A panel analysis of corporate governance spillovers among the G7, BRICS, and GIIPS countries

Abstract This study uses the Worldwide Governance Indicators (WGI) developed by the World Bank to examine a cross-country corporate governance spillover effect among the G7, BRICS, and GIIPS countries during the period of 1996–2014. The panel unit root tests show stationary panel time series properties of the WGI among these country groups. However, the results of panel vector autoregression partially support the spillover effect, which the U.S. possesses a governance-influence over the GIIPS and BRICS countries, but not for the remaining G7 countries.


Introduction
A good corporate governance paradigm is considered as an effective means to minimize the agency problem and to improve transparency. This paper aims to examine whether or not good governance practices can extend across economically relevant countries or regions. Using the Worldwide Governance Indicators (WGI) developed by the World Bank, the results show that there are governance spillovers among the G7, BRICS, and GIIPS countries. Panel unit root tests confirm stationary panel time series properties in all groups. However, panel vector autoregresssion (VAR) tests do not fully support governance spillover effect, which the U.S. has a governance-influence over the GIIPS and BRICS countries, but not for the remaining G7 countries. Kaufmann et al. (1999b) document a positive relationship between governance and economic development outcomes in 150 countries. They use the Worldwide Governance Indicators (WGI), measuring six dimensions of governance. First, Voice and Accountability captures the degree to which country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media. Second, Political Stability and Absence of Violence captures the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means (e.g., politically motivated violence and terrorism). Third, Government Effectiveness measures the quality of public and civil services, degree of independence from political pressures, quality of policy formulation and implementation, and credibility of the government's commitment to such policies. Fourth, Regulatory Quality measures the ability of the government to formulate and implement sound policies and regulations promoting private sector development. Fifth, Rule of Law measures the extent to which agents have confidence in and abide by the rules of society, and the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. Lastly, Control of Corruption indicates the perceptions of the extent to which public power is exercised for private gain. These reflect perceptions of interested parties, residents of a country, entrepreneurs, foreign investors, and civil society at large regarding the quality of governance in a country. The aggregate indicators are based on several underlying variables, taken from a wide variety of existing data sources (See, Kaufmann et al. (1999aKaufmann et al. ( ), (1999b for the governance dimensions measured and aggregation methodology).

Governance indicators as a proxy for corporate governance practices
The WGI is not without a limitation. Critics have focused on problems of bias or lack of comparability of these indicators. For example, Thomas (2010) focuses on whether the indicators have a construction validity and whether they measure what they are supposed to measure. The argument is that the WGI is based on personal and untested notions of governance and that the WGI claims to measure governance; as yet no evidence has been offered that this is true. Langbein and Knack (2010) also argue that the WGI is tautological and is not measuring what claims to measure.

Corporate governance spillover
A transmission of corporate governance among firms can be explained by the model of career concerns (Cheng, 2011;Holmstrom, 1982;Meyer & Vickers, 1997) that managers of two firms compare each other's performance and decide how much to inflate earnings based on their performance differential. The competing managers, who are motivated by their own career concerns, are more likely to inflate their own earnings to boost up stock prices. 1 Based on the micro foundation of spillovers, the corporate governance practices can extend across borders due to, for instance, the economic integration, cross-border mergers and acquisitions and financial market integration. For the past decades, free trade agreements among countries in the same region and across regions have been widely established and implemented (e.g., the European Union (EU), ASEAN Economic Community (AEC) and The North American Free Trade Agreement (NAFTA)), resulting in an increasing level of economic integration among the groups of countries under the agreements. 2 In addition, firms in the countries under the free trade agreement have been integrated through the mergers and acquisitions (M&A). Corporate governance of the firms can then be spilled over between the two merged firms. 3 With a higher level of international trade and economic integration among countries, the cross-country M&As have enforced the notion of the corporate governance spillovers. In summary, corporate governance spillovers can spread between the two countries through the cross-border merger and acquisition activity.
Financial market integration among countries has also largely increased through the world financial market during the past decades. The financial integration allows cross-border capital flows and cross-listing. For example, Liao and Ferris (2015) examine the intra-industry spillover in international equity markets by explaining that foreign cross-listing firms have to comply with the SEC and exchange regulation, subsequently pushing to a higher level of corporate governance. All in all, there is strong evidence that the level of governance practices can extend across the borders from one country to the other.

