Renders and Gaeremynck (2006 SSRN working paper 888770), in the study published as Renders, Gaeremynck, and Sercu (2010, Corporate Governance: An International Review), reported that the relation between corporate-governance ratings and firm performance for European FTSE Eurotop 300 firms (1999–2003) becomes positive and significant only after simultaneously correcting for sample-selection bias and endogeneity, using a Heckman-type selection correction combined with instrumented governance ratings. Their own analysis leaves the causal question open: the instruments (a one-year-lagged firm rating and a country governance score) are only semi-endogenous, and the authors flag generalization beyond Europe, the joint determination of governance, ownership, and capital structure, and the restriction of their analysis to the contemporaneous relation — governance may also have long-term effects on performance — as unresolved. Later work with stronger identification reaches contradictory conclusions — dynamic panel generalized-method-of-moments (GMM) studies find no causal relation (US board structure; UK code compliance), a European study of commercial ratings finds no consistent relation, and quasi-experimental studies find governance value only in specific settings and channels — while a 2025 methodological review documents that credible exogeneity of instruments remains unachieved in this literature. The open problem is to establish, with identification that withstands the documented critiques, whether better corporate governance causally improves firm performance, jointly addressing selection into rated samples, endogeneity of governance choice, and time-varying unobserved heterogeneity, and to state the institutional scope of the identified effect.

Background

Corporate governance is the set of institutions — board structure, shareholder rights, takeover defenses, disclosure — intended to align managers with shareholders when ownership and control are separated. Agency theory, the standard framework in corporate finance, predicts that better governance should improve firm performance, typically measured by valuation proxies such as Tobin's q (the market value of a firm relative to the replacement cost of its assets) or by operating performance. Empirical tests usually compress governance into a single index: for example, Gompers, Ishii, and Metrick (2003) built a shareholder-rights index for US firms and documented a strong positive association between stronger rights and firm value and returns, and commercial raters such as Deminor published annual firm-level scores (a 0–40 aggregate over more than 300 criteria) for the largest European listed firms.

Two econometric obstacles have shadowed this literature from the start. First, sample selection: governance scores exist only for rated firms, which are disproportionately large, and the process that determines whether a firm is rated is itself related to governance and performance. Estimating the relation only on rated firms therefore conditions on a non-random sample. The standard remedy is a Heckman-type correction: model observability with a Probit equation (a regression model for a binary outcome — here, whether a firm is rated) estimated on the broader universe of listed firms, then include the inverse Mills ratio (the conditional expectation of the selection equation's normal error given observability) as an additional regressor in the outcome equation. Second, endogeneity: firms choose their governance in response to performance and firm characteristics, so ordinary least squares (OLS) estimates of the governance coefficient are biased. The common remedy is instrumental-variables estimation via two-stage least squares (2SLS): replace the endogenous governance measure with the part predicted by instruments — variables correlated with governance but, by an exclusion restriction, uncorrelated with performance shocks.

Renders, Gaeremynck, and Sercu implemented both corrections jointly, adapting a three-equation framework from Mroz (1987): a Probit selection equation for whether a firm carries a Deminor rating (estimated on the broad Datastream universe of listed firms, tens of thousands of firm-year observations), a reduced-form first stage predicting the firm's rating from two instruments — the one-period-lagged firm rating and a country-level score of governance-code recommendations — and a structural performance equation containing the predicted rating and the inverse Mills ratio. On the pooled FTSE Eurotop 300 panel (1999–2003), uncorrected OLS yields small, insignificant, sometimes negative governance coefficients; the joint correction flips them positive and significant, with each single correction alone removing only part of the bias. The published version adds that, although companies in countries with stronger investor protection (legal rules protecting minority shareholders against expropriation by insiders) have higher governance ratings, the effect of governance on performance is smaller in those countries than in countries with only weak shareholder-protection regulations, and that improvements in ratings exhibit diminishing marginal performance benefits.

The authors themselves, and the wider literature, identify the weakness of this evidence: lagged ratings and country scores are choice variables or slow-moving institutional aggregates, so the instruments are at best semi-endogenous, and the exclusion restriction is contestable. This concern was sharpened by contemporaneous critiques. Core, Guay, and Rusticus (2006) showed that market-based evidence is inconsistent with a causal interpretation of the governance-performance association, and Larcker and Rusticus (2010) demonstrated that when the exclusion restriction fails, IV estimates can be more biased than the OLS estimates they replace. A further line of attack on the endogeneity problem is dynamic panel generalized method of moments (GMM), which uses lags of the firm's own variables as internal instruments and so addresses unobserved heterogeneity and simultaneity without external instruments (Wintoki, Linck, and Netter, 2012). Finally, quasi-experimental designs — governance shocks such as legal reforms, exploited with discontinuity or event-study methods — have been proposed as the most credible route (Giroud and Mueller, 2011; Black, Kim, Jang, and Park, 2015). The result is a literature in which the sign and existence of a causal governance-performance effect depend on the identification strategy used, which is precisely the unresolved situation that Renders, Gaeremynck, and Sercu left open.

Problem Statement

Establish, with credible causal identification, whether better corporate governance improves firm performance — and, in particular, whether the positive relation between corporate-governance ratings and performance that appears after jointly correcting for sample-selection bias and endogeneity in European rated-firm samples reflects causation rather than residual identification failure.

