- Email: [email protected]
Credit constraints, inequality and the growth gains from trade
Economics Letters 121 (2013) 43–47
Contents lists available at ScienceDirect
Economics Letters journal homepage: www.elsevier.com/locate/ecolet
Credit constraints, inequality and the growth gains from trade Mauro Caselli ∗ School of Economics, ASB, UNSW, Kensington, NSW 2033, Australia
highlights • • • • •
This paper studies the effect of wealth inequality on the growth gains from trade. The focus is on the role of credit constraints and financial development. The data cover manufacturing industries in developing countries. The analysis considers differences in growth rates pre- and post-trade liberalisation. High inequality has a negative impact on the growth rate in the number of firms.
article
info
Article history: Received 4 March 2013 Received in revised form 22 June 2013 Accepted 26 June 2013 Available online 2 July 2013 JEL classification: F10 F43 G20 L60 O40
abstract This paper tests the hypothesis that, in the presence of credit constraints, higher wealth inequality affects negatively the growth gains from trade liberalisation. Variations in the growth rate of value added – decomposed in the growth rate of the number of establishments and the growth rate in average size – of manufacturing industries in 34 developing countries before and after trade liberalisation are used to study the effects of inequality on the difference in growth under liberalised and nonliberalised regimes. The results show that the number of firms in industries with high dependence on external finance in countries with higher inequality grow significantly slower, in both statistical and economic terms, than in industries with low dependence on external finance in countries with lower inequality following a trade liberalisation relative to the closed-economy period. © 2013 Elsevier B.V. All rights reserved.
Keywords: Growth Inequality Trade liberalisation Credit constraints Developing countries
1. Introduction One of the foundations of modern economic theory is that free trade benefits the agents that take part in it. This simple idea can be extended to countries, even though standard trade theory is well aware that there might be winners and losers within each country. Moreover, gains from trade may not only be of a static nature, as an economy moves to a higher steady-state equilibrium, but also dynamic and associated with longer-term growth (Gustafsson and Segerstrom, 2010a,b). However, the empirical evidence seems to be less clear-cut. On the one hand, Dollar and Kraay (2004) suggest that trade liberali-
∗
Tel.: +61 293853561. E-mail address: [email protected].
0165-1765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.econlet.2013.06.038
sation has led to higher growth rates in all countries and, as a consequence, to lower poverty rates too. On the other hand, Rodríguez and Rodrik (2001) and Rodrik et al. (2004) argue that trade liberalisation may have actually played a much smaller role, if any, in raising living standards and instead they point at institutions and their improvements in the last few decades to explain higher GDP per capita. Wacziarg and Welch (2008) make a significant improvement in the trade and growth literature by making use of the timing of liberalisation in a within-country setting to identify the changes in growth and investment rates associated with discrete changes in trade policy. They find that countries on average grow faster after opening to trade, but this effect is not homogeneous. In the effort of understanding the cross-country variation in the growth outcomes of trade reforms, Caselli (2012) shows that the distribution of wealth in developing countries, proxied by land ownership, affects negatively the growth gains from trade
44
M. Caselli / Economics Letters 121 (2013) 43–47
liberalisation. He also finds preliminary evidence that the negative relationship is stronger at lower levels of financial development, i.e., when credit constraints are more binding. Therefore, this paper takes over where Caselli (2012) and Wacziarg and Welch (2008) left off by taking a closer look at the role of credit constraints. More specifically, the hypothesis to be tested empirically is whether inequality has an effect on growth, also decomposed in the intensive and extensive margins, in the aftermath of trade liberalisation in the presence of credit constraints. The paper makes use of the timing of liberalisation in a withincountry setting in combination with the methodology developed by Rajan and Zingales (1998). In their paper, Rajan and Zingales (1998) suggest that one way to check whether a channel is at work is to see whether industries that might be most affected by a channel grow differentially in countries where that channel is likely to be more operative (Rajan and Subramanian, 2008). Other empirical growth studies that have used this methodology include Braun (2003) and Braun and Raddatz (2007). For the purpose of this paper, this implies that growth rate differences between the period the economy is open and the period it is closed for industry–country pairs are regressed on the interaction between inequality at the country level and dependence on external finance at the industry level, a measure of the extent to which firms in a given industry are dependent on funds coming from an entrepreneur’s own wealth or borrowing. This approach is equivalent to ‘‘difference-in-difference-in-differences’’ thanks to the use of variation across industries and countries as well as changes between the periods before and after trade liberalisation. This makes it possible to establish a causal relationship from inequality to the growth gains of trade liberalisation through the credit constraints channel. The main finding is that industries more dependent on external financing in countries with high land inequality grow slower in terms of the number of firms after opening to trade relative to the closed-economy period. This is an important finding since the creation of new firms may, in turn, affect the potential for long-term growth and the competitiveness nature of an industry. The remainder of the paper is organised as follows. Section 2 describes the empirical specification used to test the paper’s hypothesis. Section 3 summarises the data used. Section 4 presents and discusses the results and some robustness checks. Section 5 concludes. 2. Econometric specification The starting point for the econometric model is the specification in Rajan and Zingales (1998) and Braun (2003) for panel data. This specification is augmented by terms interacting trade policy not only with land inequality but also with industry fixed effects and country fixed effects to control for different degrees of tradability across industries in all countries and different years of trade liberalisation across countries: Grow thskt = a0 + a1 Shareskt + a2 Ineqk · DepExtFins · Openkt
+ a3 FinDevkt · DepExtFins + ηs Openkt + νk Openkt + φt + ϵskt .
