CEO Diversity, Political Influences, and CEO Turnover in Unstable Environments: The Romanian Case

This work expands the literature on a less studied topic, the Chief Executive Officer (CEO) turnover in post-communist economies, analyzed during an unstable and ambiguous economic and financial environment. For the period 2005–2010, the results indicate the political inference in CEO turnover decision for the Romanian listed companies. In this period, with great turmoil in the economy determined by the financial crisis of 2008, we also find that CEO gender helps to explain the probability of changing the CEO. Moreover, this paper empirically tests if the financial and corporate governance determinants that are validated in the existing literature work for the Romanian listed companies. We reinforce that CEO turnover decision is negatively related to accounting-based performance. We find evidence of the “voting with their feet” behavior of institutional investors, and of the lack of Board of Directors monitoring. The CEO–Chairman duality and the controlling power of the largest shareholder act as entrenchment mechanisms.


Introduction
A substantial literature base devoted to the causes and consequences of CEO succession has been developed in the last decades. Due to the key economic role played by top-executive managers, CEO turnover has become a subject of widespread attention. The decision of managerial succession stands at the crossroads of corporate finance, management, corporate governance, psychology, and sociology (Campbell et al. 2011). It can be considered a strategic decision, which helps to preserve the shareholders'/stakeholders' interest whenever major objectives of the firm are not achieved, or whenever resources of the company are not used according to a mutually agreed plan.
Although there is a large amount of literature on CEO turnover in developed countries (e.g., Parrino, 1997, Kato and Long, 2006, Hazarika et al. 2012, not much has been written about the difficulties in implementing the letter and spirit of corporate governance rules in this area for a transition economy. To the best of our knowledge, this is the first study considering a wide range of financial, social, political, and corporate governance determinants on CEO turnover in Romania. Little research has been conducted for Central and Eastern European (CEE) post-communist countries (Muravyev, 2003 for Russia;Claessens and Djankov, 1999, Muravyev et al. 2010Eriksson, 2005 for Slovakia and Fidrmuc and Fidrmuc, 2007 for Czech Republic). The above-mentioned studies are not entirely applicable to the Romanian case for several reasons. For example, these studies focus mainly on the enterprise restructuring process and its consequences from a corporate governance perspective. Moreover, even if Eastern and Central European countries have some economic and recent political similarities, important differences in terms of cultural and pre-communist political backgrounds can be found (Filip and Raffournier, 2010). In the last decades, these countries have become, to a certain extent, different in terms of the degree of economic 1 and stock market development. 2 This paper contributes to CEO turnover research by taking into account three new variables, namely, the foreign origin of the CEO, the CEO's gender, and also the role of government ownership in connection with political changes. We find a positive correlation between the CEO foreign origin and the likelihood of CEO turnover. CEO gender helps explain both forced and voluntary CEO turnover: women chief executives are replaced more often than men. With regards to the specific corporate governance mechanisms in companies with governmental participation, we find that government ownership interaction with political changes is positively correlated with CEO turnover. The financial crisis also has a role in explaining the CEO turnover likelihood during the analyzed period, but only for the period of sharp downturn (the year 2009).
In addition, we analyzed the influence of "classical" corporate governance characteristics on CEO turnover in Romania. We partially validated the hypotheses usually proposed by the literature. Whenever it was necessary, we adapted them to better fit to the characteristics of the Romanian financial system and regulations.
The conclusions from this paper can be interesting for researchers, analysts, practitioners, and supervision authorities, also from the perspective of the analyzed period. This paper concentrates on a period of economic turmoil, namely, the period 2005-2010. Our choice has scientific but also practical reasons. From a scientific point of view, this paper covers the Romanian listed companies, the only relevant sample in terms of corporate governance implementation and data availability. Starting from 2011, the listing criteria and, consequently, the sample of listed companies suffered numerous changes, thus it is difficult to insure the homogeneity of the database.
In terms of practical interest for the study, our analyzed period, characterized by different researchers and analysts as turbulent in terms of economic and financial evolution, requires specific interest for corporate decisions. For example, Caruso et al. (2019) performed a model-based counterfactual exercise by estimating the model for the period 1983-2007 (pre-crisis sample) and by computing forecasts for 2008-2013 based on the pre-crisis parameters. They confirmed that households' and financial corporations' debts and house prices are weakly associated with the economic cycle in the pre-crisis sample. Additionally, an abnormal deep downfall in private investment and an increase in households' savings beyond historical regularities were registered. Finally, the major changes in the fiscal deficit-GDP and debt-GDP ratios in 2008-2009 were exceptional, similar to the fiscal consolidation that followed. Furthermore, international, but also national analysts, agree in declaring that during the year 2019, there was-and in the following years, there will be-a clear deceleration of the macroeconomic indicators. On October 1 st 2019, when Ms. Kristalina Georgieva was appointed Managing Director and Chairman of the Executive Board of the International Monetary Fund, her discourse was focused on Europe being on the verge of a new financial crisis. One month later, the same institution released a warning for European countries to put in place emergency plans, ready to be implemented, in the eventuality of what 1 The Gross Domestic Product (GDP) per capita (constant 2010 US$) varied between 5735 US$ for Serbia and 23437 US$ for Slovenia in the year 2010, in a comparision which included nine CEE countries (World Bank statistics). Romania had, in 2010, a GDP per capita of 8210 US$. Using the same CEE countries and the same database, Romania had, in 2008, the highest GDP growth rate (11.14%), but a decrease of -4.74% in 2009 (however, better than the case of the Czech Republic (-5.34%), Hungary (-6.46%), Croatia (-7.19%), and Slovenia (-8.63%)). In 2010, Romania had the slowest recovery in terms of GDP growth rate (-3.32%) out of all nine of the analyzed CEE countries. 2 Romania had, in 2005, almost the least developed capital market out of all nine of the analyzed CEE countries, with a market capitalization of listed domestic companies (% of GDP) of 16.11%, whereas Poland, Hungary, and Croatia had market capitalization around 30% from GDP. Romania had an even weaker position in 2010 (8.54%), after the fall from 2008, and the gap with countries such as Poland (39.79%) and Croatia (42.76%) increased. At the same time, it can be noticed that Romania was not among the most affected countries by the crisis from the analyzed CEE countries (a decrease of market capitalization as % from GDP around 59% in 2008), while the CEE countries with the most developed capital markets had the highest drop rates (65% and above) (using the World Bank indicators). seems to be an imminent crisis. Also, some European countries entered into technical recession according to different financial institutions: Germany in September 2019 according to the Central Bank, and Italy according to the European Central Bank. Romania is a special case, because for the previous few years, the government led a pro-cyclical policy that is likely to increase the risk of a potential recession. Several analysts warned of the unsustainability of public policies, but also of the economic growth, with the National Bank of Romania Governor Mugur Isarescu and his counselors, as well as Nouriel Roubini in a press analysis in November 2019, being among them. At the beginning of 2020, the worst-case scenario was confirmed with the spread of the Covid-19 epidemy, that caused a big fall in most important financial markets, as well as in the Romanian one. Thus, the conclusions of the present study can be very relevant to better understand and anticipate the future period (for 2020 and beyond), that Europe, in particular, will go through.