The problem requires: (i) an identification strategy for the causal effect of a defined governance measure (or an explicit governance dimension, such as board structure, shareholder rights, disclosure, or takeover defenses) on a defined performance measure, whose exogeneity assumptions withstand the critiques that have invalidated earlier instruments in this literature; (ii) a treatment of the selection problem that governance is observed only for rated or otherwise truncated samples, so that the estimate does not inherit rated-sample selection bias, or a demonstration that selection is immaterial in the setting studied; (iii) handling of the dynamic relation between governance and performance — the source models only the contemporaneous relation, whereas governance may have long-term effects on performance, and time-varying unobserved heterogeneity may confound estimates of it — and of the fact that governance, ownership concentration, and capital structure are jointly chosen; and (iv) an explicit statement of the institutional scope of the identified effect — country, legal and investor-protection environment, governance dimension, and mechanism — given the documented heterogeneity across regimes.

Both directions of conclusion are admissible: a confirmed positive causal effect with its mechanism and scope, a credibly identified null for a stated governance dimension or setting, or a demonstration that the positive joint-correction findings (including those of Renders, Gaeremynck, and Sercu) are artifacts of semi-endogenous instruments would each resolve the problem, provided the identification itself is credible. The completion boundary is a causal claim about governance and performance, together with its scope, that a specialist referee accepts as identified — not the use of any particular dataset, estimator, or number of institutional settings.

References

  1. Renders, Annelies; Gaeremynck, Ann (2006). Corporate Governance and Performance: Controlling for Sample Selection Bias and Endogeneity. SSRN working paper. DOI: 10.2139/ssrn.888770. https://doi.org/10.2139/ssrn.888770
  2. Renders, Annelies; Gaeremynck, Ann; Sercu, Piet (2010). Corporate-Governance Ratings and Company Performance: A Cross-European Study. Corporate Governance: An International Review 18(2):87–106. DOI: 10.1111/j.1467-8683.2010.00791.x. https://doi.org/10.1111/j.1467-8683.2010.00791.x
  3. Gompers, Paul; Ishii, Joy; Metrick, Andrew (2003). Corporate Governance and Equity Prices. The Quarterly Journal of Economics 118(1):107–156. DOI: 10.1162/00335530360535162. https://doi.org/10.1162/00335530360535162
  4. Core, John E.; Guay, Wayne R.; Rusticus, Tjomme O. (2006). Does Weak Governance Cause Weak Stock Returns? An Examination of Firm Operating Performance and Investors' Expectations. The Journal of Finance 61(2):655–687. DOI: 10.1111/j.1540-6261.2006.00851.x. https://doi.org/10.1111/j.1540-6261.2006.00851.x
  5. Larcker, David F.; Rusticus, Tjomme O. (2010). On the use of instrumental variables in accounting research. Journal of Accounting and Economics 49(3):186–205. DOI: 10.1016/j.jacceco.2009.11.004. https://doi.org/10.1016/j.jacceco.2009.11.004
  6. Wintoki, M. Babajide; Linck, James S.; Netter, Jeffry M. (2012). Endogeneity and the dynamics of internal corporate governance. Journal of Financial Economics 105(3):581–606. DOI: 10.1016/j.jfineco.2012.03.005. https://doi.org/10.1016/j.jfineco.2012.03.005
  7. Giroud, Xavier; Mueller, Holger M. (2011). Corporate Governance, Product Market Competition, and Equity Prices. The Journal of Finance 66(2):563–600. DOI: 10.1111/j.1540-6261.2010.01642.x. https://doi.org/10.1111/j.1540-6261.2010.01642.x
  8. Klapper, Leora; Love, Inessa (2004). Corporate governance, investor protection, and performance in emerging markets. Journal of Corporate Finance 10(5):703–728. DOI: 10.1016/s0929-1199(03)00046-4. https://doi.org/10.1016/s0929-1199(03)00046-4
  9. Black, Bernard S.; Kim, Woochan; Jang, Hasung; Park, Kyung-Suh (2015). How corporate governance affect firm value? Evidence on a self-dealing channel from a natural experiment in Korea. Journal of Banking & Finance 51:131–150. DOI: 10.1016/j.jbankfin.2014.08.020. https://doi.org/10.1016/j.jbankfin.2014.08.020
  10. Akbar, Saeed; Poletti-Hughes, Jannine; El-Faitouri, Ramadan; Shah, Syed Zulfiqar Ali (2016). More on the relationship between corporate governance and firm performance in the UK: Evidence from the application of generalized method of moments estimation. Research in International Business and Finance 38:417–429. DOI: 10.1016/j.ribaf.2016.03.009. https://doi.org/10.1016/j.ribaf.2016.03.009
  11. Núñez Izquierdo, Manuel E.; Garcia-Blandon, Josep (2017). Evaluating the link between commercial governance ratings and firm performance in a cross-European setting. Management Decision 55(10):2089–2110. DOI: 10.1108/MD-11-2016-0830. https://doi.org/10.1108/MD-11-2016-0830
  12. Khatib, Saleh F. A. (2025). An assessment of methods to deal with endogeneity in corporate governance and reporting research. Corporate Governance: The International Journal of Business in Society 25(3):606–630 (first published online 2024). DOI: 10.1108/cg-12-2023-0507. https://doi.org/10.1108/cg-12-2023-0507
  13. Bøhren, Øyvind; Ødegaard, Bernt Arne (2012). Governance and Performance Revisited. In: International Corporate Governance After Sarbanes-Oxley. Wiley, pp. 27–64. DOI: 10.1002/9781119201885.ch3. https://doi.org/10.1002/9781119201885.ch3
  14. Mroz, Thomas (1987). The Sensitivity of an Empirical Model of Married Women's Hours of Work to Economic and Statistical Assumptions. Econometrica 55(4):765–799. DOI: 10.2307/1911029. https://doi.org/10.2307/1911029