(1)
Grow thskt is the annual growth rate of industry s in country k averaged over time period t, Shareskt is the share of industry s in country k’s total value added in manufacturing at the beginning of period t and is included to control for convergence, DepExtFins represents the dependence on external finance at the industry level and is time invariant, Ineqk is land inequality at the country level and is time invariant, Openkt is a dummy that takes value 1 during the time the country is open to trade and 0 otherwise, FinDevkt is the level of financial development at the beginning of period t, ηs represents industry fixed effects, νk are country fixed effects, φt
are time period fixed effects and ϵskt is the idiosyncratic stochastic error term. Time t can only take two values based on a country’s openness status, i.e., closed or open to trade. Taking the difference of Eq. (1) under a liberalised regime, i.e. Open = 1, and the same equation under a nonliberalised regime, i.e. Open = 0, it gives the main equation to be estimated:
1Grow thsk = a0 + a1 1Sharesk + a2 Ineqk · DepExtFins + a3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk ,
(2)
where 1 represents the difference between the period when country k is open to trade and the period when it is closed to trade. In order to test whether inequality affects the ability to open a new firm after a country opens to trade, however, it is necessary to decompose the changes in the average growth rate of value added at the industry level into changes in the average growth rate of the number of firms and changes in the average growth rate in average size. This implies estimating the following equations:
1Grow thNOsk = b0 + b1 1ShareNOsk + b2 Ineqk · DepExtFins + b3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk (3) 1Grow thAv Sizesk = c0 + c1 1RelAv Sizesk + c2 Ineqk · DepExtFins + c3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk , (4) where Grow thNOsk is the average growth in the number of firms of industry s in country k, ShareNOsk is the share of industry s in country’s k total number of firms in manufacturing, Grow thAv Sizesk is the average growth in average firm size of industry s in country k and RelAv Sizesk is the average firm size of industry s in country k relative to the average firm size in country k’s manufacturing as a whole.1 Eqs. (2)–(4) can be estimated via the OLS estimator with standard errors clustered at the country level.2 3. Data sources The raw data cover an unbalanced panel of 34 developing countries and 28 manufacturing industries between 1963 and 2004, chosen according to the availability of data and whether countries switched to an open trade regime during this period.3 The analysis is limited to developing countries because land inequality is considered to be a good proxy for wealth inequality only in these countries. The number of observations (industry–country pairs) is 783 in the largest possible sample. The dependent variables are measured at the 3-digit industry level of the International Standard Industrial Classification (ISIC) as the differences in the average annual growth rate of value added, the average annual growth rate of the number of firms and the average annual growth rate in the average size of an establishment under liberalised and nonliberalised regimes and are calculated based on UNIDO’s (2012) database.4 From the same database it
1 Following Rajan and Zingales (1998), the results presented in this paper are robust to using changes in industry’s share of total manufacturing value added to control for convergence in Eqs. (3) and (4) instead of, respectively, changes in industry’s share of the number of firms and changes in relative average firm size. The additional results are available upon request. 2 The specification in differences is exactly equivalent to fixed effects considering that only two aggregate periods are included in the final specification, i.e., closed and open to trade. The results from the specification in differences are included because they are easier to interpret. 3 It should be noted that for most countries in the sample the raw data cover a shorter period. As an example, the data available for Bolivia go from 1981 to 2001, of which the first four years cover a period of closed economy and the last seventeen a period of open economy. 4 The average size in the industry is obtained by dividing the value added in the industry by the number of establishments.
M. Caselli / Economics Letters 121 (2013) 43–47
45
Table 1 Differences in industry growth rates before and after trade liberalisation. External dependence measured using All firms
Young firms
1Share
−4.14
−4.13
Ineq × DepExtFin
(0.95) −0.79 (0.50)
(0.95) −0.81* (0.47) −0.04 (0.26) Yes Yes 783 34 0.21
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
***
Yes Yes 783 34 0.21
***
***
−4.29 (1.02) −0.14 (0.29)
Yes Yes 754 34 0.21
Mature firms ***
−4.29
(1.02) −0.13 (0.28) 0.03 (0.07) Yes Yes 754 34 0.21
−4.13*** (0.95) 0.15 (0.56)
Yes Yes 755 34 0.20
−4.13*** (0.95) 0.11 (0.55) −0.13 (0.22) Yes Yes 755 34 0.20
Notes: The dependent variable in all the regressions is the difference in the average growth rate of each industry’s total value added under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. * Coefficients significantly different from zero at 10% level. *** Coefficients significantly different from zero at 1% level.
is possible to calculate industry’s share of total value added in manufacturing, industry’s share of the number of firms and average firm size by industry relative to average firm size in manufacturing as a whole.5 The Sachs and Warner (1995) Index is used to assess whether a country is deemed closed (value of 0) or open (value of 1) to trade. This paper uses a version of this dataset updated by Wacziarg and Welch (2008). A potential shortcoming of using the Sachs–Warner Index is that it does not distinguish between different industries within a country. Some robustness checks will be undertaken to control for this issue. The main industry characteristic used is the level of dependence on external finance, measured as the industry-level median of total investment minus cash flow from operations divided by total investment and taken from Rajan and Zingales (1998). Each industry’s external financial dependence is calculated on the basis of all United States (US) companies in that industry as well as separately for young and mature US companies in that same industry. The main assumptions behind using this variable to rank industries according to how much credit constraints and wealth inequality may matter for their growth are that US firms face relatively minor constraints to accessing external finance. Moreover, for technological reasons, which persist across countries, some industries may depend more than others on external finance and young and mature firms, even in the same industry, may require different levels of external finance. The basic country characteristic is the level of inequality in the distribution of land, which is measured by the Gini coefficient at the beginning of the whole period and is taken from Frankema (2006). Land inequality is measured in levels because it is only available at one point in time (around 1960). Other countries’ characteristics, measured in changes between the two trade regimes, are interacted with the external financing of an industry to control for other changes that may be occurring at the same time as trade liberalisation. The main one is financial development, proposed by Rajan and Zingales (1998) and measured as private credit based on the dataset from Beck et al. (2000). Additional regressions in the robustness checks control for changes in the level of education, calculated as the average years of schooling for the population over the age of 15 and taken from Thomas et al. (2002), who, in turn, make use of the Barro–Lee dataset (Barro and Lee, 2001), and changes in per capita real GDP at purchasing power parity (PPP), taken from Heston et al. (2006) (Penn World Table).
5 The value added data is originally in current values of the local currency. Therefore, the GDP deflator from the World Bank’s World Development Indicators is used to transform the series in constant value terms.
