1 A History of Leadership Keith Grint INTRODUCTION Wliy bulher with the beginning? Is history, as Henry Ford said, 'just one damned thing after another'? Well it might be, though note that Ford proceeded to build Ms own museum of the Ford Motor Company - to make sure his version of one damned thing after another was the accepted version. It might also be worth considering whether history provides us with examples of failed and successful leadership that we might learn from; as, George Santayana (1954) suggested, "Those who do not learn from history are condemned to repeat it.' But does history repeat itself? Marx certainly thought there was something in this with the opening line of his work on the 18th Bnmwire of Louis Napoleon; for 'Hegel remarks somewhere that all great world-historic facts and personages appear, so to speak, twice. He forgot to add [he first time as tragedy, the second time as farce.' Yet if history did repeal itself we should be able to predict events rather belter than we currently do. We should, for example, have seen the current economic problems coming if we were historians of the interwar period. Yet even though events are not identical and are thus not repeated, are there patterns that might give us clues as to what might and might not happen? Indeed, does not almost every academic journal article start with a potted history of some sorts to explain why the author has decided to focus on whatever follows? Perhaps, but, what counts as the beginning anyway? Well we can start by agreeing that 'the beginning' for leadership scholars is the beginning of recorded history, not the beginning of Homo sapiens. As far as it is possible to tell, all organizations and societies of any significant size and longevity have had some form of leadership. often, but not always, embodied in one person, usually - but nol always - a man. This does not necessarily mean that leadership has always been, and will always be, critical or essential btit it does imply that we have always had leaders. How, then, can we establish whether leadership is crucial or whether the forms and styles of leadership have changed across space and time? We have an inordinate amount of information on 'management', as constituted in the last 200 years, and we have an increasing pool of knowledge about contemporary leadership, but what do we know about leadership in classical times? To a very large extent our knowledge of leadership in ancient times is crucially dependent upon the existence of written texts, and here lies the first lesson of leadership: history is written, generally speaking, by the winners. This goes for both successful military leaders and for successful political groups. In the former category we might consider how we know so much about Alexander the Great's or Julius Caesar's victories, so little about Spartacus, and almost nothing about the hundreds of other slave revolts that regularly shook slave societies throughout antiquity (Grint. 2005; Wiedemann, 1995). The answer, of course, is that Alexander and Julius Caesar either wrote their own histories or had them professionally written on their behest, while Spartacus left no written accounts and very few other slave leaders even get a mention in the accounts of their slave owners. Thus, a preliminary warning in reading any account of classical leadership - and indeed any account of contemporary leadership - is to be wary of the sources. Accounts are not neutral carriers of factual information; rather, they are partial accounts intended to achieve a particular purpose. Horalio Nelson, for example, was known Lo have 4 THE SAGE HANDBOOK OF LEADERSHIP A HISTORY OF LEADERSHIP 5 written his own account of the various battles he engaged in, and then have his subordinate officers attach their names to the account and send them off to the newspapers and the British Admiralty as if they were written hy these subordinates (Grint, 2000: p. 247). If you want to increase- your chances of appearing a good leader in the media, write the story yourself. Whether that story ever gets written in the first place depends, to some extent, on whether the narrative contains something regarded as significant: that is to say, we tend to record only those events that are unusual or extraordinary to some degree. As a consequence we do not have vast tracts on how to run a small farm in China 2,000 years ago nor on leadership in an era of peace amongst the Celtic tribes of Gaul at the same time. But we do have records of the Celtic wars against the Romans at the tune and we do have some accounts of Chinese warlords in the same period. However, the texts relating to the wars between the Gauls and the Romans are Roman texts: first, because the Celts were largely a non-literate society where oral cultures prevailed and second, because, by and large, the Romans were victorious. Again, what tend to survive over long periods of time are materia] texts and artefacts rather than oral narratives, so our understanding of the leadership of non-literate societies is often reconstructed from the often pejorative accounts of others. From what we know of preliterate ancient civilizations from the archaeological records, any periods of peaceful coexistence with neighbouring tribes led by humanitarian leaders are few and far between: as Keeley (1996, p. 174) suggests, 'Peaceful pre-state societies were very rare; warfare between them was very frequent, and most adult men in such groups saw combat repeatedly in a lifetime.' It should already be clear that war is a critical component in the early developments in the practice of leadership. From Sargon of Akkad (c. 2334-2279 ec) in what is now the Middle East, to Ramesses 11 (Ramesses the Great) of Egypt and from the early Cretan civilizations from around 3000 Be to the Harappan civilization in the Indus valley at the same time, and across to the Huang Ho walled settlements in China, we know that military leadership played a crucial role in the quest for survival and domination. Again, this is not to insist that leadership has its origins in war or that military leadership is the most important element in classical leadership - we simply do not know enough about these times to confirm or deny this. But it remains the case that some of the most important classical writings on leadership pertain either to the conduct of war or the conduct of politics: what Clausewitz referred to as 'the continuation of war by other means.' Tiiis is particularly so for the Classical and Renaissance periods that we shall consider first, before turning to the more modern literature. CLASSICAL LEADERSHIP STUDIES Outside Europe, Kautilya's The Artlirasastra, written around 321 nc for the Mauryan dynasty in what is now India, provided an array of practical tips for leaders to consider. But probably the first prescriptive text that achieved significant success in both its own time and space - ancient China -and continues to beguile business executives to this day is Sun Tzu's (400-320 bc) The An of War. This became mandatory reading for the military leaders not just of China but of Japan under the Taiho Code of 701 (Farris, 1999, pp. 52-53), where, as in many non-Western societies, war in general and military leadership in particular became tightly incorporated into the state's governance systems. In contrast, Western approaches tended to maintain a markedly higher degree of decentralization and independence from the central authorities (Hanson and Strauss, 1999, pp. 446-448). In fact, it is not clear who the author of the aphorisms that comprise The Art of War really is, and it may be that many were written by Sun Tzu's disciples and students; indeed, the text reproduces this assumption is its conversational format with several characters participating in the discussion under 'Master Sun's' facilitation. Nevertheless, the central message about leadership - though obscure in parts - is clear: 'The responsibility for a martial host of a million men lies in one man. He is the trigger of its spirit' (Manoeuvre 20). Once this is established. The Art of War sets out to provide conversational sketches of the most crucial elements of strategy and tactics for military leaders. Ironically, to Western minds, but appropriately for the minimalist essence of its Taoist origins, one of the most important lessons in The Art of War is that fighting is the last tiling military leaders should engage in, for: 'those who win every battle are not really skilful - those who render others* armies helpless without fighting are the best of all' ('Planning A Siege'). Sun Tzu then insists that strategy is critical to success, for the art of war is the art of avoiding unnecessary colli lict. 'The Golden Bridge' is a natural consequence of this philosophy: if you must fight then avoid head-on conflicts if at all possible, since these are both expensive in resources and casualties and are far riskier than simply attacking the enemy's plans or supply lines. And if you must attack the enemy head on — but you cannot be confident about a complete rout - then you should leave a 'golden bridge', an escape route for your enemy to retreat across, otherwise your enemy will be forced to fight to the finish, and again the consequences could be problematic. For those involved in leading negotiations, this is clearly an important piece of advice: if you cannot allow your opponent to leave the negotiating table with something they value, you will find it hard to lead them to a deal. A second and paradoxical piece of advice is to burn your own bridges: in other words, commit yourself or suffer the penalty. 'When a leader establishes a goal with the troops', suggests Sun Tzu, 'he is like one who climbs to a high place and then tosses away the ladder'. This is an inversion of the golden bridge rule, but that is for your enemies not your allies and followers, for if your colleagues feel threatened but see an easy escape route they may well take it. If, however, there is no escape — what Sun Tzu refers to as 'Dead Ground' - then they will have to commit themselves to the fight for survival, and it is this commitment by followers to then leader that again reflects the Taoist roots of Snn Tzu's work. As he puts it, in 'Nine Grounds': 'Put them in a spot where they have no place to go and they will die before fleeing.' This might be a useful phrase for all leaders - if their own survival or fortunes are tied into the organizations they lead, they might find it much harder to fail and walk away with a golden parachute. Sun Tzu is also adamant that military matters should be left to the military specialist and not to their political controllers: 'To say that a genera] must await commands of the sovereign in such circumstances is like informing a superior that you wish to put out a fire' ('Offensive Strategy'). Or as is suggested in 'The Nine Variables', 'There are occasions when the commands of the sovereign need not be obeyed.... When you see the correct course act; do not wait for orders.' Now there's a radical thought - not waiting for permission for something that obviously needs doing! At roughly the same time that Sun Tzu was teaching military leadership in China, Plato (7429-347 bc) was warning the Greeks that the rise of political leadership rooted in democracy did not represent the flowering of Greek culture so much as a direct threat to Greek civilization. The electoral system for selecting leaders generated a circus rather than a forum for serious consideration, as far as Plato was concerned, for it encouraged potential leaders to pander to the bases! instincts of the mob - 'the large and dangerous animal' - that pervades much of his writing in this sphere. The mob, suggests Plato, in his Republic, would be willing to risk their society (represented as a ship) by electing whichever person promised them most. Thus, rather than sailing under the person who was best qualified to be the ship's captain (one of Plato's Philosopher Kings), democracy ensures that the popular demagogue prevails - and, of necessity, leads the ship straight onto the rocks of catastrophe. This, of course, legitimizes the leadership form used in business, where democracy is absent; though financial leadership by people best qualified to lead us (financial experts) seems to have steered us straight onto, rather than away from, the rocks of catastrophe, and only the elected political leaders seem to have saved us. But how is the best person to lead recognized? For Plato it is self-evident that we recognize the skills of people by considering their expertise: we would not ask a gardener to build us a boat any more than we would ask a farmer to run the economy. But, to Plato's intense frustration, where 'moral' knowledge is concerned, the mob assumes that everyone is an expert and therefore no-one is. The result is that the mob assumes they can recognize good political leadership when they see and hear it and are keen to put themselves forward for office, even though they would not dream of building a ship unless they were a shipbuilder. It was for precisely this reason that Plato was so firmly opposed to the Sophists and Isocrates who taught the skills of rhetoric, or public speaking, because this would simply encourage the domination of form over content. Above all, Plato feared that even those who intended to lead in a moral way for the benefit of the community would be corrupted by the system and, since leaders were vital to the health of the community, a corrupted leader would inevitably destroy 'his' own community. Aristotle (384-322 bc), one of Plato's students, agreed that Athens was indeed under attack from corrupt leaders but differed in his response to the problem. His book Rhetorica was