J-Curve in Transition Economies: A Large Meta-analysis of the Determinants of Output Changes

Abstract

Immediately after the collapse of socialism, the countries of Central and Eastern Europe and the former Soviet Union fell into a serious output decline, after which they experienced a gradual recovery. Therefore, without exception, these countries followed a J-curved growth path. However, there were marked differences among them in the length and depth of the output fall and the speed of recovery. In this paper, we perform a comparative meta-analysis of the effect size and statistical significance of structural change, transformation policy, the legacy of socialism, inflation, and regional conflict in order to elucidate the mechanism that generated the J-shaped trajectory in transition economies. The meta-synthesis, which employs 3279 estimates drawn from 123 previous studies, revealed that while the growth-enhancing effects of structural change and transformation policy were small yet significant, inflation and regional conflict had a highly significant and strongly negative effect on output. In addition, the legacy of socialism might exacerbate the decline in production in the early stages of transition. The meta-regression analysis that simultaneously controls for various research conditions and the assessment of publication selection bias provides supporting evidence for the results obtained from the meta-synthesis.

This is a preview of subscription content, log in to check access.

Fig. 1

Source: Authors’ illustration. The data is derived from the EBRD website (http://www.ebrd.com)

Fig. 2

Source: Authors’ illustration. The data are derived from the EBRD website (http://www.ebrd.com)

Fig. 3

Source: Authors’ illustration

Fig. 4

Source: Authors’ illustration

Fig. 5

Source: Authors’ illustration

Notes

  1. 1.

    In fact, according to data published by the European Bank for Reconstruction and Development (EBRD, http://www.ebrd.com), output in the three CEE countries of Bosnia-Herzegovina, Montenegro, and Serbia and the two FSU countries of Moldova and Ukraine was, in 2013 and 2015, respectively, between 8 and 35% lower than that at the end of socialist period.

  2. 2.

    Babecký and Campos (2011) involved a meta-analysis of 515 estimation results reported in 46 studies, while Babecky and Havranek (2014) employ 537 estimation results from 60 studies. In addition to these two papers, Fidrmuc and Korhonen (2006), Hanousek et al. (2011), Kuusk and Paas (2013), Iwasaki and Tokunaga (2014, 2016), Iwasaki and Kočenda (2017), Iwasaki and Mizobata (2017), Iwasaki and Uegaki (2017), Tokunaga and Iwasaki (2017), and Iwasaki et al. (2018) present meta-studies of the transition literature.

  3. 3.

    One may feel it odd that even among the then-new member states of the European Union (EU), six Central European and Baltic countries with especially strong reputations for promoting reform, as well as three FSU countries, Uzbekistan, Turkmenistan, and Belarus, where the pace of democratization and economic reform has been particularly low, are all included in the same cluster. However, as Iwasaki (2004) pointed out, the macroeconomic performance of these FSU countries, the governments of which dealt with national crises caused by the breakup of the Soviet Union by exercising strong leadership over industry, was not much more unfavorable in comparison with that of the Central European and Baltic countries, especially during the early phase of transition. Putting aside the evaluation of the reform strategy based on a statist, paternal industrial strategy, these facts can be seen as having a big impact on the results of the cluster analysis.

  4. 4.

    In Section 3 of Iwasaki and Kumo (2016), we comprehensively review the relevant literatures that argue the macroeconomic effects of these five factors.

  5. 5.

    The final literature search was performed in December 2016.

  6. 6.

    For the details of these 123 studies, see the online appendix of the paper available on the journal’s website: http://www.palgrave.com/jp/journal/41294.

  7. 7.

    Independent variables employed in previous research at frequencies similar to the above five variables are domestic investment and fiscal expenditure, but the number with empirical findings estimated to be statistically significant is much lower than that for the above five factors. The next most frequently used are education level and foreign direct investment (FDI). The empirical results of the growth-promoting effect of education are similar to those of domestic investment and fiscal expenditure, as mentioned in Introduction. FDI is regarded as a promising factor behind growth in transition economies. However, Iwasaki and Tokunaga (2014) have already examined the impact of FDI on macroeconomic growth in CEE and FSU countries through a meta-analysis. Other factors that might significantly affect macroeconomic performance in transition economies include a great shift in the labor force participation rate, large-scale international migration, and dramatic changes in the world oil price (Bah and Brada 2014; Kuboniwa 2014; Bilan and Strielkowski 2016). However, empirical evidence of the growth effects of these factors is extremely limited, and most available estimates do not cover the first decades of the transition period. Therefore, we could not consider these factors, in addition to the factors mentioned above, in the meta-analysis in this paper. We thank the referee for his/her comment on this point.