4. Results This section discusses the results obtained from estimating Eqs. (2)–(4). All the regressions include country and industry fixed effects and the standard errors shown in parentheses are clustered at the country level. All the regressions use the OLS estimator, yet due to the difference taken between the values under the liberalised and nonliberalised regimes as well as the industry and country fixed effects, the methodology is equivalent to a differencein-difference-in-differences approach. Table 1 presents the results obtained from regressing the difference in the growth of valued added at the industry level under liberalised and nonliberalised regimes on the interaction between land inequality and three different definitions of dependence on external finance as explained above. The results show that in the first four regressions presented the coefficient on the interacted term between land inequality and external financing is negative; however, only in one out of these four specifications is the negative coefficient also statistically significant at least at the 10% level (column 2). When dependence on external finance is measured using mature firms, the coefficient on the interacted term with land inequality becomes positive but highly insignificant. These results seem to suggest that there is no clear evidence indicating that industries with high dependence on external finance in countries with low land inequality grow faster in terms of overall output. However, as shown in the next tables, the picture becomes clearer when the growth rate of value added is decomposed into the growth rate of the number of establishments and the growth in average size. Regarding the other regressors, the coefficient on changes in an industry’s share of total value added in manufacturing under liberalised and nonliberalised regimes is always negative and highly significant. This suggests that conditional convergence is at work in this sample. On the other hand, the interaction between the difference in private credit under liberalised and nonliberalised regimes and external financial dependence is never significant. Table 2 turns to the regressions using the difference in the growth rate of the number of establishments under liberalised and nonliberalised regimes as the dependent variable. The coefficient on the interaction between land inequality and external financial dependence is now negative in all six columns, but it is significant at the 5% level only when external dependence is measured using young firms. The negative and significant coefficient on the interacted term implies that industries with high dependence on external finance for young firms in countries with high inequality experience lower growth rates in the number of new firms after
46
M. Caselli / Economics Letters 121 (2013) 43–47 Table 2 Differences in growth in number of establishments before and after trade liberalisation. External dependence measured using All firms
1ShareNO
−4.95
Ineq × DepExtFin
(2.51) −2.12 (1.41)
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
Yes Yes 783 34 0.47
*
Young firms *
−4.93 (2.51) −2.34 (1.46) −0.71 (0.63) Yes Yes 783 34 0.47
*
−4.48 (2.37) −0.86** (0.37)
Yes Yes 754 34 0.49
Mature firms *
−4.48
(2.37) −0.90** (0.39) −0.14 (0.14) Yes Yes 754 34 0.49
−5.06* (2.59) −0.81 (1.57)
Yes Yes 755 34 0.47
−5.05* (2.59) −0.95 (1.57) −0.47 (0.36) Yes Yes 755 34 0.47
Notes: The dependent variable in all the regressions is the difference in the growth rate of the number of establishments in each industry under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. * Coefficients significantly different from zero at 10% level. ** Coefficients significantly different from zero at 5% level.
liberalisation compared to the period when the country is closed to trade. Therefore, high inequality puts a strain on the growth gains from trade by affecting people’s ability to open up a new business in a sector where credit constraints are more binding for new firms following a trade liberalisation. In terms of the size of the coefficient and considering the specification robust to the inclusion of the interaction between the difference in private credit under liberalised and nonliberalised regimes and external financial dependence (column 4), a one standard deviation decrease in a country’s land inequality (approximately 16 points) would increase the growth rate of the number of firms of the industry at the 75th percentile of the distribution of external financial dependence by approximately 7.5% points more than that of the industry at the 25th percentile in the period following trade liberalisation relative to the previous period. Table 3 presents the results of the regressions with the difference in growth in average size as the dependent variable. In this case, the R-squared shows that the variance in the regressors can explain only around 19% of the variance of the dependent variable, compared to about 21% in the first set of regressions and over 45% in the second set of regressions. Three out of the six coefficients on the interaction terms between land inequality and the dependence on external finance are positive, but they all turn out to be insignificant due to large standard errors. This suggests that inequality does not interact with credit constraints to affect the growth opportunities of a firm that already produces once a country opens up to trade. The results presented so far may be due to other channels not related to inequality. Thus, in order to confirm the importance of inequality, this section presents a set of robustness checks that indeed points at inequality to explain growth differences following trade liberalisation. Given the results above, the focus of the following discussion is on the regressions based on the difference of the growth of the number of firms under liberalised and nonliberalised regimes as the dependent variables and external financing measured using only young firms.6 Following the discussion in Rajan and Zingales (1998), the magnitude and significance of the coefficient on land inequality in column 4 of Table 2 are checked against the inclusion of interaction terms between external financing and, separately, real GDP per capita to control for economic development and average years of
6 Additional tables of results are available upon request.
schooling to control for human capital.7 The results are not statistically different. Even after controlling for economic development and human capital, the coefficient on land inequality may still be capturing the effect of other variables, such as institutions or liberalisation occurring at different times for different industries within the same country, a potential shortcoming of using the Sachs–Warner Index as a measure of trade liberalisation. The main difficulty, however, is getting hold of reliable data to control for such variables. Therefore, additional robustness checks include region-industry fixed effects to act as a catch-all vector for all these potentially heterogeneous effects. The use of regional dummies to control for institutions is not new and, indeed, Acemoglu et al. (2001) and Frankema (2006) show that institutions vary more widely across regions than within them. It is also arguable that countries in the same region have usually adopted similar trade policies in terms of which industries to protect (see Goldberg and Pavcnik (2007)). No significant change in the coefficient on the interacted term on land inequality in the regression for the growth in the number of establishments in differences can be detected and, thus, the coefficient is still negative and statistically significant at the 5% level. A further set of regressions include fixed effects interacting industry and the decade in which the country liberalised. The rationale behind the inclusion of this variable is that certain industries might have been favoured from being exposed to international markets at a particular time and, thus, the timing of liberalisation might be correlated with land inequality. While Caselli (2012) shows that there seems to be no such correlation between land inequality and the timing of liberalisation in a similar sample of countries, it is still important to control for this potential issue. The results confirm that this is not a potential source for biased coefficients since none of the coefficients change significantly. The final robustness check looks at the role of credit constraints in more detail. A potential explanation for the findings above is that trade liberalisation may cause structural change and, thus, some firms to close as new ones open. People’s ability to open a new firm depends on an industry’s entry costs – measured by the external financing needs for new firms – that are unlikely to be paid from cash flows and need instead to be financed either from entrepreneurs’ own wealth or from borrowing. This suggests that in countries with lower financial development, i.e., where borrowing to finance new activities is constrained, entrepreneurs’
7 The difference between the two papers is that in the current setting these two variables are measured in differences between the initial value of the period when the country is open and the initial value of the period when it is closed.
