Abstract To the pantheon of grotesque and bloodthirsty dictators is now, on the front row, the Colonel Muammar Kaddafi who, for over 40 years, bought the silence of Western democracies with oil from his country and amuse the gallery with his folk outfit, his tents, his very close female guard, his reasoning and his degenerated offspring for a smoky diversion to hide the crimes of his regime. How the world has tolerated all these years, a fool like him who dictated the law to his enslaved people and executed worldwide opponents? Was it not enough to analyze his rambling speeches to understand that it was a Nero in power? Initially, this fictional character, out of "The Autumn of the Patriarch" by Gabriel Garcia Marquez, was probably not as crazy. Certainly was perceived in the eyes of the shy Lieutenant interviewed on television after the "White Revolution", a certain arrogance, but we were still far from the image of the future dictator. In 1969, the lieutenant was wearing a simple uniform, without medals, he smiled constantly and dodged most questions put to him, he did nothing of the clown lawless that some years later, terrorized his people and the international community. Who is he? Where is he from? How has he grown from a single officer to a despot? The phenomenon of Kaddafi was not new. The character looked like most dictators already known: totalitarian, he allowed no other political party, except his own, and model the constitution at his will to ensure the sustainability of his power; megalomaniac, he had delusions of grandeur and dreams of becoming the unifier of the Arab world; provocative and stubborn, he willingly employed a rhetoric that lacks of logic and insulted the intelligence; exhibitionist, he was always wearing, like Mussolini, accoutrements as surprising as ridiculous; irascible, he gesticulated and uttered during his speeches in the style of Hitler. Convinced that everything is permitted, he launched into projects and allowed excessive worst follies, like Idi Amin Dada, including the bombing of his own people. The paradox was the constant of his regime. Corrupt, he castigated the corruption and pretended to fight it by imprisoning some scapegoats; they became billionaires, he claimed to be "socialist" and was intended as poor as his subjects who stuck in sub-sustainable development. Champion of democracy, the one who was proud to say that   he granted the power to the Libyans by creating his "popular committees" eliminated his opponents, muzzled the media and cancelled the freedom of expression. For most of us, leadership has become synonymous with competence, courage, and good character. How then do we explain these atrocities? This fraud? And other scandals? In spite of the countless glaring examples of bad leadership, why do we cling to an idealized notion of leadership that is more imagined than real? This thesis is about the dark side of human condition. It paints leadership in shades of gray and black. In spite of all the work on leadership that assumes it by definition to be good, I describe how it is possible to exercise power, authority and influence in ways that do harm. This harm can be the result of the madness of one man and has calamitous consequences. The intend is here to draw attention to how and why it happens. I contend that this type of leadership is not an aberration nor a unique case but an ubiquitous and insidious part of everyday life that must be carefully examined and better understood.

Knowledge about the antecedents to and the consequences of innovation is often studied in a fragmented way, resulting in an incomplete understanding of the dynamics that drive organisational performance. The purpose of this study is to develop a comprehensive model explaining the relationship between leadership style, organisational climate, innovation and organisational performance. The systematic literature review procedure was used to identify, analyse and critically evaluate studies that examined the relationship between leadership style, organisational climate, innovation and organisational performance. Given this information as background a cross-sectional survey design was used to test the relationship between the named variables. Firstly, a measurement model was tested with data collected from 3 180 respondents, representing 52 companies. Secondly, a structural path model was tested, with data collected from 231 employees representing 112 companies. The findings of the systematic literature review revealed that empirical studies that link the four variables are scarce; in the main, combinations of three variables are found. It also revealed that measures of innovation and performance vary vastly, inhibiting the incremental development of a comprehensive empirical body of knowledge. The results of the measurement model substantiated differentiation between leadership styles and the expected positive correlation between both transformational and transactional leadership and innovative behaviour. Furthermore, the results showed that not all components of leadership impacted positively on innovative behaviours. The structural path model showed that a transformational leadership style has a direct impact on the organisational climate, innovation and organisational performance. In contrast, a transactional leadership style had a direct impact on organisational performance, but no relationship was found between transactional leadership style and organisational climate and innovation. This study is important as it provides a unified model of innovation that focuses on both antecedents, as well as the outcomes of innovation, in a more comprehensive manner than any previous study. / Graduate School of Business Leadership (SBL / DBL