  8. 8.

    The representative proxy for the socialist legacy employed in the growth models estimated in the literature subject to the mate-analysis is the number of years under socialism. The effect of inflation on output was examined using the consumer price index (CPI) in almost all of these primary works. With respect to the impact of regional conflict on growth, the vast majority of studies have employed a dummy variable whereby 1 denotes the year and country in which a conflict occurred. In regard to the proxy for transformation policy and structural change, we will mention this in the subsection of Meta-analysis of Structural Change and Transformation Policy in the next section. For a more detailed argument on the empirical strategy adopted in the literature, see Section 3 of Iwasaki and Kumo (2016).

  9. 9.

    For more details on the method of evaluating the research quality, see Appendix of this paper.

  10. 10.

    For a more detailed explanation of the meta-synthesis method and the fail-safe N, see Appendix B of Iwasaki and Kumo (2016).

  11. 11.

    Cohen (1988), who is frequently cited for assessing correlation coefficients, defines a coefficient of 0.3 as the threshold between a “small effect” and a “medium effect” and a coefficient of 0.5 as the threshold between a “medium effect” and a “large effect.” It is argued, however, that Cohen’s guidelines for zero-order correlations are too restrictive when applied to economics. This prompted Doucouliagos (2011) to propose alternative criteria to those of Cohen (1988). According to his new general criteria, the lower thresholds for small, medium, and large effects are set at 0.070, 0.173, and 0.327, respectively.

  12. 12.

    Incidentally, when we performed a meta-synthesis limited to estimates for the 1990s, when almost all of the CEE and FSU countries were either in the midst of crisis or in which output had still failed to recover to the levels at the end of socialism, as the estimation period, the synthesized effect size of structural change using the random-effects model shrank to 0.012, thereby becoming statistically insignificant. On the other hand, those of the socialist legacy and inflation both increased dramatically, to − 0.206 and − 0.413, respectively. Meanwhile, the synthesized effect size of transformation policy and regional conflict changed only slightly, to 0.170 and 0.281, respectively. These results suggest that the time-lagged effect of structural change and the time-decay effect of the socialist legacy and inflation have not been adequately captured in the earlier research.

  13. 13.

    For more details on the MRA method, see Appendix B of Iwasaki and Kumo (2016). In addition to MRA using these orthodox estimators, some meta-analysts employ some sort of model averaging approach including frequentist model averaging and Bayesian model averaging to tackle with the issue regarding model uncertainty. For instance, see Ahtiainen and Vanhatalo (2012), Babecky and Havranek (2014) and Havranek and Sokolova (2016).

  14. 14.

    Following Havranek and Sokolova (2016), we also estimated Eq. (2) by the cluster-robust WLS estimator that employs the inverse of the number of estimates reported per study as an analytical weight and obtained the similar results to those in the WLS models in Tables 5, 7, and 8.

  15. 15.

    Studies that have paid particularly close attention to the relationship between reform speed and economic growth include Heybey and Murrell (1999) as well as Bernardes (2003), Staehr (2005) and Godoy and Stiglitz (2006). Most previous studies have employed temporal differences in the degree of reform as a proxy for reform speed. For instance, see de Macedo and Martins (2008) and Segura-Ubiergo et al. (2010).

  16. 16.

    Due to space constraints, we have left some estimates out, but as was the case with Table 5, meta-independent variables that capture various study conditions are simultaneously estimated.

  17. 17.