M. Caselli / Economics Letters 121 (2013) 43–47
47
Table 3 Differences in growth in average size before and after trade liberalisation. External dependence measured using All firms
Young firms
1RelAv Size
−6.41
−6.56
Ineq × DepExtFin
(2.01) −0.23 (0.64)
(2.09) −0.39 (0.64) −0.53 (0.42) Yes Yes 783 34 0.19
***
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
Yes Yes 783 34 0.19
***
***
−8.07 (2.63) 0.02 (0.35)
Yes Yes 754 34 0.20
Mature firms ***
−8.08
(2.65) −0.02 (0.36) −0.13 (0.30) Yes Yes 754 34 0.20
−6.40*** (2.00) 0.33 (0.26)
Yes Yes 755 34 0.19
−6.47*** (2.01) 0.23 (0.26) −0.33** (0.14) Yes Yes 755 34 0.19
Notes: The dependent variable in all the regressions is the difference in the growth rate in average size in each industry (measured as value added divided by the number of firms) under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. ** Coefficients significantly different from zero at 5% level. *** Coefficients significantly different from zero at 1% level.
own wealth, and thus wealth inequality, should have a greater role in affecting the growth gains from trade. In order to test this hypothesis in more detail, the regressions in column 4 of Tables 1 and 2 are run with the current sample of countries split between those with a level of financial development at the beginning of the whole period below the median and those above the median. When the growth rate of value added in differences is used as the dependent variable, the results show that the coefficient on the interaction term between land inequality and dependence on external finance is not statistically different between the two samples. On the other hand, when the dependent variable is the difference in the growth of the number of firms, the coefficient on the interaction term with land inequality is negative and significant at 5% level only for those countries with a level of financial development below the median. This confirms that there is some evidence that land inequality affects the growth gains from trade through the credit constraints channel when domestic financial institutions are not well developed. 5. Conclusion This paper tests the hypothesis that credit constraints are the main mechanism through which wealth inequality, here proxied by land inequality, affects negatively the growth gains from trade liberalisation in developing countries. This question is important because empirical evidence shows that not all countries grow faster after opening to trade, as predicted by standard models and hoped by trade liberalisers. Therefore, understanding which factors can limit a country’s growth potential after trade liberalisation can help policymakers design policies that augment the benefits derived from trade liberalisation. The empirical analysis finds a negative but insignificant coefficient when the difference in growth of value added under liberalised and nonliberalised regimes is regressed on the interaction term between inequality and three different measures of external financial dependence, but a negative and significant coefficient is found on the same regressor when growth of the number of establishments in differences is used as the dependent variable and land inequality is interacted with dependence on external finance measured using only young US companies in each industry. This evidence suggests that in developing countries inequality has a long-term effect on the growth gains from trade liberalisation by affecting agents’ decision to start up a new business and that availability of credit to a wider range of firms, i.e., financial widening, can help alleviate this negative effect on growth.
Acknowledgements I thank participants at the ETSG, seminar participants at UNSW and Monash University and an anonymous referee for helpful comments. I gratefully acknowledge that this research was financially supported by an Australian Research Council grant to Professor Alan Woodland. All errors are mine. References Acemoglu, D., Johnson, S., Robinson, J.A., 2001. The colonial origins of comparative development: an empirical investigation. American Economic Review 91, 1369–1401. Barro, R.J., Lee, J.-W., 2001. International data on educational attainment: updates and implications. Oxford Economic Papers 53, 541–563. Beck, T., Demirgüç-Kunt, A., Levine, R., 2000. A new database on financial development and structure. World Bank Economic Review 14, 597–605. Braun, M., 2003. Financial contractibility and assets’ hardness: industrial composition and growth. Harvard University. Braun, M., Raddatz, C., 2007. Trade liberalization, capital account liberalization and the real effects of financial development. Journal of International Money and Finance 26, 730–761. Caselli, M., 2012. Does wealth inequality reduce the gains from trade? Review of World Economics 148, 333–356. Dollar, D., Kraay, A., 2004. Trade, growth, and poverty. Economic Journal 114, F22–F49. Frankema, E., 2006. The colonial origins of inequality: exploring the causes and consequences of land distribution. In: Research Memorandum GD-81. University of Groningen, Groningen Growth and Development Centre. Goldberg, P.K., Pavcnik, N., 2007. Distributional effects of globalization in developing countries. Journal of Economic Literature 45, 39–82. Gustafsson, P., Segerstrom, P., 2010a. North–south trade with increasing product variety. Journal of Development Economics 92, 97–106. Gustafsson, P., Segerstrom, P., 2010b. Trade liberalization and productivity growth. Review of International Economics 18, 207–228. Heston, A., Summers, R., Aten, B., 2006. Penn World Table Version 6.2. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. Rajan, R.G., Subramanian, A., 2008. Aid and growth: what does the cross-country evidence really show? The Review of Economics and Statistics 90, 643–665. Rajan, R.G., Zingales, L., 1998. Financial dependence and growth. American Economic Review 88, 559–586. Rodríguez, F., Rodrik, D., 2001. Trade policy and economic growth: a skeptic’s guide to the cross-national evidence. In: Bernanke, B.S., Rogoff, K. (Eds.), NBER Macroeconomics Annual 2000. In: NBER Chapters, vol. 15. National Bureau of Economic Research, Inc., pp. 261–338. Rodrik, D., Subramanian, A., Trebbi, F., 2004. Institutions rule: the primacy of institutions over geography and integration in economic development. Journal of Economic Growth 9, 131–165. Sachs, J., Warner, A., 1995. Economic reform and the progress of global integration. Working Papers 1733. Harvard Institute of Economic Research. Thomas, V., Wang, Y., Fan, X., 2002. A new dataset on inequality in education: gini and theil indices of schooling for 140 countries, 1960–2000, the World Bank. UNIDO, 2012. Industrial statistics database. 3-digit ISIC code, Rev. 2. Wacziarg, R., Welch, K.H., 2008. Trade liberalization and growth: new evidence. World Bank Economic Review 22, 187–231.
