Corporate Ownership & Control / Volume 3, Issue 2, Winter 2005-2006 (continued) 165 DETERMINANTS OF BOARD COMPOSITION: EVIDENCE FROM TUNISIAN COMPANIES Sonda Marrakchi Chtourou*, Soumaya Ayedi**, Yosra Makni Fourati*** Abstract This study focuses on the composition of boards of directors in the Tunisian context. We model the composition of the board of directors as a function of alternative governance mechanisms, some board characteristics and other control variables. On a sample of 97 Tunisian firms, we find evidence that the proportion of outsiders on the board of directors is positively associated with large block, institutional and overseas ownerships, and board size. We document that the CEO duality is associated with a decrease in the board independence. We fail to find an evidence that increased debt ratio to total assets is inversely associated with the outside board representation. While we predict a positive relationship between the board independence and the firm size, the organizational complexity and the quotation status; our results generally do not support this conjecture. Keywords : board composition, corporate governance, agency conflicts, entrenchment. * Unité de Recherche en Gestion des Entreprises, Faculté des Sciences Economiques et de Gestion de Sfax sonda.chtourou@fsegs.run.tn ** Unité de Recherche en Gestion des Entreprises, Faculté des Sciences Economiques et de Gestion de Sfax soumayaayedi@8-0.net ** Unité de Recherche en Gestion des Entreprises, Faculté des Sciences Economiques et de Gestion de Sfax ymakni@yahoo.fr 1. Introduction The board of directors has long been recognized as a major structural mechanism to curtail managerial opportunism. In the modern corporation, the separation of ownership from control results in potential agency conflicts stemming from divergence between managerial and shareholder interests. In general, the small shareholders delegate their authority to the board of directors which is charged with the task of representing the shareholders’ interests. The board delegates decision making to the managers and is responsible for determining long run targets of the company and for controlling managerial decisions. This situation leads to an agency problem, since the managers can use the firm’s assets to serve their own interests in the detriment of those of the shareholders. The central point of the effectiveness of any board of directors is its composition. From an agency perspective, the outside directors are objective and independent, especially in evaluating issues closely related to the fate of internal managers (Fama and Jensen, 1983). Outsiders have particular incentives to monitor the managers on behalf of the shareholders because of their reputation on the external labour market (Fama and Jensen, 1983). As argued by (Prevost et al, 2002a), within an agency theoretical context, the determinants of the board composition can be classified in three major areas: the alternative corporate governance mechanisms, the other board characteristics and the potentially important control variables. This study examines the determinants of the board composition in the context of the Tunisian market. Previous studies were conducted mostly in the US, UK and other comparatively large markets where the institutional environments differ greatly from that in Tunisia. Since institutional differences may have important implications for corporate governance in different countries (Shleifer and Vishny, 1997), the results of this study can thus enhance the understanding of how institutional differences impact on corporate governance. Using a cross- sectional sample of 97 Tunisian firms, we find evidence that the proportion of outsiders on the board of directors is positively associated with large block, institutional and overseas ownerships, and board size. We document that CEO duality is associated with a decrease in board independence. We fail to find evidence that increased debt ratio to total assets is inversely associated with outside board representation. While we predict a positive relationship between board independence and firm size, organizational complexity and quotation status; our results generally do not support this conjecture. The remainder of this paper is organised as follows. The next section briefly reviews the previous studies that have investigated the board composition. Section three gives a brief idea about of the formal legislative framework of the Tunisian corporate governance.