    In their meta-analysis, Babecky and Havranek (2014) focused on the difference in the short-run and long-run growth effect of structural reforms and pointed out that “on average, in the short-run reforms lead to significant costs in terms of output growth, while in the long run the effect of reforms on economic performance is positive and substantial” (p. 31). The results of our meta-analysis indicate that the difference between the reform level and the reform speed may be a more important aspect to empirically examine the effect of transformation policy on output, judging from the findings that the meta-independent variable of reform speed repeatedly shows a significant and negative coefficient, while those of length of estimation period and use of cross-section data are estimated to be insignificant (Iwasaki and Kumo 2016, Table 9). However, when most meta-independent variables of transformation policy variable type and reform speed are controlled for between-study heterogeneity using the multilevel mixed-effects RML or the random/fixed-effects panel estimator, the statistical significance of the regression coefficient drops by a large margin. This makes it likely that some caution should be exercised in the interpretation of estimation results. We are grateful to the referee for his/her comment regarding this point.

  18. 18.

    In this regard, Coricelli and Maurel (2011) give a key to untangling this puzzle, suggesting that the post-recession performance in transition economies strongly depends on the complementarities among different reform measures, to which most previous studies do not pay sufficient attention. We thank the referee for his/her insight on this point.

  19. 19.

    To estimate Eq. (3), we use either the cluster-robust random-effects estimator or the cluster-robust fixed-effect estimator according to the results of the Hausman test of the random-effects assumption. With regard to Eq. (4), which does not have an intercept term, we report the random-effects model estimated by the maximum likelihood method.

  20. 20.

    For more details on the methodology, see Appendix B of Iwasaki and Kumo (2016).

  21. 21.

    The method for assuming that the mean of the most precise 10% of estimates is the approximate value of the true effect is along the lines of Stanley (2005).

  22. 22.

    For robustness check, as done in Havranek and Sokolova (2016), we also estimated Eq. (3) by the IV method using the inverse of the square root of the number of observations as an instrument for 1/SE and obtained the similar coefficients to those in Table 10 with a slightly lower statistical significance.

  23. 23.

    We should note, however, that the overwhelming majority of previous studies assume causality between the five factors in question and the macroeconomic performance in CEE and FSU countries as conventional wisdom, and as a result, they often do not properly deal with the possible endogeneity problem. The distinction between the effects of hyper and moderate inflation on growth is also examined insufficiently. In addition, the impact of resource shift among industrial sectors on productivity is addressed only in a few studies. Addressing these shortcomings remains as a future agenda in this study field. The comment from the referee on this point is highly acknowledged.

References

  1. Ahtiainen, H., and J. Vanhatalo. 2012. The value of reducing eutrophication in European marine areas: A Bayesian meta-analysis. Ecological Economics 83: 1–10.

    Article  Google Scholar 

  2. Babecký, J., and N.F. Campos. 2011. Does reform work? An econometric survey of the reform–growth puzzle. Journal of Comparative Economics 39(2): 140–158.

    Article  Google Scholar 

  3. Babecky, J., and T. Havranek. 2014. Structural reforms and growth in transition: A meta-analysis. Economics of Transition 22(1): 13–42.

    Article  Google Scholar 

  4. Bah, E., and J.C. Brada. 2014. Labor markets in the transition economies: An overview. European Journal of Comparative Economics 11(1): 3–53.

    Google Scholar 

  5. Bernardes, L.G. 2003. Reference-dependent preferences and the speed of economic liberalization. Journal of Socio-Economics 32(5): 521–548.

    Article  Google Scholar 

  6. Bilan, Y., and W. Strielkowski. 2016. Migration in post-transition economies: Immigration surplus in Visegrad group countries. International Journal of Trade and Global Markets 9(2): 182–196.

    Article  Google Scholar 

  7. Brada, J.C., and A.E. King. 1992. Is there a J-curve for the economic transition from socialism to capitalism? Economics of Planning 25(1): 37–53.

    Google Scholar 

  8. Cohen, J. 1988. Statistical power analysis in the behavioral sciences, 2nd ed. Hillsdale: Lawrence Erlbaum Associates.

    Google Scholar 

  9. Coricelli, F., and M. Maurel. 2011. Growth and crisis in transition: A comparative perspective. Review of International Economics 19(1): 49–64.

    Article  Google Scholar 

  10. de Macedo, J.B., and J.O. Martins. 2008. Growth, reform indicators and policy complementarities. Economics of Transition 16(2): 141–164.