Contents lists available at ScienceDirect
Economics Letters journal homepage: www.elsevier.com/locate/ecolet
Credit constraints, inequality and the growth gains from trade Mauro Caselli ∗ School of Economics, ASB, UNSW, Kensington, NSW 2033, Australia
highlights • • • • •
This paper studies the effect of wealth inequality on the growth gains from trade. The focus is on the role of credit constraints and financial development. The data cover manufacturing industries in developing countries. The analysis considers differences in growth rates pre- and post-trade liberalisation. High inequality has a negative impact on the growth rate in the number of firms.
article
info
Article history: Received 4 March 2013 Received in revised form 22 June 2013 Accepted 26 June 2013 Available online 2 July 2013 JEL classification: F10 F43 G20 L60 O40
abstract This paper tests the hypothesis that, in the presence of credit constraints, higher wealth inequality affects negatively the growth gains from trade liberalisation. Variations in the growth rate of value added – decomposed in the growth rate of the number of establishments and the growth rate in average size – of manufacturing industries in 34 developing countries before and after trade liberalisation are used to study the effects of inequality on the difference in growth under liberalised and nonliberalised regimes. The results show that the number of firms in industries with high dependence on external finance in countries with higher inequality grow significantly slower, in both statistical and economic terms, than in industries with low dependence on external finance in countries with lower inequality following a trade liberalisation relative to the closed-economy period. © 2013 Elsevier B.V. All rights reserved.
Keywords: Growth Inequality Trade liberalisation Credit constraints Developing countries
1. Introduction One of the foundations of modern economic theory is that free trade benefits the agents that take part in it. This simple idea can be extended to countries, even though standard trade theory is well aware that there might be winners and losers within each country. Moreover, gains from trade may not only be of a static nature, as an economy moves to a higher steady-state equilibrium, but also dynamic and associated with longer-term growth (Gustafsson and Segerstrom, 2010a,b). However, the empirical evidence seems to be less clear-cut. On the one hand, Dollar and Kraay (2004) suggest that trade liberali-
∗
Tel.: +61 293853561. E-mail address: [email protected].
0165-1765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.econlet.2013.06.038
sation has led to higher growth rates in all countries and, as a consequence, to lower poverty rates too. On the other hand, Rodríguez and Rodrik (2001) and Rodrik et al. (2004) argue that trade liberalisation may have actually played a much smaller role, if any, in raising living standards and instead they point at institutions and their improvements in the last few decades to explain higher GDP per capita. Wacziarg and Welch (2008) make a significant improvement in the trade and growth literature by making use of the timing of liberalisation in a within-country setting to identify the changes in growth and investment rates associated with discrete changes in trade policy. They find that countries on average grow faster after opening to trade, but this effect is not homogeneous. In the effort of understanding the cross-country variation in the growth outcomes of trade reforms, Caselli (2012) shows that the distribution of wealth in developing countries, proxied by land ownership, affects negatively the growth gains from trade
44
M. Caselli / Economics Letters 121 (2013) 43–47
liberalisation. He also finds preliminary evidence that the negative relationship is stronger at lower levels of financial development, i.e., when credit constraints are more binding. Therefore, this paper takes over where Caselli (2012) and Wacziarg and Welch (2008) left off by taking a closer look at the role of credit constraints. More specifically, the hypothesis to be tested empirically is whether inequality has an effect on growth, also decomposed in the intensive and extensive margins, in the aftermath of trade liberalisation in the presence of credit constraints. The paper makes use of the timing of liberalisation in a withincountry setting in combination with the methodology developed by Rajan and Zingales (1998). In their paper, Rajan and Zingales (1998) suggest that one way to check whether a channel is at work is to see whether industries that might be most affected by a channel grow differentially in countries where that channel is likely to be more operative (Rajan and Subramanian, 2008). Other empirical growth studies that have used this methodology include Braun (2003) and Braun and Raddatz (2007). For the purpose of this paper, this implies that growth rate differences between the period the economy is open and the period it is closed for industry–country pairs are regressed on the interaction between inequality at the country level and dependence on external finance at the industry level, a measure of the extent to which firms in a given industry are dependent on funds coming from an entrepreneur’s own wealth or borrowing. This approach is equivalent to ‘‘difference-in-difference-in-differences’’ thanks to the use of variation across industries and countries as well as changes between the periods before and after trade liberalisation. This makes it possible to establish a causal relationship from inequality to the growth gains of trade liberalisation through the credit constraints channel. The main finding is that industries more dependent on external financing in countries with high land inequality grow slower in terms of the number of firms after opening to trade relative to the closed-economy period. This is an important finding since the creation of new firms may, in turn, affect the potential for long-term growth and the competitiveness nature of an industry. The remainder of the paper is organised as follows. Section 2 describes the empirical specification used to test the paper’s hypothesis. Section 3 summarises the data used. Section 4 presents and discusses the results and some robustness checks. Section 5 concludes. 2. Econometric specification The starting point for the econometric model is the specification in Rajan and Zingales (1998) and Braun (2003) for panel data. This specification is augmented by terms interacting trade policy not only with land inequality but also with industry fixed effects and country fixed effects to control for different degrees of tradability across industries in all countries and different years of trade liberalisation across countries: Grow thskt = a0 + a1 Shareskt + a2 Ineqk · DepExtFins · Openkt
+ a3 FinDevkt · DepExtFins + ηs Openkt + νk Openkt + φt + ϵskt .