Data
While other studies focus on corporate governance at the firm level based on several aspects, such as the sustainability index, individual governance index, Environmental, Social, and Governance of corporate (ESG), this study highlights on the country level by using the Worldwide Governance Indicators (WGI) of the World Bank. 4 Kaufmann et al. (2011) employ the unobserved components model (UCM) in order to create WGI by grouping 31 sources of survey data into six dimensions of governance, namely Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Government Effectiveness, Regulatory Quality, Rule of Law,and Control of Corruption. 5 The WGI data are based on the survey of participant perceptions regarding governance. The WGI is constructed from many sources of expert survey, for example, the Global Competitiveness survey, BERI survey, and Economic Freedom Index Poll. The sample in this study excludes missing data and countries that do not have stock exchanges. The remaining sample covers 182 countries over the period of 1996-2014. To provide an economic justification on the governance spillovers of the economic/financial integration, this study uses three country groups namely, the G7, BRICS, and GIIPS countries as follows.
The annual stock market index returns are calculated from its closing stock index prices, which are collected from the Refinitiv Thomson Reuters.

Methodology
Starting from a panel unit root test, we hypothesize that a cross-country corporate governance spillover exists among countries within the same economic groups. If it is the case, the panel time series using the annual WGI of these country groups (G7, BRICS and GIIPS) should be stationary.

G7 BRICS GIIPS
The tests include the LLC test (Levin et al., 2002), IPS test (Im et al., 2003) and MW test (Maddala & Wu, 1999), which assume that cross-sectional (countries) WGIs in the panel are all independent or there is no relationship among the cross-sectional (countries) WGIs.
The LLC test computes the test statistic by averaging single time-series Augmented Dickey Fuller (ADF) t-tests of all cross-sectional units (countries) assuming homogenous across cross-sectional units (countries). The null hypothesis is that each individual country corporate governance time series contains a unit root against the alternative that each country corporate governance time series is stationary. The testing model is The test procedures can be divided into three steps. The first step begins by performing separate augmented Dickey-Fuller (ADF; equation (1)) for each i cross-sectional country currency where lag order p i can be varied across i. For given time T, optimal lag p i can be determined. The two regression models are then estimated (i) using ΔCG it as dependent variable and ΔCG itÀ L (for all L = 1, . . ., p i ) and d imt as independent variables to obtain residual ê it and (ii) using CG itÀ 1 as dependent variable and ΔCG itÀ L (for all L = 1, . . ., p i ) and d imt as independent variables to get residual v itÀ 1 . Then, standardized values of the two residuals should be computed as ẽ it ¼ê it σ εi and The second step is to compute the ratio of long-run to short-run standard deviations. The longrun variance can be computed as: where � K is optimal truncated lag and w� KL Then, ratio of long-run standard deviation to innovation standard deviation is computed as ŝ i ¼σ R i �σ ε i and average standard deviation can also be computed as The last step is to compute the panel unit root test statistics by estimating pooled regression based on N e T observations of where: Then, the panel unit root t-test for H 0 : ρ ¼ 0 can be computed: Finally, to obtain asymptotic property, the adjusted t-statistic can be computed: There are two equations of the spillover tests in XTVAR. The first equation has the WGIs of the remaining G7 countries as a dependent variable and a lagged dependent variable and the U.S.'s WGI as the independent variables. The second equation uses the U.S.'s WGI as a dependent variable and the same set of independent variables. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% level, respectively. There are six dimensions of corporate governance spillovers. There are the coefficients of lag term (θ) of equation where: μ � mT and σ � mT are the mean and standard deviation adjustments obtained from LLC computations. As a result, the t � ρ is asymptotically distribution as Nð0; 1Þ.
Note that the LLC test has also been claimed that its limitations are caused by cross-sectional independent assumption and test only no unit-root of all cross-sectional units. Table 1 shows descriptive statistics of all six aspects on country governance (i.e., Voice and Accountability, Political Stability, Government Effectiveness, Regulatory Quality, Rule of Law, and Control of Corruption) of all countries. Mean values of the WGIs for the G7 and GIIPS countries are largely similar in all six aspects, while the mean figures of the WGIs for the BRICS countries are not in a uniform pattern because of differences in geography, economic background, and culture. The G7 countries have on average the WGIs higher than those of the two groups countries. This potentially shows that more advanced economies possess larger levels of corporate governance. The panel unit root tests of each aspect of WGIs are performed into three groups, namely G7, BRICS and GIIPS as shown in Table 2. The results demonstrate that most aspects are statistically significant, which is consistent to the presumptions of the models. 6