The emerging and/or frontier markets can be of great interest for both national and foreign investors, looking for new investment opportunities and for diversified portfolios based on shares issued by companies, other than those listed on mature capital markets. Therefore, we believe that our conclusions may be of great interest for scientists, but also for analysts and other finance practitioners.
The remainder of the study is organized as follows. Section 2 provides a short description of the Romanian environment in terms of corporate governance and investors' protection. In Section 3, we present previous research on the determinants of CEO turnover and we develop the tested hypotheses, focusing on CEO diversity and on political influences. Section 4 describes the research methodology and the data set. The empirical results and discussion are then presented in Section 5, while Section 6 concludes the study.

A Short Description of the Romanian Environment in Terms of Corporate Governance and Investors' Protection
In 2009, the Romanian Corporate Governance Code was adopted on a comply-or-explain basis. It proposed principles harmonized to those of the European Union. For the most liquid listed companies in Romania, Dobroţeanu et al. (2010) provided empirical evidence of a corporate governance disclosure gap. Indeed, Romanian listed companies still have a hesitant evolution regarding the implementation of the corporate governance standards.
Romania has a small capital market, even in comparison with different CEE countries, with 73 listed companies in 2010 (compared with 240 in Croatia, 390 in Bulgaria, and 570 in Poland in the same year) and with low trading volumes (1% for stock traded value, as % of GDP, compared with 15% for Poland and 11% for Czech Republic in 2010) (World Bank indicators) and relatively few disclosure requirements (including those for corporate governance). Moreover, the development of the Romanian capital market, including the corporate governance domain, took place in a peculiar political and economic context, different from other CEE countries. Due to the fact that the Romanian governments focused more on political issues than on important and necessary economic decisions for the healthy development of an ex-communist country, the economic gap of CEE countries such as the Czech Republic, Poland, and Hungary increased, and Romania joined the European Union only in 2007 (Filip and Raffournier, 2010).
Because we use market-based performance variables (alongside accounting-based performances), additional explanations are necessary for the Romanian capital market. Previous studies regarding the efficient market hypothesis (EMH) either reject it or provide evidence only for a relatively low degree of informational efficiency (Dragotă and Mitrică, 2004). The most recent studies have both interesting and intricate conclusions. Dragotă and Ţilică (2014) reject the weak form of informational efficiency, but reveal that for some assets and indexes, it is hard to reach systematic abnormal earnings. In this context, we expect some questionable results for marketbased performance in relation to CEO turnover.
Institutional investors' ownership is far less important in Romania than in developed capital markets. For the period 2005-2010, the average institutional investors' holdings for the companies listed on the Bucharest Stock Exchange (BSE) varied between 24% in 2005 and 30% in 2010, comparatively to almost 50% for the US in 2006 (Helwege et al. 2012). Also, our sample confirms the high level of ownership concentration, with an average percentage of the voting rights owned by the largest shareholder of around 47% and a median value of 52%. Thus, more than half of the listed companies have a major shareholder, controlling at least 52% of the voting rights.
Regarding the minority shareholders' rights protection index, Romania stands slightly above the average of the transition countries (41 compared with the average value of 35.4), but with a lower index value than the average of the developed countries included in the study (the average value for the developed countries is 46.9). Claessens and Yurtoglu (2013) also emphasize a weak creditors' protection in Romania and a low efficiency of debt enforcement. Weak investor protection obstructs the function of corporate governance systems. Consequently, CEO turnover will most likely be less sensitive to firm performance when the strength of investor protection decreases (Hu and Kim, 2019).

Literature Review and the Tested Hypotheses
We structured this section in two main parts. First, we review the existing literature on CEO turnover, firm performance (sub-section 3.1), corporate governance (sub-section 3.2), and also in connection with various financial indicators for leverage, dividend policy, and sector homogeneity (sub-section 3.3). There is a large body of empirical literature examining the impact of a wide range of economic, social, financial, and corporate governance determinants for CEO turnover. Then, in the second step, we propose three new determinants with corresponding hypotheses for foreign CEO, female CEO, and political influences on CEO turnover (sub-section 3.4).
Both market-based and accounting-based performance indicators are widely discussed in the literature and, in some cases, the conclusions are not similar. Market-based performance indicators are related to firm accounting performance, and also to investors' expectations regarding the perspectives of the firm, to their culture and optimism, and also to the degree of market efficiency. Weisbach (1988), Murphy and Zimmerman (1993), and Blackwell et al. (1994) claim that companies with low performance can use accounting manipulations to change investors' expectations. Altogether, Blackwell et al. (1994) affirm that accounting-based performance seems to be more important than market-based performance to explain CEO turnover, while Murphy and Zimmerman (1993) conclude that CEO replacement is inversely related to firm performance for both measures used. Bushman et al. (2010), Kaplan and Minton (2012), Eisfeldt and Kuhnen (2013), and Jenter and Kanaan (2015) underline also the importance of industry and market performance. Eisfeldt and Kuhnen (2013) emphasize that the weight-skills industry shocks determine CEO turnover, while idiosyncratic shocks do not strongly influence this phenomenon. Their model provides a broad explanation to better understand CEO turnovers for both performing and under-performing companies that experience voluntary CEO departures. Another result is that companies with low industry-adjusted return are more likely to have forced CEO turnovers, and also that industry performance is negatively correlated with the probability of forced turnovers.