written in part as an expose of 'the tricks of public speaking', which Aristotle believed were already corrupting Athenian public life. In its earliest form 'rhetoric', from the Greek pm/cmp (rhm lor) (speaker in the assembly), was the art of using speech to persuade. It is not coincidental that the art of rhetoric and the rule of democracy evolved simultaneously. As Lawson-Tancred (1991, p. 3) suggests, when political rule is through naked force, or inherited tradition, there is little need to persuade the people of one's right to rule, though in fact Ancient Greek society held oratorical skills almost as high as military prowess. The origins of rhetorical skill appear to lie in Syracuse, a Greek colonial city, and the skill moved rapidly throughout Greek society where the political and the legal system depended upon rhetorical skill. By the time of the Peloponnesian War (431-404 bc), professional teachers of rhetoric and speech writers (logographoi) had appeared, and training in THE SAGE HANDBOOK OF LEADERSHIP rhetoric became commonplace, especially for potential leaders. Classical theoreticians (Georgias, Plato, Aristotle and Cicero) were very divided in their debates about the nature of rhetoric, in particular whether it was an essentia] part of life and leadership or a sleight of hand (Wardy, 1996). Georgias had originally argued that rhetoric, the art of persuasion ill which emotion and power were influential features, was an inevitable element of all human life. For Plato, the use of rhetoric — as an act of manipulative persuasion - was both inferior to, and subversive of, philosophy, in particular the dialectical questioning method used by Socrates to establish truth through reason and rational debate. Plato was vehemently opposed to democracy and regarded the teaching of rhetoric as a dangerous malpractice: it was the mischievous tool of the demagogue - it enslaved the masses and it pleased rather than benefited the mob, for it provided even the worst political leaders with the technical skill to manipulate the masses in whatever direction they chose (Grint, 1997a; Wardy, 1996, p. 81). For Plato, the direction was crucial and rhetorical skill should aiways be subordinated to the direction, and not vice versa. But for Georgias, the Athenian right to free speech and the rule of democracy (for male citizens at least) made rhetoric a principle skill for all to master; however, it was a skill that could be used for good or evil as it did not embody any values in and of itself. Of course, heroic deeds were also important, but without their recording in rhetoric they were soon lost to history. Furthermore, Plato's representation of Socrates suggests that the teacher of philosophical dialectic did not pretend to know - as he suggests rhetoricians did - the answers, but merely taught the techniques for understanding, to attain knowledge for its own sake, and increase one's own knowledge in the dialectical process with the student (Wardy, 1996, pp. 54-70.) Whatever the results of the battle in any theoretical sense, in the hands of Aristotle, the construction of rules for rhetoric became formalized, and these rules changed little between Cicero and the late eighteenth, early nineteenth, centuries when the formal teaching of rhetoric as a university subject generally fell from favour. Aristotle had four explanations for this apparent betrayal of bis teacher: First, he still maintained that since truth and justice were stronger than lies and injustice, [hen 'false' rhetoric would not be able to overturn the former. Truth and justice, then, had a 'natural advantage' (Wardy, 1996, p. 110). Secondly, his arch opponent and intellectual competitor, Isocrates. had developed a very successful school based on the study of rhetoric - which shaded Aristotle's school with its pulling power -and, with the decline of Athenian democracy, the utility of persuasion look on a different meaning. For Aristotle, the reality that rhetoric was already being taught meant that he now had a duty to educate people about the tricks that rhetoricians would use lopeTsuade them of falsehoods. Thirdly, Aristotle argued thai rhetoric could be studied, but only in the context of its philosophical foundations and formations. Rhetoric, at the hands of Aristotle, was to become not [he pragmatic bag of tricks he associated with its Isocratic version, but with the scientific roots of knowledge that Aristotle inherited from Plato and Socrates. Fourthly, although Plato was adamant that rhetoric persuaded through emotion not reason - and hence his dislike of it— Aristotle suggested that one could appeal to the listeners' emotions - if, but only if, the intent and effect was to enable them to see the rationality of the argument. Unfortunately for Aristotle, his efforts had a limited effect in their time, not through any lack of skill on his part but because within Five years of his death the great Athenian experiment in democracy had been plunged into darkness with the rise of the tyrant Demetrius (an ex-pupil of Aristotle), and with the subsequent restructuring of the legal system the requirements for rhetorical skill were marginal at best. Yet, the last 100 year's have demonstrated without parallel that the ability to persuade people through rhetoric has altered the world beyond our wildest dreams or nightmares: would Hitler or Churchill or Obama have made such an impact without this ancient skill? RENAISSANCE LEADERSHIP STUDIES One thousand and eight hundred years after Aristotle there emerged from that same area of the Mediterranean a book that came to dominate prescriptive writing on leadership, not just in its own time but in our time too. Not that Machiavelli's The Prince was popular: on the contrary, it was the most unpopular prescriptive text of the sixteenth century. No doubt Macbiavelli would have found this doubly ironic: first, because The Prince was written to regain some political credibility and popularity with his former employers; secondly, because Machiavelli wrote it as descriptive, rather than a prescriptive, work. In other words, Machiavelli insisted that he wrote about the world of politics as it was, not as it should be in some mystical and unachievable Utopia. It was the political realism that infused The Prince that led to its instant condemnation by the religious and political leaders of the day but which also explains its popularity today (see, for example, Ledeen, 1999; McAlpine, 1997). It was, according to Machiavelli, rooted not in theory but in historical A HISTORY OF LEADERSHIP fact, yet it was prohibited by the Catholic Church under its Index of Books. It was nothing but 'A handbook for gangsters' according to Bertrand Russell, thougli Napoleon was rather more positive, suggesting it was The only book worth reading.' The Prince was written in 1513-1514, probably after Machiavelli started writing his Discourses on Livy, though the latter was not published until 1531, while the former was published the following year. As his homeland fell apart under civil war and foreign invasion, Machiavelli sought to write a guidebook for all political leaders but, in particular, for the Medicis, his patrons and erstwhile leaders of Florence. The Prince, then, was not simply a book to ingratiate the favour of the Medicis, but a call to arms to defend Florence and - through Florentine domination - Italy, from the 'Barbarians', by whom he meant the Spanish and French invaders. Machiavelli had lived through the golden age of Florence, but also through some of the excesses, invoked by critics like the Dominican preacher Girolamo Savonarola who encouraged the Florentines to burn their worldly goods - their books, pictures and jewels — in a 'pyre of vanities.' Eventually, Savonarola was himself burned at the stake, but it taught Machiavelli a lesson about the vulnerability of the 'unarmed prophet'. In 1498, Machiavelli was appointed to a senior position within the Florentine civil service as Secretary to the Ten, the committee concerned with foreign and military policy. However, his attempt to organize the recovery of Pisa using mercenary troops failed miserably and his antagonism to them - and desire to displace them with a citizens' army -played a significant part in regaining Pisa in 1509, and cemented Ms views in The Prince. Also important was his next mission, to fend off Cesare Borgia, the illegitimate son of Cardinal Rodrigo Borgia, who had become Pope Alexander VI in 1492, Cesare Borgia led the papal armies and threatened Florentine independence, but Machiavelli soon recognized a different category of leader in Cesare: for here was a man who murdered his own lieutenant (Remirro Oreo), when he appeared to be unnecessarily cruel in his control over the Romagna. As Machiavelli recalled, '... one morning Remirro's body was found cut in two pieces on the piazza at Cesena, with a block of wood and a bloody knife besides it. The brutality of this spectacle kept the people of the Romagna at once appeased and stupified.' (Vll). Cesare subsequently invited those conspiring against him to dinner, only to have them all slaughtered as they ate. Machiavelli, then, used Cesare as a good example of the real polhik of life; for he believed Cesare had restored peace through the selective use of violence. The alternative, as professed in public by most leaders at the time, was to act nobly and morally, but for Machiavelli the consequence of acting morally in an immoral world was simply to allow the most immoral to dominate. 'The fact is,' he suggests in The Prince, 'that a man who wants to act virtuously in every way necessarily comes to grief among so many who are not virtuous. Therefore, if a prince wants to maintain his rule he must learn how not to be virtuous, and to make use of this or not according to need.' (XV) Cesare Borgia was accounted cruel; nevertheless, this cruelty of his reformed the Romagna, brooght it unity and restored order and obedience. On reflection, it will be seen that there was more compassion in Cesare than in the Florentine people who, to escape being called cruel, allowed Pistoia to be devastated.,,(XVII) In effect, Machiavelli — and Machiavellians -are not necessarily suggesting that leaders should act immorally, but that to protect the interests of a community a prince has to do whatever is necessary — for the greater good. Thus, the act should be contextualized and not analysed against some mythical moral world. The problem, of course, is defining 'the greater good' and that problem continues to plague us. And, in answer to his rhetorical queslion 'Whether it is better to be loved or feared, or the reverse', Machiavelli unequivocally sides with the fear factor. The answer is that one would like to be the one and the other, but because it is difficult to combine them, it is far better to be feared than loved if you cannot be both. One can make this generalization about men; they are ungrateful, fickle, liars and deceivers... when you are in danger they turn against you. Any prince who has come to depend entirely on promises and has taken no other precautions ensures his own ruin.... The bond of love is one which men... break when it is to their advantage to do so; but fear is strengthened by a dread of punishment which is always effective. The prince must nonetheless make himself feared in such a way that, if he is not loved, at least he escapes being hated. (XVII) Mrs Thatcher might well have taken Machiavelli's advice here, for although she never did court popularity in the sense of wanting to be liked (often regarded as a fatal flaw in some accounts of leadership), she did generate so many political enemies within her own political party that eventually she ran out of allies and had no more punishments to hand out. As far as Machiavelli was concerned, his ideas amounted to no more than disclosing what was ! : SAGE HANDBOOK OF LEADERSHIP already happening rather lhari advocating something radically new. However, he has usually been interpreted - and certainly was at the time -as suggesting that leaders should always act immorally and subordinate the means to the end, since this would generate an advantage over others. But he actually insists, more clearly in The Discourses than in Tlie Prince, that this can only be legitimated by reference to the greater good of the community, and not to the benefit of the individual . Of course, how this greater good is defined is a moot question, but in The Discourses Machiavelli sets himself firmly on the side of the republic and the populace (though he was not a democrat) and against princes and those who led for their own self-interest (especially the 'idle' aristocracy). But he also suggests that there are times when the ruthlessness of a prince is necessary to restore a society back to health, something which republics find difficult to do. Thus, in the long term (and for Machiavelli the long term could only exist within an expansionist state), a republic is preferable, but an occasional prince may be a necessary evil. We might extend this idea to suggest that sometimes (in a crisis) leaders need to be ruthless commanders, but other times (facing a Wicked Problem) they need to exhibit the more collaborative style conducive to engaging disparate interests (Grint, 2005). Machiavelli drew many of his examples of leadership from the classical period of Rome and Ancient Greece and suggested that historical patterns and lessons of the kind we began with were clear for all to see. fn the next section I consider modern leadership studies, not in any detail (that is well covered in other chapters of this book), but sufficiently to establish whether such historical leadership patterns can be discerned at all. MODERN LEADERSHIP STUDIES