(1)
Grow thskt is the annual growth rate of industry s in country k averaged over time period t, Shareskt is the share of industry s in country k’s total value added in manufacturing at the beginning of period t and is included to control for convergence, DepExtFins represents the dependence on external finance at the industry level and is time invariant, Ineqk is land inequality at the country level and is time invariant, Openkt is a dummy that takes value 1 during the time the country is open to trade and 0 otherwise, FinDevkt is the level of financial development at the beginning of period t, ηs represents industry fixed effects, νk are country fixed effects, φt
are time period fixed effects and ϵskt is the idiosyncratic stochastic error term. Time t can only take two values based on a country’s openness status, i.e., closed or open to trade. Taking the difference of Eq. (1) under a liberalised regime, i.e. Open = 1, and the same equation under a nonliberalised regime, i.e. Open = 0, it gives the main equation to be estimated:
1Grow thsk = a0 + a1 1Sharesk + a2 Ineqk · DepExtFins + a3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk ,
(2)
where 1 represents the difference between the period when country k is open to trade and the period when it is closed to trade. In order to test whether inequality affects the ability to open a new firm after a country opens to trade, however, it is necessary to decompose the changes in the average growth rate of value added at the industry level into changes in the average growth rate of the number of firms and changes in the average growth rate in average size. This implies estimating the following equations:
1Grow thNOsk = b0 + b1 1ShareNOsk + b2 Ineqk · DepExtFins + b3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk (3) 1Grow thAv Sizesk = c0 + c1 1RelAv Sizesk + c2 Ineqk · DepExtFins + c3 1FinDevk · DepExtFins + ηs + νk + 1ϵsk , (4) where Grow thNOsk is the average growth in the number of firms of industry s in country k, ShareNOsk is the share of industry s in country’s k total number of firms in manufacturing, Grow thAv Sizesk is the average growth in average firm size of industry s in country k and RelAv Sizesk is the average firm size of industry s in country k relative to the average firm size in country k’s manufacturing as a whole.1 Eqs. (2)–(4) can be estimated via the OLS estimator with standard errors clustered at the country level.2 3. Data sources The raw data cover an unbalanced panel of 34 developing countries and 28 manufacturing industries between 1963 and 2004, chosen according to the availability of data and whether countries switched to an open trade regime during this period.3 The analysis is limited to developing countries because land inequality is considered to be a good proxy for wealth inequality only in these countries. The number of observations (industry–country pairs) is 783 in the largest possible sample. The dependent variables are measured at the 3-digit industry level of the International Standard Industrial Classification (ISIC) as the differences in the average annual growth rate of value added, the average annual growth rate of the number of firms and the average annual growth rate in the average size of an establishment under liberalised and nonliberalised regimes and are calculated based on UNIDO’s (2012) database.4 From the same database it
1 Following Rajan and Zingales (1998), the results presented in this paper are robust to using changes in industry’s share of total manufacturing value added to control for convergence in Eqs. (3) and (4) instead of, respectively, changes in industry’s share of the number of firms and changes in relative average firm size. The additional results are available upon request. 2 The specification in differences is exactly equivalent to fixed effects considering that only two aggregate periods are included in the final specification, i.e., closed and open to trade. The results from the specification in differences are included because they are easier to interpret. 3 It should be noted that for most countries in the sample the raw data cover a shorter period. As an example, the data available for Bolivia go from 1981 to 2001, of which the first four years cover a period of closed economy and the last seventeen a period of open economy. 4 The average size in the industry is obtained by dividing the value added in the industry by the number of establishments.
M. Caselli / Economics Letters 121 (2013) 43–47
45
Table 1 Differences in industry growth rates before and after trade liberalisation. External dependence measured using All firms
Young firms
1Share
−4.14
−4.13
Ineq × DepExtFin
(0.95) −0.79 (0.50)
(0.95) −0.81* (0.47) −0.04 (0.26) Yes Yes 783 34 0.21
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
***
Yes Yes 783 34 0.21
***
***
−4.29 (1.02) −0.14 (0.29)
Yes Yes 754 34 0.21
Mature firms ***
−4.29
(1.02) −0.13 (0.28) 0.03 (0.07) Yes Yes 754 34 0.21
−4.13*** (0.95) 0.15 (0.56)
Yes Yes 755 34 0.20
−4.13*** (0.95) 0.11 (0.55) −0.13 (0.22) Yes Yes 755 34 0.20
Notes: The dependent variable in all the regressions is the difference in the average growth rate of each industry’s total value added under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. * Coefficients significantly different from zero at 10% level. *** Coefficients significantly different from zero at 1% level.
is possible to calculate industry’s share of total value added in manufacturing, industry’s share of the number of firms and average firm size by industry relative to average firm size in manufacturing as a whole.5 The Sachs and Warner (1995) Index is used to assess whether a country is deemed closed (value of 0) or open (value of 1) to trade. This paper uses a version of this dataset updated by Wacziarg and Welch (2008). A potential shortcoming of using the Sachs–Warner Index is that it does not distinguish between different industries within a country. Some robustness checks will be undertaken to control for this issue. The main industry characteristic used is the level of dependence on external finance, measured as the industry-level median of total investment minus cash flow from operations divided by total investment and taken from Rajan and Zingales (1998). Each industry’s external financial dependence is calculated on the basis of all United States (US) companies in that industry as well as separately for young and mature US companies in that same industry. The main assumptions behind using this variable to rank industries according to how much credit constraints and wealth inequality may matter for their growth are that US firms face relatively minor constraints to accessing external finance. Moreover, for technological reasons, which persist across countries, some industries may depend more than others on external finance and young and mature firms, even in the same industry, may require different levels of external finance. The basic country characteristic is the level of inequality in the distribution of land, which is measured by the Gini coefficient at the beginning of the whole period and is taken from Frankema (2006). Land inequality is measured in levels because it is only available at one point in time (around 1960). Other countries’ characteristics, measured in changes between the two trade regimes, are interacted with the external financing of an industry to control for other changes that may be occurring at the same time as trade liberalisation. The main one is financial development, proposed by Rajan and Zingales (1998) and measured as private credit based on the dataset from Beck et al. (2000). Additional regressions in the robustness checks control for changes in the level of education, calculated as the average years of schooling for the population over the age of 15 and taken from Thomas et al. (2002), who, in turn, make use of the Barro–Lee dataset (Barro and Lee, 2001), and changes in per capita real GDP at purchasing power parity (PPP), taken from Heston et al. (2006) (Penn World Table).
5 The value added data is originally in current values of the local currency. Therefore, the GDP deflator from the World Bank’s World Development Indicators is used to transform the series in constant value terms.