We tested a wide range of accounting-and market-based performance indicators to obtain a better understanding of the main performance objectives considered by shareholders when they decide to change the CEO and, at the same time, to better fit this study to the Romanian economic environment. We used the return on equity (ROE) as an alternative to the stock return to overcome the effects of weak market informational efficiency (see, for example, the study of Dragotă and Țilică, 2014, regarding the market efficiency in Romania). We also analyzed if long-term return monitoring is more important than short-term returns, to decide forced CEO turnover using a 3year average return on assets (ROA), a 3-year average return on equity (ROE), and a 3-year average operating margin.

CEO Turnover and Corporate Governance
Country and firm characteristics in terms of corporate governance, ultimately resulting in a specific level of control of the managerial activity, are also expected to influence forced CEO turnover decisions. The existing literature proved that the CEO-Chairman duality is associated with a lower probability of forced turnovers (Kato and Long, 2006;Helwege et al. 2012;Hu and Leung, 2012;Hazarika et al. 2012). This correlation can be explained through a higher influence of the CEO on the Board of Directors and higher opportunities for the CEO to maintain informational asymmetries. On the other hand, the separation between CEO and Chairman leads to more efficient board control and improves firm performance.
CEO turnover occurrence is also related to the board's characteristics-such as size and independence. It is expected that a higher number of board members is associated with a higher level of managerial control, especially when independent directors prevail in the board. On the other hand, when the Board of Directors has numerous members, it could be more difficult to make changes in the company. Brunello et al. (2003) found a direct link between board size and the probability of forced CEO turnover. However, for Ukraine, a post-communist country, Muravyev et al. (2010) found that board size had no statistically or economically significant effects on the probability of CEO turnover.
To protect the interests of minority shareholders, the BSE Corporate Governance Code (2015) recommends that "the Board and its committees should have the appropriate balance of skills, experience, gender diversity, knowledge and independence to enable them to effectively perform their respective duties and responsibilities. It is recommended for the majority of non-executive members of the Board of Directors or Supervisory Board to be independent." Additionally, the same Corporate Governance Code (2015) recommends that "the majority of the members of the Board of Directors should be non-executive." Even though the Code does not establish an exact number or a proportion of independent directors, the Romanian listed companies have chosen to follow this recommendation. Helwege et al. (2012) provide weak evidence on board independence being associated with higher probability of CEO turnover. On the other hand, there are studies showing that companies with an independent board experience higher of forced CEO turnover to performance (Guo and Masulis, 2015).
Ownership concentration is important to determine a forced CEO turnover. Many empirical studies show that listed companies in Western Europe, but also in East Asia, the Middle East, or Latin America have large shareholders (see, among others, Claessens et al. 2000; Barca and Becht, 2001;Faccio and Lang, 2002). The presence of large shareholders is also typical for Romania and other post-communist Eastern European countries. For the companies listed on the first and second tiers on the BSE, the average first shareholder ownership is 47.42%. Concentrated ownership is considered by some authors more favorable for good corporate governance (Shleifer and Vishny, 1997), based on the agency problem arising from the separation of ownership and control. On the contrary, when using asymmetric information models, different authors prove that a higher ownership concentration affects firm performance or dividend policy (see Holmström and Tirole, 1983;Aghion and Tirole, 1997;Bolton and von Thadden, 1998, or studies for CEE countries such as Hanousek et al. 2007;Bena and Hanousek, 2008;Dragotă et al. 2013, among others). The high stake owned by the largest shareholder can also lead to a decrease in the interest of the companies in promoting firm-level corporate governance measures, as the controlling shareholder has additional and more effective instruments to monitor the firm (Bollaert and Dilé, 2009).
The literature predicts a lower probability of forced CEO turnover associated with a concentrated ownership (see, for example, Parrino 1997, Brunello et al. 2003, while Kato and Long (2006) recognize the usefulness of ownership concentration as a control variable in order to eliminate endogeneity problems.
Institutional ownership is also important to explain CEO turnover occurrence, especially due to the commonly known activism, doubled by the advanced financial skills of the institutional investors. It is expected that these "highly skilled and well-resourced professional shareholders would make informed use of their rights, promoting good corporate governance in companies in which they invest" (OECD 2011). Kaplan and Minton (2012), following the methodology applied by Cremers and Nair (2005), used the cumulative percentage of shares held in each firm by the large institutional shareholders (who own more than 5% ownership of the firm's outstanding shares). They found that the institutional investors' activism increases the probability of CEO turnover. Similar results can be found in the work of Brav et al. (2008) and Del Guercio et al. (2008). Parrino et al. (2003) found decreases of the share owned by institutional investors prior to forced turnovers, which can be considered as proof of the "voting with their feet" phenomenon.
The studies on agency models in connection with CEO ownership can be divided into at least two distinct groups. First, it is considered that CEO ownership can reduce the agency conflicts between managers and shareholders (especially with minority shareholders), because an executive with stocks (and stock options) is "in the same boat" as the rest of the owners. Contrary to this view, the second set of studies argues that an owner-manager can distort the main performance objectives of the company for private benefits, affecting the market prices and/or using his/her close ties with the controlling shareholders (Isakov and Weisskopf, 2014). This could certainly be the case of the Romanian listed companies, controlled by major shareholders. In this context, minority shareholders can still be protected if managers (or controlling shareholders) develop a reputation for treating outside shareholders well (Gomes, 2000;Maury and Pajuste, 2002).
The results of previous studies confirm the negative correlation between CEO ownership and the probability of forced turnover (see, among others, Brunello et al. 2003 andCampbell et al. 2011). Muravyev et al. (2010) also found evidence of managerial ownership supporting entrenchment. CEO ownership, like board ownership (Helwege et al. 2012), becomes a mechanism of alleviating shareholder-manager agency conflicts and is expected to be negatively correlated to CEO forced turnover.