Ironically, Thomas Carlyle - for many the first 'modern' writer on leadership - had spoken warmly at his inaugural address as Rector of Edinburgh University in 1866 of both Machiavelli and Oliver Cromwell. The latter Carlyle likened to a Machiavellian prince who was absolutely necessary at the time of the English Civil War. The irony lies in both Cromwell's refusal to take the crown after the execution of Charles I and in Cromwell's delivery of the only period in English history when England was a republic. In fact we can trace the rise of leadership studies to the modem era — that is the rise of industrial societies — to then. In the beginning was perhaps not God, but rather the god-like creatures that peppered the 1840 lectures of Thomas Carlyle, whose fascination with the 'Great Men' of history effectively reduced the role of mere mortals to 'extras'. Despite Carlyle's dislike of the early industrial entrepreneurs of Britain - die 'millocracy' as he called them - the model of individual heroism that he constructed personified a popular assumption about leadership in Victorian times: it was irredeemably masculine, heroic, individualist and normative in orientation and nature. That model seems to have prevailed throughout the latter half of She nineteenth century and was not really challenged until the first professional managerial group began displacing the original -and 'heroic' - owiter-manageTS towards the end of the nineteenth century. Then, the argument runs, the context - and thus the requirement - for leadership shifted from heroic individuals to rational systems and processes, as the scale of industry and the level of backward integration began generating huge industries (especially in the United States) that needed significant numbers of administrators to retain organizational coherence. Many of the models for such organizational leadership were derived from the army, civil service, post office and railroads and most constituted leadership as administrative positions within formal hierarchies. In turn, as the productive growth unleashed by these giants began to encourage significant market competition and eat into profit margins, attention quickly turned to cost-reduction strategies and to Scientific Management. F.W. Taylor concentrated on the control of knowledge by management at the expense of the workforce and (he deskilling of jobs in line with the expansion of the division of labour. In this case, leadership was configured as knowledge leadership, with the leaders as repositories of knowledge of production that generated power over production - in contrast to the control over production, formerly wielded by craft workers. The economic depression of the 1920s coincided with the next major shift in leadership models and, for our purposes, it was a major shift back to the role of normative power and away from the rationality of scientific systems and processes that had dominated for the previous two decades. This 'return' to a previous normative model was derived initially from the Hawthorne experiments in the t920s and 1930s at the General Electric (GE) plant near Chicago. There, Taylorist scientific experiments in the development of the optimum environmental working conditions had allegedly generated first perplexity and then a realization that work could not be measured objectively because the very act of measurement altered Ihe experience and thus the behaviour of those being measured. This 'Hawthorne Effect', as it was called then, spawned a whole series of related experiments that eventually persuaded first A HISTORY OF LEADERSHIP GE and then whole swathes of American management thai workers were normatively not rationally motivated and group- not individually-oriented in culture. Arguably, these alternating or dualist models of leadership — first the 'normative' model of Carlyle of the second half of the nineteenth century, followed by the 'rational' model of Taylor and Ford in the first two decades of the twentieth century, in turn superseded by a return to the 'normative' model of the Hawthorn experiments that solidified into the Human Relations approach of the 1930s and 1940s — reflect two broader phenomena: first, the economic cycles of the period; and, second, the political models of the period. The economic cycles formed the basis of Kondraliev's controversial theory of Long Economic Waves or Cycles; the political cycles are less controversial and more intriguing, for it seems unlikely that industry could have isolated itself from the global rise of the mass movements of communism and fascism in the late 1920s and 1930s and more likely that the leadership models embodied in these were refracted in industry through a Zeitgeist that made sense at the time. In other words, in an era when mass political movements driven by normative adherence to the collective will - but manifest in cull-like loyalty to the party leader - were so prominent, it seemed perfectly natural to assume that the best way to lead an industrial organization was to mirror this assumption: work should be normatively rather than rationally organized - by groups led by leaders who prototypically embody the same apparent desires as those held by the mass. By the time the Second World War was over, and the economic boom returned, the model that began to dominate in the West shifted once again from the normative cult of mass and heroes - that had reflected the power of communism and fascism - to one dominated by rational analysis of the situation - a scientific approach more conducive to the war-fighting capabilities of the preeminent victor, Ihe United States, and one located within its individualist culture. Thus we see the rise of the American self-actualization movement, manifest particularly in Maslow's (1954) 'Hierarchy of Needs' and in McGregor's (1960) displacement of Theory X with Theory Y. The movement away from norms and back towards the rational understanding of contexts followed the critiques of traits by Jenkins (1947) and Slogdill (1948, 1974), as well as the work of the University of Michigan and the Ohio State studies. These provided Ihe framework for a radical development: contingency theory. Under the general umbrella provided by Fred Fiddler's (1964) contingency theory, and Robert Blake and Jane Mouton's (1964) Managerial Grid, the theoretical fragility of relying upon a potentially endless list of traits and superhuman charismatics was, ostensibly, dealt a crippling blow. From then on what really mattered was not having the most charismatic leader, leading the most adoring mass of followers, but having a rational understanding of the situation and responding appropriately. These leadership theories that eschewed the dominant and proactive role of the individual leader in favour of more social or structural accounts tended to assume that the context or situation should determine how leaders respond; thus, in terms of the early contingency theories, situation X requires leadership X to ensure an appropriate - and thus rational - response. More recent developments in contingency theory, for all their more sophisticated accumulation of significant and independent variables, still reproduce this assumption that the 'correct' response is determined by the correct analysis of the situation. Since the early days of this contingency approach, we have 'progressed' by reluniing to the importance of leaders working with the (normative) 'strong cultures' beloved of Peters and Waterman (1982), then on to the (rational) pedagogy of the Reengineeriiig revolution of the 1990s, and finally on to the contemporary development of transformational and inspirational leadership theories silhouetted by the rise of terrorism, global warming and political and religious fundamentalism. Such transformations also invoke the New Public Management, since the 1980s, under which the public sector was ostensibly transformed from a lethargic and bureaucratic leviathan to an agile service deliverer through the encroachment of the market and the discipline of targets and performance management systems. Coupled with concerns about the importance of emotional intelligence, identity leadership, and the development of inspiring visions and missions, this seems to have ensured the return of the original normative trait approaches: we seem to have gone forward to the past. Thus, we were recently (back) in thrall to inspirational individuals, endowed with what list of essential competencies the contemporary leaders happen to have that are adjudged to be responsible for the results. That the competencies are dccontextualized, ahislorical and, at best, only correlated with, rather than determinants of, success, seems to be irrelevant. Indeed, we appear to have an amazing capacity lo attribute organizational success to individual competence on the basis of virtually no evidence at all. Many political commentators describe the global political situation in similar terms - dominated by the role of individual politicians. Others simply report that politicians are the world's least trusted people, again implying that it is the individual leader that matters not the situation. Yet there is considerable support for the 10 THE SAGE HANDBOOK OF LEADERSHIP r TIME 2000+ Distributed Leadership Foilowership Identity Mission-CommanrJ 1990s New Public Management BPR (Business Process Reengineering) Competencies Benchmarking Targets, Psychometrics 197Os-80s Corporate Culture Quality Circles TQM (Total Quality Management) Delayering 1950S-60S Contingency Theory, Systems Analysis Self-Actualization: Maslow & McGregor 1930-40S Hawthorne/Human Relations Mass + leadership traits + charismatics 1910-20S Taylorism & Fordism Scientific Management 1900s Rule of thumb Great Man TIME A HISTORY OF LEADERSHIP 1990-2000S New Public Management BPR (Business Process Reengineering) Competencies Benchmarking Targels Psychometrics 1900-50 Taylorism & Fordism Scientific Management Hawthorne/Human Relations Mass + leadership traits + charismatics Theory X 1.1 2000+ Distributed Leadership Followership Identity Mission-Command 1950-1980s Contingency Theory, Systems Analysis Corporate Culture Qualify Circles TQM (Total Quality Management) Delayering Decentralization Maslow & McGregor 1850-1900 Rule of thumb Great Man Figure 1.2 Binary model A: centralization-decentralization. RATIONALITY Figure 1.1 Increasingly rational leadership over time. latter approach, within which even if individual leaders do make a difference, that difference is only marginal in comparison to the influence of more structural features like the economy or religion or political party or social class or gender or any other of the myriad variables on offer. The current leadership fashion also manifests itself in some form of distributed leadership: we now need a collective approach to decision-making to counter the romance of (individual) leadership and to better cope with an increasingly complex world. Much of this work actually relates to specific contexts - either education, where professionalism is high, or the military, whose Mission-Command doctrine of highly decentralized operational leadership, combined with centralized strategic leadership, is designed to cope with the shift from conventional war-fighting to asymmetric conflict and peace keeping where decentralized decision-making is a prerequisite of success. Yet Mission-Command has been practised by the German army since the nineteenth century. It may be, then, that assumptions about the necessity of rigid command and control systems under all (military) circumstances have always been as dubious as assumptions that suggest the opposite. We might, therefore, be in as much danger of pursuing the romance of collaborative leadership (Leonard, 2007) as we previously were of pursuing the romance of individual leadership. PATTERNS OF LEADERSHIP The argument for these dualist shifts between forms of leadership is not universally accepted. Indeed, there are many ways to understand this pattern - if indeed there is a pattern. First, what we have is simply an increasingly sophisticated and rational approach to leadership across time represented by the incremental enhancements manifest in Figure 1.1. Students of history will recognize this as a Whig variant on progress across time. Of course there will be setbacks and hiccups along the way, but the future is eternally rosy and will be preferable to the past. Alternatively, there are two binary models that suggest a rather different explanation for change. Figure 1.2 suggests that the pattern is represented by a pendulum swinging between centralized and decentralized models of leadership - usually premised on assumptions about organizational learning and game playing so that what was once efficient becomes inefficient as institutional sclerosis sets in. Figure 1.3, on the other hand, retains the binary model but the causal mechanism relates to the structural binaries that constitute language: night/day, black/white, dead/alive and so forth. Here if is the relationship between science and culture as the binary pairing of linguistic oppo-sites which provides the natural barriers to change and once the efficiencies of one leadership style arc expended the pendulum swings in the opposite I i THE SAGE HANDBOOK OF LEADERSHIP TIME 3 1990s BPR, Psychometrics Competencies, Emotional Intelligence, Benchmarking Targets New Public Management 1950S-1960s Contingency Theory, Systems Analysis Self-Actualization: Maslow & McGregor 1010-20$ Taylorism a Scientific Management Fordism Figure 1.3 Binary model B: science vs culture. 