1 / 39

Should Firms Look to an Insider or an Outsider When Hiring a New CEO? Evidence from China

Should Firms Look to an Insider or an Outsider When Hiring a New CEO? Evidence from China. Feng Helen Liang Haas School of Business, UC Berkeley April 21, 2006. CEO Turnover Happens a Lot CEO turnover rate increased from 6% to 10.1% in world’s 2500 largest firms 1995-2002.

cate
Télécharger la présentation

Should Firms Look to an Insider or an Outsider When Hiring a New CEO? Evidence from China

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Should Firms Look to an Insider or an Outsider When Hiring a New CEO? Evidence from China Feng Helen Liang Haas School of Business, UC BerkeleyApril 21, 2006

  2. CEO Turnover Happens a LotCEO turnover rate increased from 6% to 10.1% in world’s 2500 largest firms 1995-2002 Source: CNET news.com

  3. It’s Costly to Find Mr./Ms. RightThe number of search firms in North America tripled 1990-2000, hitting revenues of $8.7 billion in 2000 Source: Kennedy Information LLC

  4. More outsiders hired for the top positionThe percentage of outsider succession doubled in 800 large companies 1990-2000 Source: Watson Wyatt Worldwide

  5. But does this searching effort pay off? • “Outsiders are hired to shake up the firm, but they are a high-risk gamble.” • “Outsiders generate rapid results.” • “But they lose momentum later because of lack of internal support network.” --Booz Allen Hamilton Annual Study of CEO Succession • Empirical studies found little systematic relations between successor type and performance, due to • Mean reversion • Noisy dependant variables: stock price (Huson et al 2004) • Selection

  6. What’s new in this study? • A rich data set on Chinese firms • Details on CEO background and firm operation • Control for selection between outsiders and insiders • Productivity: a more accurate performance measure • Developing country setup

  7. Why China? • Rapid economic development • World manufacturing center • Restructure in the state-owned sector  widespread managerial turnover

  8. This research asks two questions about CEO turnover and firm performance: • RESEARCH QUESTIONS: • When do firms hire an outsider over an insider in choosing a new CEO? • Are outsider CEOs better at improving productivity than insiders? What are the channels?

  9. What I found -- Total Factor Productivity before and after turnoverTFP goes up in outsider-succession firms, and goes down in insider-succession firms

  10. Findings • Firms that hire outsiders do better: • Total factor productivity improves 2-3% more • Significant in large state-owned firms, but not in private firms • Managers’ ties with government and other firms help to improve productivity

  11. Outline of the Talk • Introduction • Theory and Hypotheses • China’s context • Data • Empirical Strategy • Results • Conclusion and Discussion

  12. Theory & Hypotheses I: CEO Selection • Difference between an outsider and an insider • Smaller firms less costly to hire outsiders • Less tacit org knowledge to learn (Chung et al 1987) • Fewer insider candidates available • External turnover hurts middle-level managers’ incentives (Chan1996) • Less technology complexity  outsider • Poor pre-turnover performance  hire outsider to turn around • Strategy discontinuity  outsider • Strategy maintenance  insider • (Dalton and Kesner 1985, Weisbach 1988) • H1. A firm is more likely to choose an outsider CEO when the firm: • Is small • Does not have a R&D department • Has poor pre-turnover performance

  13. Theory & Hypotheses II: post-turnover performance • Implicit contract: • Incumbent managers are bound by an implicit contract with the workers and could entrench themselves with manager specific investment (Shleifer & Summers 1988, Shleifer & Vishny 1989, Bertrand & Mullainathan 2003) • Thus they might sacrifice shareholder’s interests via: • Suboptimal labor allocation • Buy peace with higher wages • An outsider can’t capture such rents after turnover  Outsider CEOs could allocate human resource more efficiently than insiders do • H2. Everything else equal, an outsider CEO is more likely to improve firm productivity, esp. when the firm • Has a large employment, and • Higher proportion of skilled labor

  14. Theory & Hypotheses III: post-turnover performance • Outsiders might bring valuable external resources • Ties with government regulatory agents, banks • Ties with clients and suppliers (Peng & Luo 2000) • Especially important in an uncertain institutional environment to • Obtain financing for investment • Secure supply, timely delivery, and customer loyalty • Safeguard contract • But such ties could be detrimental to firm performance (Bartel & Harrison 2005) • H3. Everything else equal, an outsider CEO is more likely to improve firm productivity, esp. when the firm • Has higher proportion of investment funding from the government and state-owned banks and • Has a larger amount of revenue in the form of customer credit

  15. I address the research questions in the context of China’s reform • WHY CHINA? – Restructure in State-owned sector and CEO turnover • The Reform started in 1978 • Before 1992  Gradualism • Productivity improved during 1980-1990 • Li 1997, Groves, Hong, McMillan and Naughton 1994, 1995 • Starting from 1992, the government gave more autonomy to the managers • The 14th Congress of the Chinese Communist Party decided to build a “socialist market economy” • Before 1992, most firms suffer labor redundancy and low productivity • They are forced to restructure, downsize, and improve productivity • Our data covers CEO characteristics and firm performance 1994-1999 • How did managers do in those firms? • Variation in firm ownership: State-owned and non state-owned enterprises may have different mechanisms

  16. Outline of the Talk • Introduction • Theory and Hypotheses • China’s context • Data • Empirical Strategy • Results • Conclusion and Discussion