Particularly when CEOs are significant shareholders, it is possible that they mainly represent the interests of the controlling shareholders, more so than those of the firms. When the CEO is appointed directly by the controlling shareholder, the criteria for measuring the CEO's performance may possibly be connected more to the effectiveness in protecting the interests of the controlling shareholder, rather than those of the firm. In our models, we took into account the simple presence of the CEO as a shareholder and then we considered his/her presence as a significant shareholder (owning 5% or more of the voting rights).

CEO Turnover, Capital Structure, Dividend Policy, and Sector Homogeneity
The agency conflict between CEO and shareholders can also be mitigated using capital structure and dividend policy. The creditors have the concern and appropriate skills to monitor the firms, and the shareholders benefit from their financial expertise (Ross, 1977). In the meantime, the CEO is less willing to undertake excessive risks and is more disciplined, which eventually leads to a lower probability of being forced to quit his/her position. On the other hand, capital structure is considered in the literature as a reflection of the financial risk undertaken by the company (Hazarika et al. 2012;Hu and Leung, 2012;Cronqvist et al. 2012). Hence, the leverage is expected to be positively associated with CEO turnover (forced or voluntary), especially in economies based on capital markets. Hazarika et al. (2012) for the US, and Hu and Leung (2012) for China provide weak evidence of a positive correlation between leverage and CEO turnover (forced and voluntary). On the other hand, Lin and Liu (2004) found a negative relation between leverage and forced CEO turnover in Taiwan.
In recent decades, financial theory has offered several explanations for dividend policy, based on signaling theories, or taking into consideration the agency conflicts between corporate insiders and outside shareholders. The dividend signaling hypothesis predicts that dividends convey information to shareholders/investors regarding the firm's future prospects, and only good-quality firms can use it (Bhattacharya, 1979). Even when managers are not sure whether their company has the ability to generate cash flows in the future, they may keep dividends constant, or even gradually increase payments in order to avoid sending a negative signal to the market (Zwiebel, 1996). Myers (2000) created a link between maintaining leading positions in the companies and the trustworthy perspective of dividend payments. Moreover, other studies argue that top-executive managers protect minority shareholders through dividend policy to create a reputation for treating outside shareholders well (Gomes, 2000). Parrino et al. (2003) provide evidence that firms experience dividend cuts prior to forced CEO turnover.
Based on these arguments, we expect a company that pays dividends to have a lower probability of forced CEO turnover. High dividend yields increase the demand for the company's shares, which is expected to mitigate the potential conflict between managers and shareholders. Parrino (1997) introduced sector homogeneity as a determinant of CEO turnover. He measured sector homogeneity through a volatility indicator of the company's stock return related to industry return. The intuition behind this assumption refers to the fact that in more homogeneous industries, the managerial skills required are similar, and managers can be easily replaced by others from the same industry. The low liquidity of the Romanian capital market did not allow us to follow the same methodology, but we indirectly studied the role of sector homogeneity on CEO turnover.

Particularities of the Romanian Economic Environment
Due to the fact that Romania still has a poor corporate governance culture and a relatively high perceived level of corruption (Claessens and Yurtoglu 2013), we analyzed the possibility of political influences on CEO turnover decisions.
State ownership, and sometimes the state as the controlling shareholder, can significantly influence the decision to change the CEO or other board members. If the CEOs of privately controlled firms are politically connected, those companies could have increasing access to credit, as well as to regulatory favors and government financial assistance. In these cases, the likelihood of CEO turnover is lower. Also, if the CEOs are politically connected, this fact causes weaker turnover -performance sensitivity (Cao et al. 2016) and the criteria for executive evaluations may be different. State-owned companies are more likely to apply government-oriented executive evaluations that focus on political performance rather than on their economic performance (Liu and Zhang 2018).
Taking into account several references of CEO appointments based on political criteria in different Romanian listed companies where the state is a significant or controlling shareholder, we expect a greater likelihood for CEO turnover after political parties (coalitions) change in parliamentary elections. If the political forces in power change, more frequent CEO turnover could follow if political pressure is exercised in the economic environment. Based on these considerations, our first tested hypothesis is: H1: Forced CEO turnover probability is higher, subsequent to political changes in companies with governmental ownership, especially when the government is an important shareholder.

Particularities Related to Post-Communist Economies
We also tested for potential inefficiencies in the labor market. A general practice of postcommunist countries is to reproduce the mechanisms from developed market economies, usually without adapting them to their socio-cultural values (see, in this context, the complex analysis of King and Szelenyi 2005). The problem can be even more stringent when it is cumulated to the propensity toward the conservatism of experienced people. In this case, a foreign manager can bring an advantage in terms of experience in a highly competitive environment and knowledge regarding the newly implemented systems.
On the other hand, the appointment of foreign CEOs is not without risk, because they are not accustomed to the local economic environment. Due to the scarcity of the labor market, a bigger effort may be made to appoint a foreign manager, and therefore we can expect a lower probability for a forced turnover. However, difficulties to adapt to the new environment may worsen the performance of the company and eventually lead to a forced CEO turnover. A correlation between a CEO's origin and CEO turnover points to a possible gap in the managerial skills between domestic and foreign CEOs. The second tested hypothesis is: H2: The foreign origin of a CEO is correlated with the likelihood of CEO turnover.
Some previous empirical evidence shows that women have behavioral and attitudinal differences compared to males (Beck et al. 2018). Female CEOs/directors tend to be less overconfident in their decision making than men, and are more likely to express their independent views than male directors. Furthermore, Srinidhi et al. (2020) show that female CEOs/directors from US companies manage to overcome the obstacles of being in the minority and not having the reputational advantages of long-serving male directors, by acting as norm change catalysts to achieve substantial governance changes. In this context, the boardroom dynamics could be changed in the presence of a female CEO/director and, consequently, the intensity of discussions around difficult (financial) decision problems could be different (Kim and Starks 2016;Chen et al. 2019). In an unstable economic and financial environment resulting from a deep financial crisis, such as the one started in 2008, removing a female CEO may (or may not) be more likely, depending on the goals that the company management proposes in such a turbulent period.