20OO+ Distributed Leadership Foliowership Identity Mission-Command 1980s Corporate Culture Transformational Quality Circles TQM (Total Quality Management) Delayering 1930^0s Hawthorne/Human Relations Mass + leadership traits + charismatics 1900s Rule of thumb Great Man A HISTORY OF LEADERSHIP 13 Political Everts Terrorism, Uncertainty, Fundamentalism Rise of Bush & Blair Globalization & Rise of Japan, post Cold War, Thatcherism US domination & Cold War 2000+ 1990-2000 198O-90S 195< >-70s Leadership Model Distributed Leadership, Foliowership, Identity, Mission-Command New Public Management, BPR (Business Process Reengineering), Competencies, Benchmarking Targets, Psychometrics Transformational Leadership, Corporate Culture, Quality Circles, Tough Love Contingency Theory, Systems Analysis, Self-Actualization: Maslow & McGregor linguistic direction until that mode is also exhausted. Fourthly, a political model situates the changes not against the binary limits of language but rather the political machinations of the wider context, in which what seems 'normal' only appears so when framed by the political ideologies of the day. This approach, represented in Figure 1.4, is effectively locked into, and explained by, the Zeitgeist of the day: 'the spirit of the times'. Thus, Taylorism emerged as the norm, not because it was scientific and therefore rational, but because in an era where scientific breakthroughs were changing the world of work and where the eugenics movements started to dominate American culture, it seemed natural to assume that there was one best way to allocate, control and lead labour, and that allocation should be based on 'natural' differences that should be exploited for the benefit of all. In effect, as Taylor argued, if you could design jobs requiring virtually no intelligent action on the part of the worker, and where no communal interest prevailed, then you could minimize the cost of labour, the rate of error, and the possibilities of trade unionism. Similarly, when communism and fascism began to ensnare European politics, it probably seemed inevitable that the best way to lead was not through scientific management of the individual but through the manipulation of the emotional mass by the charismatic leader. Once the Second World War was over, the dominant power - the United States - reproduced its own scientific and individualist approach as the default leadership model and only when the threat from Japan (re)emerged in the 1980s was there a significant shift towards a more cultural approach to leadership. Once again, when these ran out of steam in the 1990s, the move was back towards the scientific end of the spectrum, as New Public Management and target measurement took over to be toppled only now as the spread of moral panics about global warming, terrorism, swine flu, the credit crunch and so on shift the debate back towards the cultural school. Finally, of course, it might well be that there is no pattern at all in the data itself as an object fact, but rather the patterns are more likely to be the consequence of the prior assumptions and cultures of the analyst. In effect, the more scientifically inclined amongst us might be inclined to see greater rationality in leadership styles across time; the more liberal amongst us might see the spread of collaborative styles as proof positive of their deeply held antipalhy to individual leadership manifest in heroic men; the more cynical amongst us might perceive none of these patterns but just an accumulation of historical detritus strewn around by academics and consultants hoping, at most, to make sense of a senseless shape or, at least, to make a living from constructing patterns to sell. Now there's a thought.... CONCLUSIONS Attempting to cover 3.000 years of history in less than 8,000 words or so inevitably requires a significant degree of omission, but what might the critical poults of this review be? Depression, Communism, Fascism Hawthorne/Human Relations, Mass + traits -charismatics Global Competition, WWi 1910-20S Industrialization -1800s TIME Figure 1.4 Political Zeitgeist. First, that while situations are always changing, there are perennial issues in leadership and our historical knowledge of these can play a role in preventing the repetition of mistakes, even if we cannot guarantee success. So perhaps one of the reasons history does not repeat itself precisely is because we have some capacity to iearn from our mistakes and stop it repeating itself. Secondly, what we think leadership is, is necessarily related to the cultural mores that prevail at the time. Thus, what appears 'normal' at the time - for example, using targets to ensure compliance with the requirement of political leaders -can often appear extraordinarily naive when considered retrospectively. The problem, of course, is that we cannot step outside our own Scientific Management, Taylorism & Fordism 1900s: Rule of thumb, Great Man milieu to reflect upon ourselves as disembodied and disinterested scholars. This, surely, is where the study of history comes into its own: not in analysing ourselves against an objective system of measurements, but in recognizing that we have been heading somewhere similar before - and if we do not take certain kinds of action we might end up in a very similar place, Thirdly, that an awareness of how often in the past leadership has failed or appeared terminally flawed does not necessarily mean that we should abandon the task of generating leadership for the public good; it just means that we need to be more alert to what is likely to happen unless we actively prevent it. That includes recognizing that flawed leadership is part of the human condition, that ■ 14 THE SAGE HANDBOOK OF LEADERSHIP there are no perfect leaders or perfect leadership systems out there to be imitated and, to quote an old Hopi Indian saying, 'We are the ones we have been waiting for.' REFERENCES Blake, R.R. and Mouton, J.S. (1964) The Managerial Grid (Houston: Gulf). Farris, W.W. (1999) 'Japan to 1300'. In K, Raaflaub and N. Rosenstein (eds), War and Society in the Ancient and Medieval Worlds (London: Harvard University Press). Fieldler, F,E. (1964) A Theory of Leader Effectiveness (New York: McGraw-Hill). Grint, K. (1997a) Fuzzy Management (Oxford: Oxford University Press), Grint, K. (2000) The Arts of Leadership (Oxford: Oxford University Press). Grint, K. (2005) 'Problems, Problems, Problems: The Social Construction of Leadership', Human Relations, 58(11); 1467-1494. Hanson, V.D. and Strauss, B.S. (1999) 'Epilogue'. In K. Raaflaub, and N. Rosenstein (eds), War and Society in the Ancient and Medieval Worlds (London: Harvard University Press), Jenkins, W.O. (1947) 'A review of leadership studies with particular reference to military problems'. Psychological Bulletin. 44: 54-79. Keeley, LH. (1996) War Before Civilization (Oxford: Oxford University Press), lawson-Tancred-H. (1991) 'Introduction'. In Aristotle's The Art of Rhetoric (Harmondsworth: Penguin). Ledeen, fvl.A. (1999) Machiavelli on Modern Leadership (New York: Truman Talley Books). Leonard, P. (2007) 'The Romance of Collaborative Leadership'. Paper given at the British Academy of Management Conference (Warwick University). McAlpine, A. (1997) The New Machiavelli (London: Aurum). McGregor, D. (1960) The Human Side of Enterprise (New York: McGraw-Hill). Maslow, A. (1954) Motivation and Personality (New York: Harper). Peters, T. and Waterman, R.H. (1982) In Search of Excellence (London: Harper & Row). Santayana, G, (1954) The Life of Reason or The Phases of Human Progress (New York: Charles Scribner's Sons). Stogdill, R.M. (1948) 'Personal Factors Associated with Leadership: A Survey of the Literature', Journal of Psychology, 25, 35-71. Stogdill, R.M. (1974) Handbook of Leadership: A Survey of Theory and Research (New York: Free Press) Wardy, R. (1996) The Birth of Rhetoric: Georgias, Plato and Their Successors (London: Routledge), Wiedemann, T.E.J. (1995) Emperors and Gladiators (London: Routledge). Research Methods in the Study of Leadership Alan Bryman INTRODUCTION In this chapter, I intend to review and comment upon the range of research methods that are employed by leadership researchers. This will require a consideration of broader issues such as epislemology that abut and relate to discussions of research methodology. I am by no means the first writer to review the methods employed by leadership researchers and, early on in this article, other reviews will be examined. Leadership researchers employ a wide range of research methods, although the d iversity of those methods is of relatively recent origin since the field lias largely been dominated by a single method of data collection - the self-compietion/self-adiTiims-tered questionnaire. As Avolio et al. have observed: 'The quantitative strategies for studying leadership have dominated the literature over the past 100 years' (2009a; p. 442). The self-administered questionnaire has been one of the most commonly employed tools among the quantitative strategies to which Avolio et ai. refer and has been the instrument employed in traditions that have defined the field of leadership research over the years; for example, the Ohio State tradition. When Lowe and Gardner (2000) reviewed the research methods employed in articles during the first 10 years of the life of The Leadership Quarterly, they found tiiat 64% employed a questionnaire-based approach. Thus, the leadership field is significantly reliant upon a single research method for its findings. It is not surprising, therefore, that Hunter et al. (2007) and Friedrich et al. (2009) refer to the questionnaire-based approach as the 'typical leadership study'. The dominance of this tool and other quantitative instruments reflects the wider epistemological orientation of many leadership researchers in that it exemplifies the commitment to a natural science mode! of the research process and to positivism in particular. This predilection is also significant for its consequences, because it has implications for the evaluation of research, especially research employing qualitative methods. It has also meant that the field of leadership research is a relatively late adopter of such methods (Bryman, 2004), although it is probably not significantly later in this regard than the oilier areas of organization studies and organizational behaviour-. However, it could be argued that the account of the typical leadership study provided in the previous paragraph represents a North American view of the methodological slate of die field. Thus, when David Collmson and Keith Grint, both co-editors of this volume, started the Europe-based journal Leadership, they outlined in their opening editorial that among their aims was that of increasing the field's methodological diversity. They wrote: In our view the understanding of leadership is best enhanced by the encouragement of a diversity of theoretical positions and research methods and the exploration of a great variety of research contexts and settings. Our vision is inclusive, not exclusive; one of radical heterogeneity, not simply a different form of homogeneity. (Collinson and Grint, 2005, p. 7, emphasis added) Thus, their mission for the new journal was not the eradication of the typical leadership article. 16 THE SAGE HANDBOOK OF LEADERSHIP RE5EARCH METHODS IN THE STUDY OF LEADERSHIP 1? but instilling greater diversity of methodological approach in the field. An analysis of all empirical articles published in Leadership in its first five years of publication revealed that it had a different methodological profile from The Leadership Quarterly, as gleaned from Lowe and Gardner's (2000) article (Bryman, 2011). Whereas Lowe and Gardner found 64% of articles to contain data derived from questionnaires, only 22% of articles in Leadership included data from this same method. By contrast, 51% of the articles in Leadership contained data derived from semi-structured interviewing. Lowe and Gardner did not have a special category for this method, but they do report that 20% of all articles included interview data. This latter figure is likely to include data from other kinds of interview (e.g. structured interview). Thus, there is a considerable difference between articles published in The Leadership Quarterly during its first decade and articles in Leadership in terms of the methods of data collection used. However, the data for The Leadership Quarterly relate to the period up to 1999. An analysis by Gardner et al. (Unpublished work) of articles published in The Leadership Quarterly in its second decade does not allow a comparison with the proportion of articles based on questionnaire data, although the authors do note that the typical empirical article is based on a single method - the survey. Further, the analysis revealed that quantitative research intensified its hold in the second decade when compared to the first decade's articles. PREVIOUS OVERVIEWS OF LEADERSHIP RESEARCH METHODS There have been a small number of overviews of the researcb methods employed in leadership research. Because there are different ways of classifying researcb approaches, these reviews have employed different categorizations. For example, Antonakis et al. (2004) began their review by distinguishing between quantitative and qualitative research methods for studying leadership but then note: 'Because the vast majority of research that is conducted in the leadership domain is quantitative in nature and because theory can be tested appropriately only with quantitative methods...' they focus upon the latter (Antonakis et al., 2004, p. 55). Within the category of quantitative research methods they distinguish the following: laboratory experiments; field experiments; field studies; and survey research. Essentially, this classification operates at the level of research design rather than research