Literature Review Women on boards and in TMTS and firm performance46 are that an executive s gender is random that boards discriminate based on gender that women self select into particular types of firms Huang Kisgen 2013 or that the selection of a female leader might be related to an emphasized focus on female customers Strøm et al 2014 More over time delayed data collection of independent female representation gender diversity and de pendent variable firm financial performance is a central requirement for the proof of causation Kroell et al 2014 Potential effects of women in executive positions on firm performance will be come apparent only after a certain period of time has elapsed In summary parameter estimates might be bia sed and inconsistent due to endogeneity making reliable inference difficult or impossible Roberts Whited 2012 It stands to reason that the me thod applied has a notable influence on the results of empirical studies and also on the interpretation of results and the conclusions drawn from them However within the framework of the present li terature review I may only highlight key aspects of endogeneity and sketch important trends in de aling with potential endogeneity The early papers in this review do not take the aspect of endogeneity into account e g Erhardt et al 2003 Shrader et al 1997 Siciliano 1996 Single OLS regression for instance does not expli citly control for endogeneity and can thus produce biased coefficient estimates Carter et al 2003 However during the past decade studies have applied other methods to their panel data set underscoring the efforts to enhance validity and re liability of the models and to rule out endogeneity Dobbin and Jung 2011 underline the importance of controlling for endogeneity Early cross sectional studies suggested that board gender diversity has positive effects on both profits and stock perfor mance However the use of panel data and statistical methods designed to rule out endogeneity suggested that fema le directors tend to have neutral or ne gative effects The big picture seems to be that gender board diversity does not help firms and it may hurt them Dobbin Jung 2011 pp 836 In order to control for possible endogeneity be tween gender diversity variable and firm value performance various regression models and sta tistical test procedures are applied Lagged dependent variables and the inclusion of fixed effects are used striving to mitigate omitted variables and to address unobserved changes over time as well as industry and firm specific charac teristics e g Adams Ferreira 2009 Carter et al 2010 Fixed effects estimation can possibly im prove unobservable heterogeneity bias The un derlying strong assumption of exogeneity might be unrealistic though as current values of the in dependent variable are not likely to be completely independent from past values of the dependent variable Wintoki et al 2012 Instrumental variables IVs are another possibility to handle endogeneity An instrumental variable replaces the endogenous independent variable in the regression It should be sufficiently correlated to the endogenous independent variable but must not be correlated to the error term Happ 2016 IVs are often used in two stage least squares 2SLS regressions Two stage least squares regres sion is employed by Campbell and Minguez Vera 2008 Carter et al 2003 Bøhren and Strøm 2010 or Khan and Vieito 2013 and three stage Women on boards and in TMTs and firm performance

Vorschau DIRK-Forschungsreihe Band 21 Workforce diversity and personal policies Seite 46
Hinweis: Dies ist eine maschinenlesbare No-Flash Ansicht.
Klicken Sie hier um zur Online-Version zu gelangen.