Moreover, Johnson and Powell (1994), Bajtelsmit and VanDerhei (1997), and also Francis et al. (2015), Zigraiova (2015), and Skała and Weill (2018) proved that women tend to have a stronger risk aversion than men. Especially in the macroeconomic context of the analyzed period, a positive correlation between CEO gender (i.e., female CEO) and the likelihood of turnover can be related to women's risk aversion, but a discrimination hypothesis cannot be rejected. However, the evidence on gender differences in risk aversion is mixed. Croson and Gneezy (2009) conclude that the difference in risk aversion between men and women becomes less significant for managers and professionals in finance. The hypothesis of gender discrimination inside the company can be disputable, because they have already been appointed as CEOs. However, the hypothesis of more subtle gender discrimination from outside the company cannot be rejected. An example of such a form of discrimination is evidenced in the work of Muravyev et al. (2009). It occurs through the more financial constraints imposed on companies managed by female CEOs and, respectively, a lower number of approved loans or higher credit costs relative to male managers. These differences are more important in weakly competitive financial systems and in emergent countries and tend to diminish in developed countries. Consequently, our third tested hypothesis is: H3: Forced CEO turnover probability is higher when the CEO is a woman.
These hypotheses were tested in an economic environment that can be considered as ambiguous and unstable, with a first sub-period of exceptional economic growth, followed by a severe financial crisis. The financial crisis, which started in 2008, brought a general deterioration of the economic conditions and we could thus expect an increase in the probability of forced CEO turnover (Kaplan and Minton 2012). At the same time, financial difficulties augment the occupational stress of CEOs and may lead to more frequent occurrences of CEO turnover. In times of crisis, companies are confronted with a great deal of ambiguity in the strategic planning process and in their business environment, and so should try to rely more on intuition and creativity in many parts of the management process, including in decisions regarding their top managers and whether to change them (e.g., CEO turnover). Based on the extensive literature in the field, Ogilvie (1998) showed that imagination and creativity become more suitable to manage unstructured, random, and contradictory data from the environment. In this context, we analyzed how the Romanian listed companies decide to manage a very turbulent period, with this study focusing on the criteria for CEO turnover (gender, nationality, and/or the level of political intrusion into a company's life).

Methodology and Data
To test our hypotheses, we used two different methodological approaches, which were also applied in the previous literature. First, similar to Brunello et al. (2003) and Hu and Leung (2012), among others, we applied a balanced panel data analysis. The (forced) CEO turnover dummy variable is equal to 1 if the firm had at least one (forced) CEO replacement during the year, and 0 otherwise. Second, we considered each CEO departure separately, and we applied an unbalanced panel data analysis (as a robustness check procedure), due to the fact that there are companies in our sample that registered more than one CEO turnover during the same year.
First, we identified and classified the reasons for CEO turnover using the methodology also applied by Warner et al. (1988) and Dherment-Ferere and Renneboog (2002), with necessary adjustments for the Romanian case, described below. Additionally, we considered the reasons for CEO turnover motivated by political manipulations and the battle for power as in Cannella and Shen (2001). We identified 12 reasons for CEO turnover (including both forced and voluntary CEO departures): retirement, health problems, political reasons, promotion, persona non-grata, change of ownership, common succession, pressure from shareholders, separation of the positions of CEO and Chairman of the Board, CEO resignation, other reasons, and no public reasons. A more detailed presentation for these reasons for our sample of Romanian listed companies can be found in the Appendix.
Second, we divided these turnovers into two classes: forced and voluntary. However, identification of forced departures can sometimes be very difficult when press releases are the main source of documentation, because they rarely describe them as such. To isolate most likely forced departures, firstly we followed Fee and Hadlock's (2004) methodology for classifying turnover as forced. If an article in the business press indicates that the CEO was banished, forced, or equivalent, then turnover was defined as forced. Secondly, similar to Conyon and Florou (2002), we considered the turnover as forced if the CEO was replaced for reasons such as political misunderstandings, conflicts with the Board of Directors, weak performance, personality differences, and scandals. Thirdly, we followed Dahya et al.'s (2002) methodology, considering dismissal and resignation as forced turnover. Thus, we considered the following reasons as forced turnovers: persona non-grata, change of ownership, pressure from shareholders, political reasons (except for the CEO of Oil Terminal who was elected in the Romanian Parliament, due to incompatibility reasons), CEO resignation (due to conflicts with the Board of Directors), as well as the cases in which no reason was given for CEO replacement, in case of weak pre-succession performances.
The sample contained 58 CEO turnover events that occurred in 36 companies. In the balanced panel data approach, they represent 53 company year CEO turnover observations out of the 426 observations collected between 2005 and 2010 for the whole sample of 71 companies. Hence, CEO turnover occurred in 12.47% of the cases and forced CEO turnover in 6.81% of the observations.
In the balanced panel analysis, we considered both accounting-based and market-based returns, computed at the end of the year previous to that when the CEO turnover occurred. In subsidiary, we also took into account in our models the annual returns of the year when the CEO was changed. We presumed that effects of the dismissed CEO's activity might also occur after the CEO was changed, due to his/her strategic position in the company. In the unbalanced panel analysis, the performance indicators were determined for 12 months before the date of the succession announcement and, because we did not exclude observations for CEOs who served less than a year, we computed the return for the period when they held the position and annualized the result.
To compute industry annual return, we used all of the Romanian companies listed on the BSE and on the over-the-counter market (former RASDAQ) that had the same first two industry code digits. It has been proven that this classification offers a higher homogeneity of returns than statistical clustering or three-or four-digit classification (Chan et al. 2007, Clarke 1989). To minimize the possibility of biases in our results due to outliers, the performance variables, the price/book value, and the leverage ratios were winsorized at 1% and 99%. We did not include the same regression variables that had a correlation coefficient equal to or greater than 0.4. Following the literature and the hypotheses presented in Section 2, we applied a binary logit model to estimate the forced CEO turnover equation. Logit regressions are commonly used in management turnover studies in order to show that the likelihood of turnover is sensitive to changes in firm performance (see Powers, 2005 for an extensive literature in this field). Consistent with previous literature, we included as independent variables several accounting-based and market-based performance indicators, various corporate governance indicators, variables related to the financial features of the firm, and industry dummies. Additionally, we included new variables for foreign CEOs and female CEOs, for the financial crisis that started in 2008, as well as an interactional variable for government ownership and political changes.