method - for example, the questionnaire can be (and indeed has been) used in connection with all four of these designs. Kroeck et al. (2004) also use the quantitative/ qualitative distinction as a starting point but have more to say about the nature of qualitative research and its underlying research methods. They distinguish between qualitative research based on ethnography, content analysis and observation. Under the umbrella of quantitative methods, they distinguish between paper-and-pencil questionnaire, computer-based questionnaire and simulation/ assessment centre. In contrast to Antonakis et al., these distinctions emphasize methods of data collection rather than research design but are not without problems. First, content analysis is regarded by many writers as a quantitative rattier than a qualitative method when it is implemented in the traditional fashion. More will be said about this issue below. Secondly, a similar* point may be made about the category of 'observation'. Kroeck et al, associate observation with the qualitative research technique of participant observation, but there are also structured forms of observation, such as that used by Lutbans and Lockwood (1984), which are typically viewed as associated with quantitative research (Bryman & Bell, 2007). A more recent contribution by Mumford et al, (2009) provides what is probably the most detailed overview by distinguishing between the following methods: survey studies of leader behaviour; attribute studies; experimental studies; qualitative studies; and historiometric studies. It is not obvious where observational studies of leadership would fit into this scheme and it could be argued that it is a classification based on two different things - leadership foci (e.g. behaviour and attributes) and diverse research designs and methods. However, it is a comprehensive approach and provides a reasonably balanced account of qualitative research. It is striking that these classifications often work with a distinction - either explicitly or implicitly - between quantitative and qualitative research. The quantitative/qualitative distinction is also implicit in the reviews of qualitative leadership research by some writers (e.g. Bryman, 2004; Klenke, 2008). Qualitative research methods tend to be grouped together and juxtaposed against several quantitative research methods. This is not unreasonable given the heavy reliance on the latter in leadership research, but a difficulty with the category 'qualitative research' in the leadership field (as well as elsewhere) is that it is very broad and as such does not fully take account of the fact that it relates to quite widely differing studies (Bryman, 2004). Examples of qualitative study might include: cross-sectional designs using qualitative interviews which can look very much like survey studies but without the numbers; single or multiple case studies using participant observation; single or multiple case studies using qualitative interviews; and single or multiple case studies using documents. Thus, the category 'qualitative research' obscures considerable variety in approach and its growing use in and contribution to the leadership field may require a retlunk concerning such issues as the evaluation of studies prior to publication. The classification of research methods in the leadership field presented in this chapter are summarized in Table 2.1. The listing is not exhaustive and it is doubtful whether any classification could be. Further, while the approaches are largely to do with research methods for the collection and/or analysis of data, experiments are strictly speaking a research design (Bryman and Bell, 2007), However, they are worth identifying separately for reasons that will hopefully become apparent during the discussion. It is also worth pointing out that single and multiple case studies will be discussed en passant as they are often linked to particular" research methods. Thus, ethnographies are almost always single or multiple case studies; also, semi-structured interview studies may be conducted either within a cross-sectional or a case study design. Some of these distinctions will be brought out in the discussion below. QUESTIONNAIRE STUDIES As previously noted, the use of the self administered or self-completion questionnaire in leadership research is pervasive. Questionnaire instruments like the Leader Behavior Description Questionnaire (LBDQ) used by the Ohio Stale leadership researchers, the Least-Preferred Co-worker (LPC) used to measure leadership orientations within Fiedler's Contingency Model, and the Multifactor Leadership Questionnaire (MLQ) used to measure transactional and transformational leadership are among the most intensively used and well-known scales in the field of organizational behaviour. Such instruments more or less define the field of leadership research. They have been widely used because their known psychometric qualities (their reliability and validity) have meant that researchers have been able to use them with an awareness of the advantages and limitations of doing so. The research streams associated with these questionnaires are covered in this volume in Chapters 17 and 22 by Diaz-Saenz and Yukl, respectively. Questionnaire instruments like the LBDQ and MLQ are employed to provide measures of leader behaviour. Typically, subordinates/followers complete questionnaires for each leader in a sample and the questionnaire scores are aggregated for each leader to give him or her an overall profile in terms of the leader behaviour dimensions. Questionnaire data are subject to some well-known limitations, some of which are outlined in the remainder of this paragraph. Response rates to questionnaires can be low, especially when administered by mail. Data from scales used in questionnaires can be affected by response sets, like acquiescence effects (e.g. a tendency either to agree or disagree with Likert-type inventories). One particular kind of effect that is well attested in leadership research is the role of implicit leadership theories, whereby it has been shown that when answering batches of questions about a Table 2.1 Methodological approaches to leadership discussed in the chapter Methodologies! approach Examples) discussed in this chapter Questionnaire Experiment Observation - structured observation Observation - observation in qualitative studies Interview-structured interview Interview - qualitative interview Content analysis - traditional content analysis Content analysis - historiometric studies Content analysis - qualitative/textual content analysis Discourse analysis Meta-analysis/systematic review Mixed methods research Greene (1975) Barting et al. (1996); Howell and Frost (1989); Lowin and Craig (1968); Rush et al. (1977) Luthans and Lockwood (1984) Maitlis and Lawrence (2007); Roberts and Bradley (1988) Mizrahi and Rosenthal (2001) Maitlis and Lawrence (2007); Roberts and Bradley (1988); Treviho et al. (2003) Bligh et at, (2004); Meindl et al. (1985) Mumford (2006) Boje et al. (this volume - Chapter 38); Jackson and Guthey (2007); Mumford and Van Doom (2001) Alvesson and Sveningsson (2003) Avolio et al, (2009b); DeGroot et al. (2001); Lord et al. (1986) Berson and Avolio (2004); Holmberg et al. (2008); Rowland and Parry (2009)_ 18 THE SAGE HANDBOOK OF LEADERSHIP leader's style, questionnaire respondents reply in terms of known characteristics of that leader. Thus, if a leader is believed to be ineffective, their answers with respect to that leader will be affected by this knowledge, regardless of the leader's acttial behaviour. The problem of common method (or same-source) variance/bias is also recognized in questionnaire surveys. This problem arises when respondents supply data relating to both the leadership variables (e.g. leader behaviour) and the outcome measures (e.g. organizational commitment or effectiveness of the leader) in a study. There is some disagreement over how problematic common method bias is (Spector and Brannick. 2009) but, to the extent that it is regarded as a problem, possible remedies have been proposed (Podsakoff et al., 2003). Because of beliefs in some quarters that in common method studies of leadership the relationships between leadership variables and outcomes tend to be inflated, research that uses measures of outcome that are not based on the same method and source are more robust. One difficulty with the process of aggregation of followers' scores for each leader is the suggestion that it ignores crucial differences between subordinates in their relationships with the leader, an insight that played a significant role in the development of the leader-member exchange (LMX) model discussed by Smriti Anand and colleagues in Chapter 23. Further, with questionnaire research based on a cross-sectional design, a recurring problem is the question of causality. It has been shown that the so-called outcomes of leadership variables that have been identified in much cross-sectional survey research may actually be independent variables, because leaders are likely to change their behaviour as a result of the performance of their followers. It is precisely for this reason that some mcthodologists deny that it is ever possible to establish causality in organizational behaviour studies through non-experimental studies (e.g. Stone-Romero, 2009). Questionnaires offer several advantages to the researcher and this endears them to many leadership researchers in spite of their limitations. They are invariably relatively cheap and quick to administer. A questionnaire is also a very flexible instrument that can be used to collect data on both leadership and various other variables which might be outcomes of leader behaviour or moderators of the leadership-outcome relationship. Respondents tend to be familiar with the instrument, so they do not require familiarization to use it. It can be administered to quite large samples at relatively little additional cost and because it rarely takes long to complete, a questionnaire is usually acceptable to organizations whose senior managers might otherwise worry about implications of completing questionnaires for getting work done. Questionnaires usually contain large numbers of closed questions, so coding is unlikely to be time-consuming and the risks associated with coding inconsistencies are reduced. Also, questionnaires have proved flexible in terms of the context of completion - they can be completed in a variety of locations and with different media (e.g. pens, computer, online). The absence of an interviewer removes the risk of interviewer effects arising from interviewers' characteristics and the demand effects arising from the presence of an interviewer. These advantages must be regarded as sufficiently compelling to leadership researchers for them to be used so extensively, given the limitations of this instrument outlined above. Questionnaires can be administered within the context of a longitudinal design as well as a cross-sectional design. Indeed, some of the problems with inferring causality from cross-sectional surveys have come to light as a result of the use of a longitudinal design. An early example of the use of this design is that of Greene (1975), who administered the LBDQ and measures of job satisfaction and subordinate performance (rated by peers) on three occasions with one month between each wave of data collection to two of (he subordinates of each of 103 fust-line managers in three firms. The findings confirmed the causal inference that consideration influences the job satisfaction of subordinates, but also that subordinate performance has an impact on both consideration and initiating structure. Such findings point to the internal validity problems of cross-sectional survey design studies based on questionnaires. However, the longitudinal design docs not get over the causality problem according to writers who valorize the experimental approach. As Stone-Romero (2009, p. 312) suggests, the issue of temporal precedence is not addressed by a longitudinal design because the apparent causal connection between an earlier measure (e.g. performance) and a later one (e.g. consideration) is a sequence that is part of an ongoing series of interconnections that includes prior and subsequent variables that have not been observed. EXPERIMENTAL RESEARCH Experimental studies would seem to provide the obvious solution to the causality problem identified in the previous section in relation to questionnaire research. In an experiment, the researcher manipulates the putative independent variable and observes its effects. The issue of temporal precedence and the problem that there could be RESEARCH METHODS IN THE STUDY OF LEADERSHIP 13 rival explanations of the causal relationship arc eliminated provided that there is equivalence in the experimental manipulations and random assignment to experimental and control groups. Indeed, one of the early studies to draw attention to the possibility that leadership researchers using cross-sectional designs were incorrectly inferring causality was based on experimental research. Lowin and Craig (1968) manipulated levels of 'subordinate' performance in a laboratory experiment. They found that when confronted with a poor performance level, 'leaders' exhibited higher levels of initiating structure and closer supervision and lower levels of consideration. Such a finding casts doubt on the notion that studies showing that leader behaviour and performance are related imply that leadership is necessarily the independent variable. Since then, experimental studies have been employed to show how large an impact different types of leader behaviour have on a variety of dependent variables. The experimental