Comprendre ce qui amène un leader à émettre des comportements de leadership transformationnel fascine les chercheurs et praticiens depuis plusieurs années (Bommer, Rubin, & Baldwin, 2004; Bono & Judge, 2004; Shamir & Howell, 1999 ; Stogdill, 1948; Yukl, 1999). Or, jusqu’à présent, ces facteurs sont encore bien peu étudiés et compris comparativement aux conséquences de ce style de leadership. Dans cette lignée, la présente thèse répond à différents enjeux soulevés par les auteurs à ce sujet (Dinh & Lord, 2012; Zaccaro, 2007) en cherchant à préciser le rôle joué par différents antécédents individuels et contextuels du leadership transformationnel. Cet objectif sera poursuivi par l’adoption d’une perspective interactionniste qui intègre des antécédents de personnalité et de contexte ainsi par l’évaluation de la personnalité à l’aide d’une modélisation de la personnalité orientée vers le critère à prédire (variable composite). La présente thèse est composée de trois articles poursuivant les objectifs suivant : 1) Effectuer une synthèse de la littérature empirique portant sur les antécédents individuels et contextuels du leadership transformationnel; 2) Vérifier les liens empiriques entre la personnalité mesurée à l’aide de variables composites, plus précisément le modèle des Great Eight de Bartram (2005), et le leadership transformationnel; 3) Tester empiriquement l’effet d’interaction entre les variables de personnalité et les variables contextuelles pour prédire le leadership transformationnel. Le premier article vise d’abord à circonscrire et organiser les connaissances empiriques actuelles provenant d’une quarantaine d’articles concernant les antécédents du leadership transformationnel. L’article s’organise en trois thèmes principaux : les antécédents individuels de personnalité, les antécédents contextuels et l’étude des interactions entre le volet individuel et contextuel. Plusieurs constats et pistes de recherches sont discutés et mettent la table pour les deux articles subséquents. Ainsi, le second article s’intéresse au potentiel explicatif d’un modèle de personnalité orienté vers le critère pour prédire le leadership. Plus spécifiquement, le modèle des Great Eight proposé par Bartram (2005) est mis en relation avec les comportements de leadership transformationnel et de récompense contingente. Les résultats, obtenus auprès de 113 gestionnaires et de leurs 799 subordonnés, donnent peu d’appui à la valeur ajoutée du modèle utilisé, mais indiquent que certaines tendances de personnalité sont associées au leadership. Des analyses supplémentaires permettent de nuancer la compréhension des effets observés dans la documentation scientifique et offrent quelques pistes de groupements de traits pouvant prédire les différents comportements de leadership. Le troisième article s’inspire de la théorie de l’activation des traits (Tett & Burnett, 2003) pour vérifier l’effet combiné de la personnalité du gestionnaire et du contexte dans lequel il évolue en vue de prédire le leadership transformationnel. Les résultats (ngestionnaires = 89; nsubordonnés = 643) n’offrent qu’un appui modéré au rationnel sous-jacent du modèle de l’activation des traits. Toutefois, il en ressort que l’aspect relationnel du gestionnaire (opérationnalisé par le composite de personnalité « soutien et coopération ») est associé à l’émergence du leadership transformationnel uniquement lorsque les facteurs contextuels (considération organisationnelle, latitude décisionnelle) sont perçus positivement par le gestionnaire. L’étude permet donc d’éclaircir une part de la variabilité observée dans les études antérieures concernant la tendance relationnelle du gestionnaire, en soulignant sa sensibilité à des facteurs contextuels positifs. / Understanding what can influence a leader to demonstrate transformational leadership behaviours has intrigued researchers and practitioners for several years (Bommer, Rubin, & Baldwin, 2004; Bono & Judge, 2004; Shamir & Howell, 1999; Stogdill, 1948; Yukl, 1999). Yet, until now, these factors were little researched and poorly understood in comparison to the consequences of this type of leadership. Along these lines, this thesis answers to numerous issues raised by authors on this topic (Dinh & Lord, 2012; Zaccaro, 2007) by seeking to clarify the role of different individual and situational antecedents of transformational leadership. This objective is pursued through the adoption of an interactionist perspective which integrates individual and contextual antecedents to the assessment of personality supported by a criterion-centric personality model (compound variable). This thesis is composed of three articles that pursue the following objectives: 1) Review empirical literature on individual and contextual antecedents of transformational leadership; 2) Validate the empirical links between a new personality model (i.e. compound variable) and transformational leadership; 3) Test empirically the correlations between personality and contextual variables to predict transformational leadership. The first article aims to review and organize the current empirical knowledge from more than forty articles regarding transformational leadership indicators. The article is built on three major themes: the individual personality factors, the contextual elements and the study of the interaction between both. Multiple findings and research leads are discussed and set the stage for the following articles. In turn, the second article investigates the explanatory potential of a criterion-centric personality model to predict leadership. Specifically, the Great Eight model, put forward by Bartram (2005) is leveraged to understand the relationships between behaviours tied to transformational and transactional leadership. Results from a survey of 113 managers and their 799 subordinates provide little support for the explanatory value of this model, but indicate that some personality traits are associated with leadership. Further analysis allows us to put the results in perspective with the academic literature and provide some insights regarding the potential for some personality trait groupings to predict leadership behaviours. The third article builds on the traits activation theory (Tett & Burnett, 2003) to validate the combined effects of a manager’s personality and the context in which he or she operates to predict transformational leadership. The results ((nmanagers = 89; nsubordinates = 643) offer limited support to the rationale underlying the traits activation theory. Yet, it appears a manager’s relational traits (operationalized by the personality compound variable “supporting and cooperating”) are correlated to demonstrations of transformational leadership, when contextual factors (organizational consideration, decision latitude) are perceived positively by the manager. As a result, this research sheds some light on an element of variability put forth in previous studies regarding a manager’s relational inclinations by highlighting its sensitivity to positive contextual factors.