The use of composite variables in logit regressions was proved to be problematic by Powers (2005) due to the non-linearity of the model employed. However, in our analysis, the composite variable acts as a stand-alone variable, as its components were not used separately in the regression. Consequently, it can be interpreted as such, the sign of its coefficient being statistically significant.
Our sample included all firms listed, during the 2005-2010 period, on the first and second tiers of the BSE, except for the BSE itself (a listed company which became public in June 2010). We studied public companies because, for small and/or unlisted Romanian companies, it would be very difficult to collect these data, and the corporate governance mechanisms are expected to not be properly implemented. The sample cannot be extended due to the small dimension of the Romanian capital market. The 36 companies that experienced CEO turnovers represent one half of all listed firms on the first and second tiers of the BSE in 2010, with average sales of 904 million RON, an average number of employees of 1728, and an average stock market capitalization of 985 million RON. The analyzed time span includes a period of exceptional development of the Romanian capital market (2005)(2006)(2007)(2008), but also one of economic recession (2009)(2010). Besides, this period is particularly interesting due to the ongoing corporate governance reforms and the adoption of the Romanian Code of Corporate Governance in 2009.
Most of the data regarding the CEO's identity, CEO ownership, board independence, the reasons for CEO turnover, the succession type (voluntary/forced), and the exact date when the CEO was changed were hand-collected through corporate websites. Additionally, different Romanian business newspapers (Ziarul Financiar, Capital, Săptămâna Financiară, and Business Magazin) and news agencies (Mediafax and Daily Business) were accessed. Data for corporate governance variables and the accounting and financial data were from quarterly and annual company reports provided by the BSE. The explanatory variables are described in Table 1. Notes: Furthermore, we constructed two variables for the second largest shareholder ownership, as a percentage of common shares, and respectively as a dummy variable equal to 1, if the second shareholder's ownership equals or exceeds 10% of the total voting rights and 0 otherwise. We found no empirical evidence of a correlation between these variables and the CEO turnover occurrence. Some descriptive statistics for the variables used in the model are presented in Table 2. These data emphasize that the average values for the firm's performance ratios are low, reflecting a negative influence of the global financial crisis starting in late 2008. The same trend is sustained by the price to book median value of 0.95, suggesting that, for more than half of the companies in the sample, the market value of the company is lower than the book value of shareholders' equity. This finding may also be related to the low liquidity and low degree of informational efficiency of the Romanian capital market.  Table 2 shows that 11% of the listed companies have foreign CEOs. We found evidence that the foreign origin of a manager is associated with the fact that the company is part of a foreign group. Almost 12% of the managers in our sample were women. This is not a high figure, but compared to other statistics, we can expect a relative significance of the manager's gender on the likelihood of CEO turnover occurrence. For instance, Kato and Long (2006) emphasize that in 2002, the proportion of female CEOs was only 4% in China.
The average first shareholder's ownership was around 47%, with a median value slightly above 50%. This result is important to characterize the corporate governance context of the Romanian market. It is presumably associated with important agency problems, especially between controlling and minority shareholders, and with high opportunities of extracting private benefits from control on a capital market with high control premiums. 3 The average percentage of votes held by all institutional investors, each of them holding more than 5% of the votes, was 26.07%, but in half of the cases, it was below 11%. The average ownership of institutional investors was around half of the average institutional ownership for US firms in the work by Helwege et al. (2012). The standard deviations for all the variables regarding institutional ownership recorded high values, showing significant differences between the listed companies in our sample. Only 29% of the companies had institutional investors who own more than 50% of the equity. This may lower the incentives of institutional investors to be actively involved in controlling the companies efficiently. Due to the fact that, in the existing literature, it was shown that the role and involvement of institutional investors in the decision to change CEOs (forced) is even greater as they hold more shares (Kang et al. 2018), the results for this independent variable in the case of Romania are somewhat uncertain.
The average board size is around five members, with a maximum value of eleven members. These values are somewhat similar to the size in other one-tier board systems, such as in the US (Yermack 1996;Denis et al. 1997;Denis and Sarin 1999), Belgium (Renneboog 2000), Italy (Brunello et al. 2003), and China (Kato and Long 2006). We can bring into the discussion the arguments of Yermack (1996) for smaller and more efficient boards that monitor the executive managers more effectively. Thus, smaller board size can emphasize the negative relation between firm performance and CEO turnover. Table 3 presents evidence on the relation between firm performance, CEO characteristics, political influence, industry sector, or financial crisis and CEO turnover likelihood. In accordance with H1, the interactional variable for political change and governmental ownership is positively and significantly correlated with forced CEO turnover. This relation is proof of the political influence on the Board of Directors' decisions, especially when the Romanian government owns an important share of total equity. Consistent with H2, the foreign origin of the CEO for the Romanian listed companies is positively and significantly correlated with the likelihood of forced CEO turnover (see Table 3). The 3 Dragotă et al. (2013) documented an average value of control premium around 115% and a median one of 25% for 173 tender offers carried through the BSE and RASDAQ for the period 2000-2011.

Results and Discussion
probability of forced turnover was higher (i.e., 15%-20%) if the CEO was of foreign origin. This result is in line with Boenisch and Schneider (2013), who found that one of the communist legacies is the formation of strong closed informal connections, difficult to breach, and it can also be a clue for explaining foreign managers' difficulties to adapt to the Romanian environment.
We found a positive and significant correlation between the female CEO dummy and CEO forced turnover likelihood (similar to the results from Kato and Long 2006). The probability of forced turnover was higher (i.e., 7%-8%) when the CEO was a woman.
We also found evidence that the crisis dummy is positively and significantly correlated with forced CEO turnover. The probability of forced CEO turnover was around 7% higher in 2009, amidst the crisis, than in other years from the database. This result is in accordance with Kaplan and Minton (2012), concluding that the expectations regarding the industry and market performances (decreasing during an economic crisis) negatively influence the probability of forced CEO turnover. An unstable economic and financial environment also puts pressure on the Romanian listed companies' management, and, consequently, the probability of forced CEO turnover is higher than in more stable economic times.