    Article  Google Scholar 

  11. Doucouliagos, H. 2011. How large is large? Preliminary and relative guidelines for interpreting partial correlations in economics, School Working Paper No. SWP 2011/5. Melbourne: School of Accounting, Economics and Finance, Faculty of Business and Law, Deakin University.

  12. Fidrmuc, J., and I. Korhonen. 2006. Meta-analysis of the business cycle correlation between the Euro area and the CEECs. Journal of Comparative Economics 34(3): 518–537.

    Article  Google Scholar 

  13. Godoy, S., and J.E. Stiglitz. 2006. Growth, initial conditions, law and speed of privatization in transition countries: 11 years later, Working Paper No. 11992. Cambridge: National Bureau of Economic Research.

  14. Hanousek, J., E. Kočenda, and M. Maurel. 2011. Direct and indirect effects of FDI in emerging European markets: A survey and meta-analysis. Economic Systems 35(3): 301–322.

    Article  Google Scholar 

  15. Havranek, T., and A. Sokolova. 2016. Do consumers really follow a rule of thumb? Three thousand estimates from 130 studies say “probably not”, Working Paper No. 15/2016. Prague: Institute of Economic Studies, Faculty of Social Sciences, Charles University in Prague.

  16. Heybey, B., and P. Murrell. 1999. The relationship between economic growth and the speed of liberalization during transition. Journal of Policy Reform 3(2): 121–137.

    Article  Google Scholar 

  17. Iwasaki, I. 2004. Evolution of the government–business relationship and economic performance in the former Soviet states: Order state, rescue state, punish state. Economics of Planning 36(3): 223–257.

    Article  Google Scholar 

  18. Iwasaki, I., and E. Kočenda. 2017. Are some owners better than others in Czech privatized firms? Even meta-analysis can’t make us perfectly sure. Economic Systems 41(4): 537–568.

    Article  Google Scholar 

  19. Iwasaki, I., and K. Kumo. 2016. Decline and growth in transition economies: A meta-analysis, Working Paper No. 2016-9. Tokyo: Center for Economic Institutions, Institute of Economic Research of Hitotsubashi University.

  20. Iwasaki, I., and S. Mizobata. 2017. Post-privatization ownership and firm performance: A large meta-analysis of the transition literature. Annals of Public and Cooperative Economics. https://doi.org/10.1111/apce.12180.

    Article  Google Scholar 

  21. Iwasaki, I., and T. Suzuki. 2006. Radicalism versus gradualism: An analytical survey of the transition strategy debate. Journal of Economic Surveys 30(4): 807–834.

    Article  Google Scholar 

  22. Iwasaki, I., and M. Tokunaga. 2014. Macroeconomic impacts of FDI in transition economies: A meta-analysis. World Development 61: 53–69.

    Article  Google Scholar 

  23. Iwasaki, I., and M. Tokunaga. 2016. Technology transfer and spillovers from FDI in transition economies: A meta-analysis. Journal of Comparative Economics 44(4): 1086–1114.

    Article  Google Scholar 

  24. Iwasaki, I., and A. Uegaki. 2017. Central bank independence and inflation in transition economies: A comparative meta-analysis with developed and developing economies. Eastern European Economics 55(3): 197–235.

    Article  Google Scholar 

  25. Iwasaki, I., S. Mizobata, and A. Muravyev. 2018. Ownership dynamics and firm performance in an emerging economy: A meta-analysis of the Russian literature. Post-communist Economies. https://doi.org/10.1080/14631377.2018.1442036

    Article  Google Scholar 

  26. Kornai, J. 1994. Transformational recession: The main causes. Journal of Comparative Economics 19(1): 39–63.

    Article  Google Scholar 

  27. Kuboniwa, M. 2014. A comparative analysis of the impact of oil prices on oil-rich emerging economies in the Pacific Rim. Journal of Comparative Economics 42(2): 328–339.

    Article  Google Scholar 

  28. Kuusk, A., and T. Paas. 2013. A meta-analysis-based approach for examining financial contagion with special emphasis on CEE Economies. Eastern European Economics 51(3): 71–90.

    Article  Google Scholar 

  29. Segura-Ubiergo, A., A. Simone, A. Gupta, and O. Cui. 2010. New evidence on fiscal adjustment and growth in transition economies. Comparative Economic Studies 52(1): 18–37.