manipulation can be achieved in different ways, such as by using actors to deliver different styles of behaviour or by using different written scenarios that are presented to experimental participants. As an example of the former, Howell and Frost (1989) used actors to simulate charismatic leadership, consideration, and initiating structure and obsci'ved the relative impacts of these three styles on task performance among students carrying out a business task. Rush et al. (1977) provide an example of the use of a scenario within the context of an experiment. Undergraduate students at a US university were given a brief description of a person in a supervisory role. No indication was given of the nature of the individual's behaviour qua leader. Students were randomly assigned to one of the three conditions: high, average or tow departmental performance. The student participants were then asked to describe the leader in terms of the Ohio State LBDQ. Rush et al. found that the performance cue affected the ratings of the leader: for example, with the high departmental performance scenario, participants were more likely to describe the leader as high in terms of both consideration and initiating structure. Since one of the fairly consistent Findings of the Ohio researchers was that leaders high in terms of both consideration and initiating structure were most effective, a finding like this implies that when people answer questions about their leaders using an instrument like the LBDQ, they are at the very least partly affected by their prior knowledge of the leader's performance or that of his/her section. In other words, when such questionnaires are being completed, followers are only partly answering questions in terms of the actual behaviour of the leaders concerned. Field experiments are also found in the leadership field. One of the main difficulties for a field experiment, as compared lo a laboratory one, is that the experimenter usually enjoys less control over experimental arrangements, which can have adverse implications for the internal validity of the study. Most notably, it can be difficult to ensure that participants are randomly assigned to the different treatment conditions, so that frequently such studies are better referred to as quasi-experiments (Stone-Romero, 2009). An example of a field experiment that did entail random assignment can be found in a study of the impact on managers in Canadian bank branches of training in transformational leadership (Barling et al., 1996). Managers were randomly assigned lo a control group and to an experimental group mat received training in transformational leadership. Managers trained in transformational leadership were found to be more likely than those not trained to be perceived as transformational and to have followers who were committed to the organization. There was also evidence that the financial performance of their branches was superior. As noted above, experiments are often rooted within a quantitative research paradigm because the researcher is more likely to be able to impute causality by eliminating rival explanations of an apparent causal relationship than with a cross-sectional design. However, this advantage is frequently viewed as at a cost - a low level of mundane realism, particularly when the experiment is a laboratory one. Further, it is likely that only a limited range of leadership issues are likely to be amenable to experimental manipulation, whereas the use of questionnaires within a cross-sectional design is open to a wide range of research questions. In addition, the frequent use of students as experimental participants in laboratory studies has often been taken to place limits on the external validity of the study. OBSERVATION Given that much of the shidy of leadership is actually about the behaviour of leaders, it is perhaps surprising that observation is rarely used. Instead, researchers typically use questionnaires lo tap leaders' behaviour. Lowe and Gardner (2000) found in their review of a decade of articles in The Leadership Quarterly that in only 8% of articles was observation used for the collection of data. It is worth drawing a distinction between structured observation, which is invariably conducted within a quantitative research approach, and participant observation, which is a qualitative research method. 20 THE SAGE HANDBOOK OF LEADERSHIP RESEARCH METHODS IN THE STUDY OF LEADERSHIP 21 Structured observation Lulhans and Lockwood (1984) devised the Leader Observation System (LOS). This instrument was developed following extensive informal observation of managers. Following this phase, leaders' behaviour was grouped into 12 categories (e.g. planning/coordinating, staffing, exchanging routine information). Each of these is broken down into components ('behavioural descriptors') that allow observers to code ihe presence or absence of each form of behaviour within a 10-minute time slot. Each manager was observed for a 10-minute period every hour for two weeks. The authors were then able to relate managers* behaviour to their performance. It was found that leaders who gave greater emphasis to conflict management, socializing/politicking, interacting with outsiders and decision-making, bul who were less concerned with staffing, motivating/reinforcing and monitoring, were more likely to be successful (Luthans et al., 1985). Such findings are highly redolent of those engendered by questionnaire researchers, but are based on actual behaviour. There are likely to be advantages to using a research method that emphasizes whal leaders do rather than what they or others say they do, but it tends to be difficult to code behaviour in a reliable manner. It is also far more time-consuming than data collection through questionnaires and this has almost certainly played a significant part in the lack of take-up of structured observation. On the other hand, it gets around some of die problems that have been identified thus far concerning cross-sectional design studies using questionnaires, such as common method bias and contamination through implicit leadership theories, though it does not get around the problem of causa] direction. Observation in qualitative studies Observation in the course of qualitative case studies is sometimes employed by qualitative researchers, but it is rarely the sole or even ihe main method of data collection in such studies (Bryman, 2004). Maitis and Lawrence (2007) conducted multiple case study research on three British symphony orchestras. The researchers were interested in the circumstances that prompt leaders and others to engage in organizational sensegiving and the circumstances that are favourable to sensegiving. In addition to semi-structured interviewing and examination of documents, the researchers engaged in 'observation of meetings, rehearsals, and orchestra tours' (Maitlis and Lawrence, 2007, p. 60). The authors go on to write: The first author observed 107 meetings, including meetings of various groups within each organization (e.g., the executive team, the board) and meetings between orchestra leaders and external stakeholders (e.g.., flinders, collaborating organizations).... The meeting observation provided many rich opportunities to witness leaders and stakeholders engaging in sensegiving on a range of topics and to observe conditions associated with these instances of sensegiving. (Maitlis and Lawrence, 2007, p, 60) As with this study, il is unusual for observation to be the primary method. A rare example of a leadership study in which observation figures sufficiently strongly for it to be considered an ethnography is Roberts and Bradley's (1988) case study of someone who was a superintendent of a suburban school district in the USA and who later became a commissioner at state level. The research was conducted over several years and at state level, Roberts was a participanl observer in a task force and the commissioner was also observed 'making presentations to teachers and superintendents on both formal and informal occasions' and 'as she conducted meetings with her cabinet and staff and during legislative hearings, press conferences, and informal interactions with members of the Department of Education' (Roberts and Bradley, 1988, p. 259). In addition, numerous interviews were conducted over the years with the focal leader as well as with others. The research was able to show how, in her earlier role, she had been a highly charismatic leader who inspired others but, in her later role, she lost a good deal of her lustre. The authors were therefore able to examine the conditions that inhibited the transfer of charisma to her new role. There are several reasons for [he lack of use of observation in qualitative studies (Bryman, 2004), but a key reason is that (possibly), it is not always clear whal is to be observed. It is one thing to observe leaders; it is quite another thing to observe leaders/?//;. How is the observer [o know when leadership is being exercised or exhibited? In some instances, it may be obvious, but in others it may not be clear whether whal is being witnessed is leadership. If the view is taken that everything a leader does is suffused with symbolism, it could be argued that every acl is a potential act of leadership, but thai too causes possible difficulties for Che researcher because it is then difficult to know what not to observe. Such considerations, along with the time-consuming nature of intensive observation, have almost certainly limited the use of the method, even since calls for its greater use (Conger, 1998). However, participant observation offers great potential to leadership researchers because it could be especially helpful in gaining an appreciation of contextual factors and how these interact with the tasks of leadership. Further, il is especially likely to be able to offer insights into informal leadership, an area that has been neglected by the tendency for questionnaire researchers to concentrate their attention on formally designated leaders. INTERVIEW STUDIES In a similar way to observation, interviews can be categorized as either structured interviews, which are employed mainly within a quanlilalive research approach, or various kinds of qualitative interview (e.g. the semi-structured or unstructured interview). The structured interview is not a prominent technique in leadership research, in spite of the dominance of the quantitative research approach. Lowe and Gardner (2000) calculaled that 20% of The Leadership Quarterly articles were based on interviews, but this figure almost certainly included the small number of articles based on qualitative interviewing that had appeared in the journal during the period in question. Structured interviewing Mizrahi and Rosenthal (2001) employed a staictured interview in then study of social change coalitions in the New York/New Jersey area. The researchers interviewed leaders about their leadership strategies in seeking to develop successful coalitions and how ihey themselves defined success. The researchers were interested in how leaders overcame the obstacles to the formation of coalitions. An inter-view schedule of over- 600 items (mosl of which were closed questions) was used to interview 41 coalition leaders. Through the interviews, Mizrahi and Rosenthal were able to establish that persistence and vision play a crucial role in successful coalition formation. The relative lack of use of structured interviewing in leadership research is possibly surprising in view of the popularity of the method in some social sciences such as sociology. It may be that this reflects the way in which psychology has played a particularly prominent role in the field and (he relative lack of use of the method among psychologists. Qualitative interviewing Among qualitative studies of leadership, the situation could hardly be more different. Of Ihe 66 articles based on qualitative research on leadership uncovered by Bryman (2004). 56 derived at least in part from qualitative intei"viewing. The reason for its popularity among qualitative leadership researchers is similar to the reasons for the popularity of questionnaires among quanlilative researchers; il is a flexible research instmmenl that can be applied to a diversity of topics. It can also be applied to different research design contexts. Thus, it can be employed in the context of a single case study (e.g. Roberts and Bradley. 1988) or a multiple case study (e.g. Maitlis and Lawrence, 2007) or it can be employed within the context of either a cross-sectional or longitudinal design, [hough Ihe latter is very rare. A good example of the use of qualitative interviewing in the context of a cross-sectional design is the study by Trcviho et al. (2003) of 20 corporate ethics/compliance officers and 20 senior executives. The goal of the study was to uncover what is meant by elhical leadership. The researchers used a semi-structured interview administered to question their 40 informants. Three-quarters of the interviews were conducted by telephone and the rest in person. The interviews suggest lhal an ethical leadership orientation is revealed in a focus on people, role modelling of elhical behaviour, having integrity, creating and institutionalizing values and using rewards and punishments to influence ethical behaviour. Also, the qualitative interview yields dala that can be applied to a number of different analytic strategies, such as qualitative thematic textual analysis, traditional quantitative content analysis, discourse analysis and narrative analysis. As such, it is capable of generating data that are amenable to different kinds of analytic approach. Tills feature contribules to ils popularity among leadership researchers inclined towards a qualitative research approach. CONTENT ANALYSIS The term 'content analysis' often refers to a range of different techniques. According 10 one of the earliest texts on [he subject, traditional content analysis is a 'research technique for the objective, systematic and quantitative description of the manifest content of communication' (Berelson, 1952, p. 18). Given this definition, il is puzzling [hat il is referred to as a qualitative method by several writers (e.g. Insch et al., 1997; Lowe and Gardner. 