Our results are in line with the results from the existing literature and show that CEO turnover (particularly forced replacements) is inversely related to firm and industry performances (Hu and Leung, 2012;Kato and Long, 2006;Muravyev et al. 2010, among others). We also found that companies with low industry-adjusted return are more likely to have forced CEO turnovers (see Eisfeldt and Kuhnen, 2013). Due to the anomalies of the Romanian capital market, we expected the accounting-based performance indicators to be more reliable than the market-based returns. We used similar models for forced CEO turnover and the three different market-based performance measures (price to book value, total stock return, and capital gains return), but we found no significant relation with the probability of forced CEO turnover 4 .
The average marginal effects explain how sensitive the probability of CEO forced turnover is to independent variables. Some differences can be noticed in the amplitude of the average marginal effects, depending on the performance measures used. A 1% decrease in the return was correlated with an increase of the probability of CEO forced turnover of around 0.3% if we measured firm performance using ROE, 0.5% if the performance indicator was ROA, and 0.54% for the operating margin. These results suggest that the Romanian listed companies put less emphasis on shareholders' interests than on the whole capital invested by creditors and shareholders for CEO turnover decisions.
A 1% decrease of the three-year average return was correlated, with slightly lower increases in the likelihood of forced CEO turnover than in the case of the performance indicators computed for the previous year. This result could be an argument for the assumption that for the Board of Directors, the short-term perspective of the performance is prevalent before the long-term perspective in their forced CEO replacement decisions. Our results are similar to Kaplan and Minton (2012), but with lower magnitude than those of Eisfeldt and Kuhnen (2013). One possible explanation for the difference in the magnitude of the changes can be the occurrence of the financial crisis during the analyzed period.
We found almost similar results for the likelihood of forced CEO turnover when we used contemporary firm performance measures. When we used three-year average ROA and ROE, the results were no longer statistically significant.
Regarding the scarcity of the managerial labor market, we found evidence that the utilities, energy, and pharmaceutics sectors are more likely to force CEOs to leave their positions. Our result is in accordance with Parrino (1997), who found that the utilities and energy sectors are among the most homogeneous.
With regards to the relation between CEO turnover and different corporate governance characteristics, first we found that the CEO-Chairman duality is inversely related to the likelihood of forced CEO turnover (Hazarika et al. 2012;Helwege et al. 2012;Hu and Leung 2012). This correlation can be explained through a higher influence of the CEO on the Board of Directors and higher opportunities for the CEO to maintain informational asymmetries. For our sample, the probability of forced CEO turnover was lower, by around 10%, for companies where CEOs held a dual position, compared to the rest of the sample.
Second, the presence of the controlling shareholder was associated with a lower probability of forced CEO turnover (see Brunello et al. 2003;Parrino 1997), which is consistent with a greater probability of type II agency conflicts. All other variables related to ownership had no statistical significance.
Third, similar to Parrino et al. (2003), the "voting with their feet" behavior can be validated for the Romanian listed companies. By some means, this result is expected for a transitional country with an enforced Corporate Governance Code only from 2009. Moreover, the results may be due to the lower holdings of the institutional investors compared to developed countries. For US companies, Helwege et al. (2012) showed that although the net change in institutional ownership prior to forced CEO turnover is negative and significant in the first sub-period of their analysis, the extent of "voting with their feet" is small and declines over time, and so becomes irrelevant in the second sub-period. The authors signal that the role of institutional investors in forced CEO turnover tends to diminish and that research must move toward other determinants. Such empirical evidence could also explain a lower statistical significance for this variable, registered for our Romanian sample for the period 2005-2010.
Fourth, we found no influence of board size and independence (as in Muravyev et al. 2010). This result can be explained by a limited effectiveness of corporate governance mechanisms, or through the activism of the first large shareholders, who put in place other mechanisms to control the managers. Additionally, we cannot exclude the assumption of a passive attitude of the outside directors. The results are somewhat similar to Miyajima et al. (2018), where the independent outside directors have no significant effect of enhancing CEO turnover sensitivity to ROE. The CEO turnover sensitivity to ROA is higher when boards have three or more independent outside directors than when boards have only one outside director.
Leverage is negative and statistically significantly correlated with the likelihood of forced CEO turnover (see Table 3). This link shows the existence of agency conflicts between managers and shareholders, which can be alleviated using the supplementary monitoring realized by the creditors. The creditors have the concern and appropriate skills to monitor the firms, and the shareholders benefit from their financial expertise.
Due to a relatively high homogeneity in our sample regarding the size of the Romanian listed companies, firm size becomes less important, but remains significant in explaining the CEO turnover decisions. The results for average marginal effects show that a 1% increase of the logarithm of net sales was associated with a decrease of around 2.5% of the likelihood of forced CEO departure.
Additionally, we analyzed separately voluntary and forced CEO turnovers. For voluntary CEO turnover, we found similar results as for all CEO turnovers in regard to the foreign origin of CEOs and ownership, supporting the explanations mentioned above. We found no statistically significant relation between firm performance and the likelihood of voluntary CEO turnovers, neither for accounting-based nor market-based returns. This result is consistent with mainstream literature, concluding that forced CEO turnovers are related to firm and industry performances, whereas other forms of managerial turnover are not linked to firm performance (see Eisfeldt and Kuhnen 2013).
The unreported results for the pooled data analysis were similar for CEO turnover, and also for forced and voluntary CEO turnovers. 5 Additional robustness tests were performed by gradually introducing into the logit models the explanatory variables for both specifications (balanced and unbalanced panel data). The results confirm the significance of all variables.

Conclusions
We revisited a widely studied topic in the literature-the determinants for the likelihood of CEO turnover. To the best of our knowledge, this is the first paper to focus on CEO turnover in Romania using a detailed and comprehensive analysis and considering financial, social, political, and corporate governance determinants, applied to an intricate period of time, with both exceptional development and economic recession.