    Article  Google Scholar 

  30. Staehr, K. 2005. Reforms and economic growth in transition economies: Complementarity, sequencing and speed. European Journal of Comparative Economics 2(2): 177–202.

    Google Scholar 

  31. Stanley, T.D. 2005. Beyond publication bias. Journal of Economic Surveys 19(3): 309–345.

    Article  Google Scholar 

  32. Stanley, T.D., and H. Doucouliagos. 2012. Meta-regression analysis in economics and business. London: Routledge.

    Google Scholar 

  33. Stanley, T.D., and S.B. Jarrell. 2005. Meta-regression analysis: A quantitative method of literature surveys. Journal of Economic Surveys 19(3): 299–308.

    Article  Google Scholar 

  34. Tokunaga, M., and I. Iwasaki. 2017. The determinants of foreign direct investment in transition economies: A meta-analysis. The World Economy 40(12): 2771–2831.

    Article  Google Scholar 

Download references

Acknowledgements

This research work was financially supported by grants-in-aid for scientific research from the Ministry of Education and Sciences in Japan (Nos. 23243032, 26245034), the Joint Usage and Research Center of the Institute of Economic Research, Kyoto University (FY2016), and Center for Economic Institutions of the Institute of Economic Research, Hitotsubashi University. We thank Tomáš Havránek, Robert J. Johnston, Martin Paldam, Tom D. Stanley, Manabu Suhara, Paul Wachtel (the editor), two anonymous referees of the journal as well as participants in the study meeting at the Institute of Economic Research, Hitotsubashi University, Kunitachi, May 25, 2016, and the MAER-NET Colloquium at Hendrix College, Conway, September 16, 2016, and the Kyoto International Conference “Frontier of Transition Economics” at the Campus Plaza Kyoto on February 23–25, 2017 for their helpful comments and suggestions. We also would like to thank Eriko Yoshida for her research assistance and Dawn Brandon for her editorial assistance. Last but not least, we also wish to express our deepest respect to the authors of the literature subject to the meta-analysis in this paper. The usual caveats apply.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Ichiro Iwasaki.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOCX 54 kb)

Appendix: Method for Evaluating the Quality Level of a Study

Appendix: Method for Evaluating the Quality Level of a Study

This appendix describes the evaluation method used to determine the quality level of the studies subjected to our meta-analysis.

For journal articles, we used the rankings of economics journals published as of November 1, 2012, by IDEAS—the largest bibliographical database dedicated to economics and available freely on the Internet (http://ideas.repec.org/)—as the most basic information source for our evaluation of quality level. IDEAS provides the world’s most comprehensive ranking of economics journals, and as of November 2012, 1173 academic journals were ranked.

We divided these 1173 journals into 10 clusters, using a cluster analysis based on overall evaluation scores, and assigned each of these journal clusters a score (weight) from 1 (the lowest journal cluster) to 10 (the highest).

For academic journals that are not ranked by IDEAS, we referred to the Thomson Reuters Impact Factor and other journal rankings and identified the same level of IDEAS ranking-listed journals that correspond to these non-listed journals; we have assigned each of them the same score as its counterparts.

Meanwhile, for academic books and book chapters, we have assigned a score of 1, in principle; however, if at least one of the following conditions is met, each of the relevant books or chapters uniformly received a score of 4, which is the median value of the scores assigned to the above-mentioned IDEAS ranking-listed economics journals: (1) The academic book or book chapter clearly states that it has gone through the peer review process; (2) its publisher is a leading academic publisher that has external evaluations carried out by experts; or (3) the research level of the study has been evaluated by the authors as being obviously high.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Iwasaki, I., Kumo, K. J-Curve in Transition Economies: A Large Meta-analysis of the Determinants of Output Changes. Comp Econ Stud 61, 149–191 (2019). https://doi.org/10.1057/s41294-018-0058-4

Download citation

Keywords

  • Output changes
  • Transition economies
  • Meta-analysis
  • Publication selection bias
  • Central and Eastern Europe
  • Former Soviet Union

JEL Classification

  • E31
  • O47
  • O57
  • P20
  • P21