2000). Content analysts may work on unstructured, qualitative material, bill lire approach to analysis is emphatically quantitative, as Berelson's classic definition demonstrates Indeed, there is a sense in which content analysis 22 THE SAGE HANDBOOK OF LEADERSHIP exemplifies quantitative research, with its emphasis on quantification and the reliability and validity of measurement. Further, the application of techniques like structural equation modelling and log-linear analysis to the quantitative data (Insch et a]., 1997) is far more redolent of quantitative than qualitative research. Traditional content analysis Content analysis can be, and indeed has been, applied to a diversity of raw materials in the leadership field. Newspaper and magazine articles, speeches, mission statements, and letters have provided some of the more common types of media to which content analysis has been applied in the leadership field. For example, Meindl et al. (1985) content analysed newspaper articles and dissertations to examine the degree to which a focus on leadership fluctuated in terms of how well the economy or particular firms or industries were faring. This analysis provided the foundation for the influential notion of the romance of leadership. Bligh et al. (2004) used computerized content analysis to examine how far the rhetoric and speeches of President George W. Bush changed after 9/11. The researchers were able to show that Bush displayed more charismatic imagery after the crisis and indeed that he was portrayed in the media in more charismatic terms than previously. Content analysis offers many advantages to the researcher, one of the most significant of which is that when applied to materials like speeches or media articles, unlike the research methods discussed thus far, it is a non-reactive measure (Bryman and Bell, 2007). It also lends itself, where She data are available, to studies of changes in communication and other kinds of content. Its chief limitation is essentially that it is difficult to determine the meaning of the findings. This can be discerned in the following passage in which Bligh et al. grapple with the meaning of their findings on Bush's charismatic imagery: But what does this shift in rhetoric imply about his charismatic stature? Did the President's underlying personality actually become more charismatic since the early days of his administration, finally blossoming during the days and weeks following the crisis?.,.. It may be that the President had the 'right stuff all along, with the crisis of 9/11 finally revealing a side of his (charismatic) personality not previously exposed to the American public. (Bligh et al., 2004, p. 227) However, if content analysts are cautious not to overinterpret their findings, content analysis offers considerable potential to the leadership researcher. There is always a risk that the content analyst is tempted to draw more inferences about the results than are warranted. It is for this reason that Berclson's definition emphasizes the manifest content of the material being examined. The search for latent content carries the risk of drawing unsustainable inferences, as well as creating potential coding problems. Historiometric research Historiometric studies have gained in popularity in recent years and are in many ways a form of content analysis. According to Mumford et al. (2009, p. 120): 'Historiometric studies represent an attempt to quantify historic observations on historic data bearing on leadership behavior and performance'. Mumford (2006) reports the results of such an examination. He analysed the biographies of 120 outstanding leaders, which allowed him to categorize the leaders into one of three types (charismatic, ideological and pragmatic). The behaviour and accomplishments of the leaders were examined, as well as some of the defining features of their earlier years. Leaders' biographies were coded in terms of a host of different dimensions. For example, the quality of leader-follower relations was coded in terms of whether the leader kept in contact with close followers after leaving power and whether the leader remained close to followers after leaving power (Mumford, 2006, p. 93). Prologue or epilogue chapters were scrutinized for tins information and then coded. Socialized leaders were more likely than personalized leaders to maintain relationships (and were also found to perform better more generally - see Howell [1988] on the personalized/socialized distinction). One issue with an analysis such as this is whether a single biography of a leader can ever be a sufficient proxy for the individual's life as a leader. To take as an example one of the leaders in Mumford's sample of 120 in which the present author has a long-standing interest - Walt Disney (Bryman, 1995) - the biography used (Mosley, 1985) was very controversial at the time it was published and the Disney family were reportedly distressed about it, though an even less flattering biography was still to come (Eliot, 1993). Relatedly, Mumford et al. (2009) point out that historiometric studies tend to be leader-centric in that they will focus on what the leader said and did but will relatively rarely provide She perspective of the follower. Therefore, historiometric studies will typically not provide significant insights into followers' reactions to leaders. On the other hand, such studies do provide an understanding of very RESEARCH METHODS IN THE STUDY OF LEADERSHIP ?3 creative leaders to whom access is normally difficult or impossible. Further, they allow us to study leadership over time, in that they give us access to such topics as formative influences and changes in leadership orientation over the individual's life course. Qualitative/textual content analysis of texts Qualitative/textual content analysis is not a distinct method like traditional content analysis. It comes in several guises, but the various approaches exhibit the common feature of the researcher extracting themes from the data. It can be discerned in Jackson and Guthey's (2007) analysis of visual images to explore what they call the 'celebrity CEO backlash', the process by which CEOs who would at one time have been shown in photographs as powerful superstars are increasingly being depicted in unflattering ways. Mumford and Van Doom (2001) analysed various primary sources to examine 10 cases of exceptional leadership exhibited by Benjamin Franklin. Through this exercise, the researchers were able to explore Franklin's credentials as a pragmatic leader. As Mumford and Van Doom point out, this is a significant issue because of a tendency to associate outstanding leadership with charismatic or transformational leaders. Mumford's (2006) previously mentioned historiometric study also makes this point. A further example of qualitative/textual content analysts can be discerned in Chapter 38, in which David Boje and his co-authors use a variety of sources, such as periodical articles, television and newspaper articles, and web sources, to develop the notion of virtual leadership. Such studies can be very fertile in terms of generating new concepts or areas of enquiry (e.g. CEO celebrity backlash or virtual leadership) and in developing notions like pragmatic leadership. As such, they can be a source of theoretical creativity. The close relationship between the sources and the conceptualization provides a strong sense of the face validity of the concepts. It is sometimes suggested that the work lacks robustness: what, for example, might provide counter-examples or material that might shed doubt on the findings that are developed? DISCOURSE ANALYSIS Discourse analysis is a family of techniques used to examine the part played by language in human affairs. There is no one version of discourse analysis (Alvesson and Karreman. 2000) but all the approaches share a predilection for analyses of language in use. 'Language' here may involve natural language in the course of conversations, the linguistic framing of answers to interview questions, rhetorical devices used in speeches, or the ways in which issues are framed in documents. Crucial to discourse analysis is the notion that discourse is not 'just' a neutral medium of communication. Instead, discourse analysts view social reality as socially constructed and propose that discourse plays a central role in the constitution of the social world. As such, the orientation of discourse analysis is somewhat different from that of many other qualitative methods. Instead of treating the social world as given, practitioners emphasize how everyday notions and objects are constituted through discourse. In Chapter 36 of this volume, Gail Fairhurst makes a distinction between Big D and Little D discourse. Studies working within the latter (c/is-course) approach tend to emphasize fairly finegrained analyses of language in the context of interaction within organizations. By contrast, Discourse studies derive from a Foucauldian tradition that focuses on 'historical constellations of talk patterns, ideas, logics, and assumptions that constitute objects and subjects in particular ways'. Such studies, while attending to the detail of language, tend to be somewhat more broad brush in orientation. An interesting example of a discourse analytic study is Alvesson and Sveningsson's (2003) research on the language of leadership and management. This article begins with a Big D issue -the way in which in the 1980s and early 1990s the popular literature on leadership developed a Discourse of leadership that distinguished it from management and in the process positioned leadership as essentially effective leadership. Alvesson and Sveningsson report a discourse analysis of qualitative interview transcripts of the accounts by managers of the nature of their leadership. They note that managers, when describing their leadership, were inconsistent in their accounts. For example, managers often repeated and claimed adherence to the Discourse of the day that positioned leadership (as distinct from management) as vision, but, when given the opportunity to elaborate on what that entails and how it is done, they turned to an emphasis on administrative and operational issues. It would seem that the managers were attracted to the Discourse of the day which positioned them as strategic masters but that when discussing their own leadership in some detail, the emphasis on vision gave way to the very immediate organizational demands and constraints they faced. In this study then, the authors 24 THE SAGE HANDBOOK OF LEADERSHIP RESEARCH METHODS IN THE STUDY OF LEADERSHIP 25 undertook a discourse study of the .Discourse of leadership. One of the chief arguments often levelled at discourse analysis generally is that its focus tends to occlude the operation of wider structures and the use of power (e.g. Reed, 2000). This orientation does not sit well with leadership theorists and organizational researchers more generally of a critical persuasion. However, as can be seen in the Alvesson and Sveningsson (2003) study, discourse analysis can be deployed in such a way that it can interrogate and illuminate larger-scale issues and concerns such as the capacity of wider' Discourses to influence and, in this case, cunfuse the discourses of managers. Also, writers whose research is influenced by critical discourse analysis tend to be less implicated in this accusation (Phillips and Di Domenico, 2009). META-ANALYSIS Reviews of leadership literature have long played a significant role in leadership research. For example, the negative reviews by Stogdill (1948) and Mann (1959) ot extant trait research on leadership have often been held responsible for sounding its death knell aird for ushering in the emphasis on leadership style (e.g. Bryman, 1986). However, by pooling the findings from groups of studies investigating a topic, meta-analyses can often provide a better overview. Thus, as John Antonakis points out in Chapter 20, when Lord et al. (1986) conducted a meta-analysis of the same studies as those reviewed by Mann (1959). intelligence was found contra Mann to be related to leadership emergence. Leadership is a field that lends itself quite well to meta-analysis because researchers are interested in impacts of leadership variables on various outcomes. Meta-analysis can also add to understanding because its capacity to pool a large number of studies can mean that moderating factors (e.g. research method used or situational factors) that are not revealed through individual studies can be examined. An example of a metaanalysis in the field of leadership is the review of published studies of the organizational outcomes related to charismatic leadership (DcGioot et al. 2000). Most of the measures of charismatic leadership were based on the Multifactor Leadership Questionnaire (MLQ), referred to by Jay Conger in this volume in Chapter 7. The meta-analysis was able to confirm diat charismatic leadership is related to: leader effectiveness; subordinate effectiveness; subordinate satisfaction; and subordinate commitment. The authors were able to identify that common method bias had some impact on the findings, with evidence of the relationship between charismatic