For the Romanian case, our study took into consideration, for the first time, explanatory variables for the foreign origin of a CEO, CEO gender, and the incidence of political intrusions in corporate decisions in companies with governmental ownership. Thus, the occurrence of CEO turnover was positively correlated with the foreign origin of the manager, supporting the hypothesis that foreign managers experience difficulties in adapting to the Romanian economic environment. The political influence on the CEO turnover decision was revealed when firms had large state ownership. We found evidence that female CEOs were more likely to leave their positions than male CEOs. We also showed that the financial crisis boosted the probability of CEO forced turnover.
We examined, as well, the role of other important corporate governance variables in the CEO turnover process, and the results are mixed. The empirical evidence confirmed that some reforms have been implemented. Two particular entrenchment mechanisms were emphasized, namely, the CEO-Chairman duality and the controlling power of the largest shareholder. Additionally, we emphasized the "voting with their feet" behavior of the institutional investors. The CEO ownership is not significant as an entrenchment mechanism for forced CEO turnovers, but we found some intriguing evidence of its direct relation to the likelihood of voluntary CEO turnover occurrence. Other well-known corporate governance mechanisms, such as the second largest shareholder and the institutional investors' activism, seem ineffective in the sample considered.
As for the relation to several performance indicators, our results were similar to those identified in the previous financial literature for other countries. The CEO turnover decision was strongly and negatively related only to accounting-based performance ratios. Whatever model applied, the market-based performance variables did not have a significant influence, and a possible explanation could relate to the features of the Romanian capital market.
As per many empirical studies, this one has some limitations. The most important limitation concerns the sample, which was relatively small (426 observations, with 58 CEO turnover events) compared to similar studies in this field. This limit can usually be found in empirical studies that analyze emerging economies, especially those of Eastern and Central Europe. Due to the fact that their capital markets are newly created or recently reopened, in many cases, the whole number of listed companies is small and thus it is not possible to collect data sets as large as those of studies conducted in highly developed countries (Filip and Raffournier, 2010). In our study, this criticism should be mitigated, given that the sample includes all of the Romanian listed companies (except the BSE).
Our results can be useful for practitioners, as well as for academics. The conclusions are important for investors, and especially for foreign investors not accustomed to the Romanian economic environment. Additionally, our findings provide a better understanding of the managerial labor market in a transitional Eastern European country and evidence that it is still subject to political interference in corporate decisions of state-owned companies subsequent to a change of the political party in power.
We expect the development of post-communist economies to determine substantial improvements in corporate governance mechanisms and practices, and the CEO turnover phenomenon to follow the trends revealed in the literature for developed countries. It would be interesting for future analysis to mainly capture this evolution for a better understanding of the causes that trigger it.
Author Contributions: All authors contributed to the design, introduction, literature review, results, and discussion. The database and methodology section was mostly written by A.C.S and R.M. The original draft preparation, the review and editing processes, and the funding acquisition were mostly done by I.M.D. and A.C.S. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the Center of Financial and Monetary Research (CEFIMO). All opinions expressed are those of the authors and have not been endorsed by CEFIMO.

Acknowledgments:
The authors would like to express their gratitude to the Bucharest Stock Exchange (BSE) and Tradeville representatives, who provided part of the preliminary database. The results and conclusions of the paper are entirely the authors' responsibility and the BSE cannot be held liable for any of the opinions stated in this research paper. The authors also wish to thank Victor Dragotă, Elena Dumitrescu, Elena Ţilică, Radu Ciobanu

Conflicts of Interest:
The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; to publish the results.

Appendix. The main reasons for CEO turnover for the Romanian listed companies 2005-2010
1. Retirement (4 cases). We consider retirement before the age of 63 as forced retirement. 2. Health problems (2 cases). We consider these cases as unforced reasons. 3. Political reasons (4 cases). Hu and Leung (2008) define three types of government intervention: (1) the government can play the role of "invisible hand" and senior executives of state-owned companies are political appointments; (2) the government has no control over the selection and appointment processes; or (3) the appointment is "just window dressing, as politically affiliated directors are puppets of management." In Romania, the first case is the most common practice for state-owned companies. In our sample are two state-owned companies, Transelectrica and Transgaz, with CEO replacement in 2009. The replacement was made in both cases by the Ministry of Economy in the first months after the new political party won the general election. Additionally, in both cases, there was an exchange between the position of the CEO and the deputy director. It can be noticed that the deputy directors were previously former CEOs in the same companies. The new senior executives appointed are controversial people, influential politicians with strong connections to the political party who won the general election. 4. Promotion (3 cases). This refers to a CEO who was promoted and made "a step forward" to take a new position within the company. From our sample, the most relevant example is the CEO from OMV, who resigned following the decision of the OMV group to become senior counsellor of the Chairman of the OMV group. 5. Persona non-grata (4 cases). Multiple reasons can be considered to justify the decision to replace the former CEO, from prolonged conflict with the board to conflict of interests or fraud; all of them are seen as forced and illegal decisions by the CEO who was changed. Often, the real reason for CEO replacement is not publicly announced, and thus the mass-media becomes the main source to learn about the internal conflicts within the company. 6. Change of ownership (4 cases). If the ownership changes, CEO turnover is inevitable. In this study, we consider these replacements as involuntary. 7. Common succession (4 cases). In our sample, these CEOs just finished their four-year mandate and took other functions after their contract expired. 8. Pressure from the shareholders (16 cases). This is the most common reason for CEO replacements for the Romanian listed companies for the period 2005-2010. These cases are as follows: CEO dismissal due to low performance, change of jobs to improve firm performance, and a new approach of the company or CEO turnover decided by the majority shareholder(s). 9. Separation of the Positions of CEO and Chairman of the Board (5). The BSE Corporate Governance Code recommends avoiding CEO duality, as per many other similar Codes, in an international context. On the contrary, the Romanian listed companies can be considered as following the CEO-Chairman duality pattern. 10. CEO resignation (3). Only one case can be considered a CEO resignation, due to prolonged conflicts with the Board concerning the company's future direction. For the other two cases, the public announcement mentioned only that it was the decision of those CEOs to resign from their current positions. 11. Other reasons (2). We include two cases of CEOs who decided to remain only shareholders in their companies and to be actively involved in other businesses. 12. No reason (8). From 59 cases of CEO turnover, in 8 cases, the companies did not give any public motivations for these decisions.