  17. Data • Collected by the Chinese Academy of Social Science (CASS), with the researchers in University of Michigan, UC San Diego, and Oxford University (Li 1997, Groves et. al. 1994,1995) • Firm operation and manages’ characteristics • Firm Operation: • 800 firms in 1994 -1999 • Four provinces and in 36 industries: • production input and output, cost and revenue, wages, etc • 55% SOEs, 45% NSOEs • 4800 firm-year observations • Managers’ characteristics: • The new manager’s characteristics, the old manager’s status upon turnover, incentive, relation with government, etc • Limitation: Retrospective data in a balanced sample  survival bias

  18. Descriptive Statistics

  19. CEO turnover by year

  20. Proportion of Insider and Outsider CEOs after turnover Turnover rate: 85% Outside Succession ratio: 70%

  21. Compensation, Age, and Education of Insider and Outsider CEOs: Mean and Difference Insider and Outsider Successors Look Similar *** p<0.01, ** p<0.05, * p<0.10

  22. Outline of the Talk • Introduction • Theory and Hypotheses • China’s context • Data • Empirical Strategy • Results • Conclusion and Discussion

  23. Two Steps of the Estimation • Selection of CEO • A Logit model • Matching using the selection estimation results • Post-turnover performance • Total Factor Productivity estimation using the matched sample • Control for unobservables

  24. CEO Selection Estimation • Conditional on turnover, the choice between insider and outsider successor is estimated with a logit model: • Pr(OUTSIDERi = 1 | CEO turnover) = Where X include • Independent Variable: • Firm Size in 1994 • Return on Asset in 1994 • R&D Department dummy • Control variable: industry dummy, province dummy, ownership

  25. Productivity Estimation • Translog production function to estimate productivity: Yit = Ait * F(Zit) • Where F(Zit) is a translog productivity function including labor hour, capital, and material • Dependant Variable: Yit -- Output • Ait include: • Independent Variables: • New CEO dummy • OUTSIDER dummy (+) • OUTSIDER dummy interactive terms with employment, College Graduate Ratio (+) • OUTSIDER from Government or Industry dummy (+) • OUTSIDER from Gov or Industry dummy interactive terms with Gov financing and Custom Credit (+) • Control Variables: • Firm fixed effect, year dummy • CEO age, education, and tenure at the firm

  26. Empirical Strategy – Challenges • Two challenges: • Selection • Firms choosing external CEOs are different from those promoting internal CEOs • Unobserved contemporaneous shocks ωit • observable to managers but not to the researcher • influence turnover and output simultaneously • e.g. technology shock, demand fluctuation • not captured by firm fixed effects and year dummies

  27. Empirical Strategy – control for selection • Construct a control group • Treatment group: firms that hired outsiders • Control group: firms that hired insiders but have the same probability of hiring an outsider • Method • Propensity score estimated with initial conditions (Rosenbaum-Rubin1984) • Nearest neighbor matching: logit estimation with higher order terms and input variables • Selection bias is reduced if • What we see is what determined the selection – observables are sufficient statistics of the probability of treatment • Matched sample covariates are “balanced”

  28. Empirical Strategy – control for unobserved shocks • Unobserved shocks (Olley-Pakes 1996, Levinsohn-Petrin 2003): • Proxy for the shocks with a proxy variable (investment or intermediate input) and a state variable (capital) • ωit = h(INVit, Kit) • h(.) is a non-parametric estimator; a third order polynomial function is used • ωit is identified if: • Investment and intermediate inputs (energy consumption) change monotonically with ωit • Lit Mit respond to ωit immediately, while Kitresponds after a lag

  29. Outline of the Talk • Introduction • Theory and Hypotheses • China’s context • Data • Empirical Strategy • Results • Conclusion and Discussion

  30. Result on CEO Selection based on firm initial conditions in 1994 Outsider succession is more likely in smaller firms without R&D dept, and with poor pre-turnover performanceDependent Variable: 1 if the new CEO is an outsider *** p<0.01, ** p<0.05, * p<0.10

  31. Distribution of ROA in matched sample

  32. Distribution of capital in matched sample

  33. Distribution of employment in matched sample

  34. Total Factor Productivity before and after turnoverTFP goes up in outsider-succession firms, and goes down in insider-succession firms

  35. Results on post-turnover performance –Outsider CEOs improves productivity more, esp. in state-owned firms *** p<0.01, ** p<0.05, * p<0.10

  36. When do outsiders do better? – Total employment and skilled laborOutsider CEOs improves productivity more, esp. in firms w/ larger employment, but not necessarily w/ more skilled labor *** p<0.01, ** p<0.05, * p<0.10

  37. When do outsiders do better? – Linkage to government and industryOutsider CEOs linkage to government and industry matters, but not necessarily through better use of GOV funding and customer credit *** p<0.01, ** p<0.05, * p<0.10

  38. Industry AnalysisOutsider CEOs do better in low-tech industries *** p<0.01, ** p<0.05, * p<0.10

  39. Conclusion and Caveats • Firms that hire outsiders increase productivity by 2-3% more • More evident in state-owned enterprises, where the manager’s pursuit of side goals might be more pervasive • An outsider might improve productivity via better allocation of labor, though not of skilled labor • An outsider might improve productivity via external connection with the government and other firms • Caveats • Matching: I assume observed variables provide sufficient info for matching • Survival bias: good firms survived outside succession, bad firms exit • Productivity versus profit (ROA)

More Related