4. Results This section discusses the results obtained from estimating Eqs. (2)–(4). All the regressions include country and industry fixed effects and the standard errors shown in parentheses are clustered at the country level. All the regressions use the OLS estimator, yet due to the difference taken between the values under the liberalised and nonliberalised regimes as well as the industry and country fixed effects, the methodology is equivalent to a differencein-difference-in-differences approach. Table 1 presents the results obtained from regressing the difference in the growth of valued added at the industry level under liberalised and nonliberalised regimes on the interaction between land inequality and three different definitions of dependence on external finance as explained above. The results show that in the first four regressions presented the coefficient on the interacted term between land inequality and external financing is negative; however, only in one out of these four specifications is the negative coefficient also statistically significant at least at the 10% level (column 2). When dependence on external finance is measured using mature firms, the coefficient on the interacted term with land inequality becomes positive but highly insignificant. These results seem to suggest that there is no clear evidence indicating that industries with high dependence on external finance in countries with low land inequality grow faster in terms of overall output. However, as shown in the next tables, the picture becomes clearer when the growth rate of value added is decomposed into the growth rate of the number of establishments and the growth in average size. Regarding the other regressors, the coefficient on changes in an industry’s share of total value added in manufacturing under liberalised and nonliberalised regimes is always negative and highly significant. This suggests that conditional convergence is at work in this sample. On the other hand, the interaction between the difference in private credit under liberalised and nonliberalised regimes and external financial dependence is never significant. Table 2 turns to the regressions using the difference in the growth rate of the number of establishments under liberalised and nonliberalised regimes as the dependent variable. The coefficient on the interaction between land inequality and external financial dependence is now negative in all six columns, but it is significant at the 5% level only when external dependence is measured using young firms. The negative and significant coefficient on the interacted term implies that industries with high dependence on external finance for young firms in countries with high inequality experience lower growth rates in the number of new firms after
46
M. Caselli / Economics Letters 121 (2013) 43–47 Table 2 Differences in growth in number of establishments before and after trade liberalisation. External dependence measured using All firms
1ShareNO
−4.95
Ineq × DepExtFin
(2.51) −2.12 (1.41)
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
Yes Yes 783 34 0.47
*
Young firms *
−4.93 (2.51) −2.34 (1.46) −0.71 (0.63) Yes Yes 783 34 0.47
*
−4.48 (2.37) −0.86** (0.37)
Yes Yes 754 34 0.49
Mature firms *
−4.48
(2.37) −0.90** (0.39) −0.14 (0.14) Yes Yes 754 34 0.49
−5.06* (2.59) −0.81 (1.57)
Yes Yes 755 34 0.47
−5.05* (2.59) −0.95 (1.57) −0.47 (0.36) Yes Yes 755 34 0.47
Notes: The dependent variable in all the regressions is the difference in the growth rate of the number of establishments in each industry under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. * Coefficients significantly different from zero at 10% level. ** Coefficients significantly different from zero at 5% level.
liberalisation compared to the period when the country is closed to trade. Therefore, high inequality puts a strain on the growth gains from trade by affecting people’s ability to open up a new business in a sector where credit constraints are more binding for new firms following a trade liberalisation. In terms of the size of the coefficient and considering the specification robust to the inclusion of the interaction between the difference in private credit under liberalised and nonliberalised regimes and external financial dependence (column 4), a one standard deviation decrease in a country’s land inequality (approximately 16 points) would increase the growth rate of the number of firms of the industry at the 75th percentile of the distribution of external financial dependence by approximately 7.5% points more than that of the industry at the 25th percentile in the period following trade liberalisation relative to the previous period. Table 3 presents the results of the regressions with the difference in growth in average size as the dependent variable. In this case, the R-squared shows that the variance in the regressors can explain only around 19% of the variance of the dependent variable, compared to about 21% in the first set of regressions and over 45% in the second set of regressions. Three out of the six coefficients on the interaction terms between land inequality and the dependence on external finance are positive, but they all turn out to be insignificant due to large standard errors. This suggests that inequality does not interact with credit constraints to affect the growth opportunities of a firm that already produces once a country opens up to trade. The results presented so far may be due to other channels not related to inequality. Thus, in order to confirm the importance of inequality, this section presents a set of robustness checks that indeed points at inequality to explain growth differences following trade liberalisation. Given the results above, the focus of the following discussion is on the regressions based on the difference of the growth of the number of firms under liberalised and nonliberalised regimes as the dependent variables and external financing measured using only young firms.6 Following the discussion in Rajan and Zingales (1998), the magnitude and significance of the coefficient on land inequality in column 4 of Table 2 are checked against the inclusion of interaction terms between external financing and, separately, real GDP per capita to control for economic development and average years of
6 Additional tables of results are available upon request.
schooling to control for human capital.7 The results are not statistically different. Even after controlling for economic development and human capital, the coefficient on land inequality may still be capturing the effect of other variables, such as institutions or liberalisation occurring at different times for different industries within the same country, a potential shortcoming of using the Sachs–Warner Index as a measure of trade liberalisation. The main difficulty, however, is getting hold of reliable data to control for such variables. Therefore, additional robustness checks include region-industry fixed effects to act as a catch-all vector for all these potentially heterogeneous effects. The use of regional dummies to control for institutions is not new and, indeed, Acemoglu et al. (2001) and Frankema (2006) show that institutions vary more widely across regions than within them. It is also arguable that countries in the same region have usually adopted similar trade policies in terms of which industries to protect (see Goldberg and Pavcnik (2007)). No significant change in the coefficient on the interacted term on land inequality in the regression for the growth in the number of establishments in differences can be detected and, thus, the coefficient is still negative and statistically significant at the 5% level. A further set of regressions include fixed effects interacting industry and the decade in which the country liberalised. The rationale behind the inclusion of this variable is that certain industries might have been favoured from being exposed to international markets at a particular time and, thus, the timing of liberalisation might be correlated with land inequality. While Caselli (2012) shows that there seems to be no such correlation between land inequality and the timing of liberalisation in a similar sample of countries, it is still important to control for this potential issue. The results confirm that this is not a potential source for biased coefficients since none of the coefficients change significantly. The final robustness check looks at the role of credit constraints in more detail. A potential explanation for the findings above is that trade liberalisation may cause structural change and, thus, some firms to close as new ones open. People’s ability to open a new firm depends on an industry’s entry costs – measured by the external financing needs for new firms – that are unlikely to be paid from cash flows and need instead to be financed either from entrepreneurs’ own wealth or from borrowing. This suggests that in countries with lower financial development, i.e., where borrowing to finance new activities is constrained, entrepreneurs’
7 The difference between the two papers is that in the current setting these two variables are measured in differences between the initial value of the period when the country is open and the initial value of the period when it is closed.
