ABSTRACT
This paper aims to perform a large-scale meta-analysis to examine the relationship between ownership concentration and firm performance in emerging economies of Central and Eastern Europe and the former Soviet Union. A meta-synthesis of 1517 estimates collected from 69 previous studies indicated the presence of a statistically significant and positive effect of ownership concentration on firm performance. The synthesized effect size, however, is only modest at best. A meta-regression analysis conducted to identify the factors underlying the small effect size revealed that differences in target industries, estimation periods, design of ownership variables, data sources, estimators, and choices of control variables could have had systematic and profound effects on the empirical results presented in previous studies. We have also noted that publication selection bias is strongly suspected in this research field, and that, due to the magnitude of this bias, existing studies cannot be expected to provide genuine evidence regarding the effect of ownership concentration on firm performance in European emerging economies. Further empirical studies are required to identify the true effect in this region.
Acknowledgments
We would like to thank Chris Doucouliagos, Tomáš Havránek, Evžen Kočenda, Xinxin Ma, Martin Paldam, Tom D. Stanley, two anonymous reviewers of the journal as well as the participants at the Second World Congress of Comparative Economics at the Higher School of Economics in St. Petersburg, June 15–17, 2017 and the MAER-Net 2017 Colloquium at Zeppelin University in Friedrichshafen, October 12-15, 2017 for their valuable comments and suggestions. We also thank Eriko Yoshida for her research assistance and Mai Shibata and Tammy Bicket for their editorial assistance. Finally, we wish to express our deepest respect to the authors of the literature subject to the meta-analysis in this paper. Needless to say, all remaining errors are solely our responsibility.
Notes
1. For details about the academic debate on the effect of ownership concentration, see Wang and Shailer (Citation2015). In addition, focusing on the management of Russian firms during financial crisis, Iwasaki (Citation2016) theoretically addressed the virtues and vices of large shareholding.
2. This includes studies on corporate governance in Chinese companies.
3. Hanousek, Kočenda, and Shamsur (Citation2015) examined corporate efficiency in old and new Europe and concluded that majority ownership does not ensure efficiency.
4. The question of how differences in the types of owners and privatization methods employed can affect firm performance in transition economies has been examined by Iwasaki and Mizobata (Citation2018) by using a meta-analysis approach similar to the one described herein.
5. A meta-analysis conducted by Heugens, van Essen, and van Oosterhout (Citation2009), on the other hand, included only studies focusing on companies in Asia.
6. These emerging markets include 14 counties in South America, Asia, the Middle East, and Africa and the four CEE/FSU countries (i.e., the Czech Republic, Hungary, Poland, and Russia).
7. Concerning the privatization methods in CEE and FSU, Iwasaki and Mizobata (Citation2018) illustrated country specificity and its efficiency. In contrast to CEE/FSU countries, China gradually changed its corporate structure, particularly the enactment of corporate law, in 1994, and has maintained state control in key industries and state restrictions over private ownership.
8. In the selection of literature, we did not perform a so-called “self-screening,” referring to the third-party evaluation of the publication media and the research content that may lead to a kind of publication selection bias. As described later, we have rather adopted the approach of testing the possible influence of differences in research quality on empirical results by meta-regression analysis that adopts a series of meta-independent variables designed to control for various aspects of precedent works.
9. For more details on the method of evaluating the quality level of the study, see the Appendix.
10. To estimate Eqs. (4) and (5), we use either the cluster-robust random-effects estimator or the cluster-robust fixed-effects estimator according to the results of the Hausman test of the random-effects assumption. We also report the results of the Breusch-Pagan test and F test for reference. With regard to Eq. (6), which does not have an intercept term, we report the random-effects model estimated by the maximum likelihood method.
11. Cohen (Citation1988), 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. Doucouliagos (Citation2011) argues, in this regard, that Cohen’s guidelines for zero-order correlations are too restrictive when applied to economics and proposes to use the 25th percentile, 50th percentile (median), and 75th percentile of a total of 22141 PCCs collected by himself as alternative criteria. According to his new guidelines, for general purposes, 0.070, 0.173, and 0.327 are considered to be the lower thresholds for small, medium, and large effects, respectively.
12. In fact, in the case when a squared term is simultaneously estimated with a single term of an ownership variable, the synthesized effect sizes of the single term and the squared term of the ownership variable by a random-effects model are −0.055 (z=-3.928, p=0.000) and 0.056 (z=3.794, p=0.000), respectively, whereas in the case when a squared term is not simultaneously estimated, the synthesized effect size of the ownership variable is 0.012 (z=8.223, p=0.000).
13. The analytical approach whereby the mean of the most precise 10% of estimates is regarded as the approximate value of the true effect was originally proposed by Stanley (Citation2005).