Empirical results
To further examine the cross-country spillover effect, this study employs the panel VARs to determine the interdependence and dynamic relationship of the WGIof the examining G7, BRICS and GIIPS countries and the WGI of the foreign (US) country. Table 3 shows the estimated XTVAR results that the dependence of the WGIof the GIIPS and BRICS countries on the Government Effectiveness, Regulatory Quality and Control of Corruption aspects with those of the U.S. However, the results between the U.S.'s WGI and WGI of the remaining G7 countries indicate insignificant interdependent and dynamic relationship between the two variables. These results can be implied that U.S. has a governance-influence over the GIIPS and BRICS countries but has no impact on the remaining G7 countries. According to panel VARs, the results partially confirm that the economic and financial integration leads to governance spillovers.

Conclusion
This study reveals an empirical evidence of a cross-country corporate governance spillover effect. It uses a panel data technique to determine corporate governance spillover among countries within the same economic/regional groups. Based on the panel unit root tests, the results show a stationary panel time series of the WGIs among the G7, BRICS, and GIIPS countries, supporting the hypothesis that there exists a corporate governance spillover among the economically integrated countries. However, the results of the panel VARs partially support the governance spillover hypothesis. Specifically, the VARs suggest that the U.S. has a governance-influence over the GIIPS and BRICS countries, but has no impact on the remaining G7 countries. The limitation of this paper is that we compare other countries with the U.S., which the results might not be similar when comparing to other countries. This calls for further investigation.

Notes
1. The inflating earnings minimizes bad perceptions on a firm's performance (Holmstrom, 1982;Meyer & Vickers, 1997;Stein, 1989). Graham et al. (2005) document that career concerns and external labor market reputations are the first concern for managers. Gibbons and Murphy (1990) and Jenter and Kanaan (2015) reveal the significant relationship between relative performance evaluation in stock price performance and observably poor labor market outcomes such as being fired. 2. Petri et al. (2012) find that the ASEAN markets, especially the ASEAN-5, have been integrated and converged in terms of economic growth (both productivity and unemployment). Trade among ASEAN has increased from about 18% in 1985 to more than 30% in 2015. 3. Mostly, an acquiring firm with a high level of corporate governance transfers its governance practice to the target firm with a lower level of corporate governance. Bris et al. (2008) show that the spillover of corporate governance is implemented via transfers of accounting standards and shareholder protection, subsequently improving the Tobin's Q. Based on the law hypothesis, a positive spillover states that a spillover of corporate governance caused by M&A spreads from a high level bidder firm in order to improve a relative low level corporate governance of target firm (Goergen and Renne, Lim et al., 2008;Martynova & Renneboog, 2008). 4. The sources of WGI data are available upon request. 5. For the methodological construction of the WGI indicators, see Kaufmann et al. (2011). 6. The results of panel cointegration are in line with the those of panel unit root. The results are available upon request.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
The authors received no direct funding for this research.