M. Caselli / Economics Letters 121 (2013) 43–47
47
Table 3 Differences in growth in average size before and after trade liberalisation. External dependence measured using All firms
Young firms
1RelAv Size
−6.41
−6.56
Ineq × DepExtFin
(2.01) −0.23 (0.64)
(2.09) −0.39 (0.64) −0.53 (0.42) Yes Yes 783 34 0.19
***
1FinDev × DepExtFin Country fixed effects Industry fixed effects Number of observations Number of countries R2
Yes Yes 783 34 0.19
***
***
−8.07 (2.63) 0.02 (0.35)
Yes Yes 754 34 0.20
Mature firms ***
−8.08
(2.65) −0.02 (0.36) −0.13 (0.30) Yes Yes 754 34 0.20
−6.40*** (2.00) 0.33 (0.26)
Yes Yes 755 34 0.19
−6.47*** (2.01) 0.23 (0.26) −0.33** (0.14) Yes Yes 755 34 0.19
Notes: The dependent variable in all the regressions is the difference in the growth rate in average size in each industry (measured as value added divided by the number of firms) under liberalised and nonliberalised regimes. Standard errors clustered at the country level are shown in parentheses. ** Coefficients significantly different from zero at 5% level. *** Coefficients significantly different from zero at 1% level.
own wealth, and thus wealth inequality, should have a greater role in affecting the growth gains from trade. In order to test this hypothesis in more detail, the regressions in column 4 of Tables 1 and 2 are run with the current sample of countries split between those with a level of financial development at the beginning of the whole period below the median and those above the median. When the growth rate of value added in differences is used as the dependent variable, the results show that the coefficient on the interaction term between land inequality and dependence on external finance is not statistically different between the two samples. On the other hand, when the dependent variable is the difference in the growth of the number of firms, the coefficient on the interaction term with land inequality is negative and significant at 5% level only for those countries with a level of financial development below the median. This confirms that there is some evidence that land inequality affects the growth gains from trade through the credit constraints channel when domestic financial institutions are not well developed. 5. Conclusion This paper tests the hypothesis that credit constraints are the main mechanism through which wealth inequality, here proxied by land inequality, affects negatively the growth gains from trade liberalisation in developing countries. This question is important because empirical evidence shows that not all countries grow faster after opening to trade, as predicted by standard models and hoped by trade liberalisers. Therefore, understanding which factors can limit a country’s growth potential after trade liberalisation can help policymakers design policies that augment the benefits derived from trade liberalisation. The empirical analysis finds a negative but insignificant coefficient when the difference in growth of value added under liberalised and nonliberalised regimes is regressed on the interaction term between inequality and three different measures of external financial dependence, but a negative and significant coefficient is found on the same regressor when growth of the number of establishments in differences is used as the dependent variable and land inequality is interacted with dependence on external finance measured using only young US companies in each industry. This evidence suggests that in developing countries inequality has a long-term effect on the growth gains from trade liberalisation by affecting agents’ decision to start up a new business and that availability of credit to a wider range of firms, i.e., financial widening, can help alleviate this negative effect on growth.
Acknowledgements I thank participants at the ETSG, seminar participants at UNSW and Monash University and an anonymous referee for helpful comments. I gratefully acknowledge that this research was financially supported by an Australian Research Council grant to Professor Alan Woodland. All errors are mine. References Acemoglu, D., Johnson, S., Robinson, J.A., 2001. The colonial origins of comparative development: an empirical investigation. American Economic Review 91, 1369–1401. Barro, R.J., Lee, J.-W., 2001. International data on educational attainment: updates and implications. Oxford Economic Papers 53, 541–563. Beck, T., Demirgüç-Kunt, A., Levine, R., 2000. A new database on financial development and structure. World Bank Economic Review 14, 597–605. Braun, M., 2003. Financial contractibility and assets’ hardness: industrial composition and growth. Harvard University. Braun, M., Raddatz, C., 2007. Trade liberalization, capital account liberalization and the real effects of financial development. Journal of International Money and Finance 26, 730–761. Caselli, M., 2012. Does wealth inequality reduce the gains from trade? Review of World Economics 148, 333–356. Dollar, D., Kraay, A., 2004. Trade, growth, and poverty. Economic Journal 114, F22–F49. Frankema, E., 2006. The colonial origins of inequality: exploring the causes and consequences of land distribution. In: Research Memorandum GD-81. University of Groningen, Groningen Growth and Development Centre. Goldberg, P.K., Pavcnik, N., 2007. Distributional effects of globalization in developing countries. Journal of Economic Literature 45, 39–82. Gustafsson, P., Segerstrom, P., 2010a. North–south trade with increasing product variety. Journal of Development Economics 92, 97–106. Gustafsson, P., Segerstrom, P., 2010b. Trade liberalization and productivity growth. Review of International Economics 18, 207–228. Heston, A., Summers, R., Aten, B., 2006. Penn World Table Version 6.2. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. Rajan, R.G., Subramanian, A., 2008. Aid and growth: what does the cross-country evidence really show? The Review of Economics and Statistics 90, 643–665. Rajan, R.G., Zingales, L., 1998. Financial dependence and growth. American Economic Review 88, 559–586. Rodríguez, F., Rodrik, D., 2001. Trade policy and economic growth: a skeptic’s guide to the cross-national evidence. In: Bernanke, B.S., Rogoff, K. (Eds.), NBER Macroeconomics Annual 2000. In: NBER Chapters, vol. 15. National Bureau of Economic Research, Inc., pp. 261–338. Rodrik, D., Subramanian, A., Trebbi, F., 2004. Institutions rule: the primacy of institutions over geography and integration in economic development. Journal of Economic Growth 9, 131–165. Sachs, J., Warner, A., 1995. Economic reform and the progress of global integration. Working Papers 1733. Harvard Institute of Economic Research. Thomas, V., Wang, Y., Fan, X., 2002. A new dataset on inequality in education: gini and theil indices of schooling for 140 countries, 1960–2000, the World Bank. UNIDO, 2012. Industrial statistics database. 3-digit ISIC code, Rev. 2. Wacziarg, R., Welch, K.H., 2008. Trade liberalization and growth: new evidence. World Bank Economic Review 22, 187–231.