersectoral transfers of resources, increasing returns to scale, embodied technical progress, and diffusion of technology.

2

The majority of existing literature shows capital elasticity of 0.3 to 0.5 in industrial countries. The share of physical capital for industrial countries is likely to be lower than for developing countries where the marginal product of capital is higher (Boseworth and Collins, 1996, p. 155).

3

The United States saw per capita income growth averaging 2.4 percent a year between 1950 and 1973; over the same period, per capita income grew on average by 5 percent a year in Germany; and by slightly more than 8 percent in Japan.

5

The GMM panel estimator extracts consistent and efficient estimates of the impact of reforms and institutions on growth. It exploits the time-series variation in the data, accounts for unobserved country-specific effects, and controls for endogeneity of all the explanatory variables.

6

Based on Monte Carlo simulations, Hauk and Wacziarg (2004) argue that taking account of all the advantages and limitations of the different estimation procedures, the cross-section OLS estimator that averages data over longer periods might be the most efficient.

7

The EBRD market reform index ranges from 1 to 4.3, where 1 represents conditions before reform in a centrally planned economy with dominant state ownership of the means of production, and 4.3 indicates that the country’s structural characteristics are comparable to those prevailing on average in market economies.

8

A high rate of inflation is harmful to growth because it raises the cost of borrowing and thus lowers the rate of capital investment. At the same time, highly variable inflation makes it difficult and costly to forecast accurately costs and profits and hence investors and entrepreneurs may be reluctant to undertake new projects. Likewise, given that financial resources in the form of domestic savings and loans are limited, a larger fiscal deficit will mean that more of those limited resources must be devoted to financing the budget deficit. Fewer resources will thus be available for private sector investment.

9

Stock and Yogo (2005) develop formal tests based on the F-statistic for the null hypothesis: (1) the bias of 2SLS is > 10 percent of the bias based on OLS. The F-test rejects the null of weak instrumental variables at the 5 percent level if F>10.3. They also consider the null hypothesis that (2) the null rejection rate of the nominal 5 percent 2SLS t test concerning a has a rejection rate 10 percent or greater. In this case, the F-test rejects the null of weak IVs at the 5 percent level if F>24.6. The adverse effect of weak IVs on 2SLS depends on the degree of endogeneity present as measured by Puv, the correlation between the structural and reduced form errors u and v. But Puv is difficult to estimate precisely when the IVs are weak. In particular, Puv cannot be consistently estimated under weak IVs asymptotics. The F test of Stock and Yogo (2005) is designed to be valid for any value of Puv. For more details on the F-test, see Appendix II.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Papers
IMF Staff Papers, Volume 56, No. 4
Author:
International Monetary Fund. Research Dept.
Volume/Issue:
Volume 2009: Issue 004
Publisher:
International Monetary Fund
ISBN:
9781589069107
ISSN:
1020-7635
Pages:
295
DOI:
https://doi.org/10.5089/9781589069107.024
  • Figure 1.
    View in gallery
    Figure 1.

    Contributions to Average Real GDP Growth, 1996–2006

    (In percentage points of GDP)

  • Figure 2.
    View in gallery
    Figure 2.

    Total Factor Productivity Growth, Historical Perspective

    (In percentage points)

  • Figure 3.
    View in gallery
    Figure 3.

    Two Views of Lost Output Recovery

  • Figure 4.
    View in gallery
    Figure 4.

    Different Measures of Market Reforms and Institutional Quality are Correlated

    (Four-year nonoverlapping averages, 1991–2006)

Figure 1.

Contributions to Average Real GDP Growth, 1996–2006

(In percentage points of GDP)
Figure 2.

Total Factor Productivity Growth, Historical Perspective

(In percentage points)
Figure 3.

Two Views of Lost Output Recovery
Figure 4.

Different Measures of Market Reforms and Institutional Quality are Correlated

(Four-year nonoverlapping averages, 1991–2006)