leadership and job satisfaction being especially affected. Meta-analysis suffers from certain difficulties, most notably the 'file drawer' problem, which refers to the fact that not all research gets into the public domain and that it is likely that research studies showing weak or no impacts may be particularly affected. DeGroot et al, did try to find unpublished research but were unsuccessful. However, they did adjust their data for the possible effects of this problem. The file drawer problem would have been a particular difficulty for Avolio et al. (2009b) who conducted a metaanalysis of experimental and quasi-experimental studies of the impact of leadership inteiTentions. In addition to searching databases, Avolio et al. wrote emails to 670 people who worked in leadership and related areas, asking them to review for inclnsiveness a list of leadership impact studies. The authors found 200 studies that met their criteria and, of these, 16% were unpublished, suggesting that at least in part the researchers had been able to address the file drawer problem. Interestingly, one of the main findings of this review was that while leadership interventions do have demonstrable impacts, those impacts tended to be stronger when the research was underpinned by a leadership theory that emphasized behavioural change rather than when the theory emphasized emotional pr cognitive change. MIXED METHODS RESEARCH Mixed methods research has increasingly come to refer to research that combines quantitative and qualitative research. Methods can be and often are mixed within these two research strategies but with the arrival of a mixed methods industry in the form of a journal, handbook, a regular conference, and several books, it tends to denote a mixture of quantitative and qualitative approaches (Bryman, 2009). Mixed methods research is also being advocated by leadership researchers (e.g. Hunt and Conger, 1999), and others see it as likely to be an increasingly important approach. Avolio et al. (2009a, pp. 441-2), for example, write: 'We expect to see a greater use of mixed-methods designs in future research.... [Ijncreasing attention is being paid to cases and qualitative research that should now be integrated with quantitative approaches'. When Bryman (2004) reviewed qualitative research on leadership, he uncovered 12 articles that used a mixed methods approach. It is striking that trine of these articles were published during the brief period 2000-03. This evidence of growing popularity reflects a growing openness among leadership researchers to both the use and the publication of qualitative research and an increased awareness among researchers generally in the potential of mixed methods research (Bryman, 2009). There are a number of ways of combining quantitative and qualitative research. In his review, Bryman (2004) used the scheme for classifying mixed methods research developed by Greene et al. (1989) and found that three approaches were particularly prominent: triangulation (using quantitative and qualitative research to cross-validate findings); expansion and complementarity (one set of data is used to expand on or complement the other set); and different issues (using quantitative and qualitative research to examine different aspects of the research topic). The last of these approaches was not a feature of the Greene et al. scheme (Bryman, 2009), Mixed methods studies typically involve a cross-sectional survey design that combines either a structured interview or more usually a self-administered questionnaire for the quantitative data with semi-structured interviews for the qualitative data (Bryman, 2006), and the leadership field seems to fit with this tendency (Bryman. 2004). For example, in a study that was published in 2004, and which was therefore outside the date range of the review, Berson and Avolio (2004) report the results of a study of transformational leadership and the dissemination of organizational goals m a large Israeli telecommunications company. This investigation entailed a self-administered questionnaire completed by managers and others to collect data on managers' leadership styles (using the MLQ) and communication style. In addition, semi-structured interviews were carried out with managers and their direct reports. They write: 'By using both quantitative aird qualitative methods, we were able to depict specific relationships, and then to provide a deeper examination of those relationships' (Berson and Avolio, 2004, p. 642), which is indicative of using one data set to expand on and complement the other, Tliis article was highlighted as one of a number that could be used as a template for exploring the levels of analysis issue that has preoccupied many leadership researchers in recent years, but not always in a technically appropriate manner (Yammarino et al., 2005). Rowland and Parry (2009) included observation of team meetings, semi-structured interviews and questionnaires in then grounded theory-based study of organizational design and leadership in teams in Australian higher education, as did Holmberg et al. (2008) in their investigation of the role of leadership in the implementation of evidence-based treatment practices for drug abuse and criminal behaviour in Sweden. A rather striking mixed methods approach call be discerned in Goodall's (2009) study of the leadership of top universities and business schools. Against a background in which it is sometimes argued thai universities would be belter led by non-academics, Goodall shows that better universities and business schools tend to be led by top scholars. The study combines quantitative and qualitative data. The former are derived from a secondary analysis of biographical information on leaders gleaned from publications like Who's Who and from the examination of citations, an approach usually referred to as bibliometrics. The qualitative data derived from interviews with 26 top leaders of universities in the UK and USA (vice-chancellors/principals/presidents) or deans of business schools. She also conducted 12 interviews in connection with a case study of the hiring of a UK vice-chancellor. The quantitative biblio-metric data provided the confirmation of the relationship between the excellence in scholarship and university performance, while the qualitative data allowed some of the factors lying behind this relationship to surface. While there has undoubtedly been a noticeable increase in mixed methods research to examine leadership, it is also the case that in much of it, the qualitative research component occupies a handmaiden role (Bryman, 2007), largely being used to illustrate or add to or support the quantitative findings rather than for what it can offer in its own right. Rowland and Parry's (2009) study represents an exception in adopting a clear constructionist orientation. The tendency for qualitative research to assume a somewhat subsidiary role in many studies probably reflects the training and general orientation of many leadership researchers and the perceived expectations of journal.*.. CONCLUSIONS Although it is possible, and almost certainly valid, to write about a typical leadership study (Friedrich et al., 2009), it is clear that this is considerable diversity in methodological approach within the field and one gets a sense that that diversity is increasing. In pan, the glowing diversity reflects a greater acceptance of qualitative methods. It is likely that the growing awareness of research methods which have been neglected by leadership researchers, like qualitative ones, will also suggest new ways of thinking about leadership. When qualitative research methods have been used in leadership research, it has often been in a way that 16 THE SAGE HANDBOOK OF LEADERSHIP is either subsidiary to the quantitative component in mixed methods research or mimics some of the features of quantitative research so that it looks like quantitative research hut without numbers (Bryman, 2004). However, an investigation like that of Alvesson and Sveningsson (2003), which was discussed above, treats leadership as a social construction and, as such, is more in tune with the ethos of qualitative research. Thus, as the appreciation of qualitative research grows and intensifies, it is feasible that the growing awareness of the potential of qualitative and other underused methods will stimulate leadership researchers to think about leadership in new ways and suggest new research questions. However, it is important not to characterize the field's growing diversity as purely methodologically driven. As the leadership field becomes more theoretically diverse, writers are likely to feel impelled to explore more methods and to be innovative. For example, seeking to apply complexity theory to leadership led Dooley and Lichtenstein (2008) to develop methods to observe in real time events and interaction that take place in very, very brief time slots. By studying numerous time slots and the bits of interaction that take place within them, the researchers expect to be able to examine how far particular leadership patterns are influenced by context or are constant. As we have seen, observation is not uncommon in leadership research (and indeed was behind Bales and Slater's f 1955J classic study of leadership emergence), but the microscopic approach taken by Dooley and Lichtenstein in their working through of some of the implications of complexity theory points to considerable experimentation. Similarly, towards the end of their chapter (Chapter 34) on complexity theory in the present volume, Uhl-Bien and Marion note that the working through of the implications of this perspective is likely to be more reliant on 'more qualitative and agent-based modeling approaches' than on the traditional questionnaire approach. They suggest that this is likely to be so because the ontology of the notion of complexity does not lend itself to the positivism with which the questionnaire approach is associated. Their suggestion that the complexity theory approach is likely to be more reliant on a process than a variance theory stance (Langley, 2009) further points to a greater reliance on qualitative research. In this way, the emergence of a new theoretical stance is acting as a spur to methodological innovation. To take another example, heavily influenced by process philosophy, Wood and Ladkin (2008) developed the idea of the 'leaderM moment', a slice of time in which leadership emerges and conies into being. S uch a view departs significantly, as in Dooley and Lichtenstein, from the more common view of leadership as part of an ongoing set of relationships that are founded on influence. To gam some leverage on the notion of the leadeiful moment, the authors asked managers to 'photograph things, people or moments they perceived as inextricably linked with the experience of leadership, but which might usually go unnoticed' (Wood and Ladkin, 2008, p. 18). The managers were later interviewed about their photographs, a method often called 'photo-elicitation'. Thus, the development of new theoretical approaches frequently prompts researchers to develop new ways of collecting data about their object of interest, especially when, as with these two research areas, more conventional ways of collecting data do not look promising. Leadership researchers have long been aware that the field is piobably over-reliant on questionnaire studies. Nowadays, that awareness is being accompanied by a preparedness to exptore new or different ways of getting to grips with the complex and elusive phenomenon of leadership. At the same time, little has changed since Conger's (1998) call for greater use of participant observation in die study of leadership. As noted above, ethnographic studies have considerable potential in the field in helping us to appreciate how feadership takes place, the 'leaderful moments' that undoubtedly occur in organizations, how context and leadership are intertwined, and the fact that leadership may occur anywhere and be exhibited by anyone -not just where leadership researchers assume it will take place. Leadership researchers have been engaged for some time in a collective mea culpa about the dominance and limitations of the questionnaire in their fietd. Now it is time to do something about it, particularly at a juncture at which new theoretical approaches and influences appear to be influencing the field more and more. REFERENCES Alvesson, M. and Karreman, D. (2000) 'Varieties of discourse: on the study of organization through discourse analysis'. Human Relations, 53(9), 1125-49. Alvesson, M. and Sveningsson, S. (2003) 'Good visions, bad micro-management and ugly ambiguity, contradictions of (non-)leadership in a knowledge-intensive organization', Organization Studies, 24(6), 961-88. Antonakis, J., Schriesheim, C.A., Donovan, I A., et af (2004) 'Methods for studying leadership'. In J. Antonakis, A.T. Ciancioloand R.J. Sternberg (eds), The Nature of Leadership. Thousand Oaks, CA: Sage, pp. 48-70. Avolio, B.J., Walumbwa, F.O. and Weber, T.J. (2009a) 'Leadership: current theories, research, and future directions', Annual Review of Psychology, 60, 421-49. Avolio, B.J., Reichard, R.J., Hannah, ST., Walumbwa, F.O. and Chan, A. (2009b) 'A meta-analytic review of leadership RESEARCH METHODS IN THE STUDY OF LEADERSHIP 27 impact research: experimental and quasi-experimental studies', The leadership Quartetly, 20, 764-84. Bales, R.F. and Slater, P.E. (1955) 'Role differentiation in small decision-making groups'. In T. Parsons and R.F. Bales (eds), Family, Socialization and Interaction Process. New York: Free Press. Barling, J., Weber, T. and Kelloway, E.K. (1996) 'Effects of transformational leadership training on attitudinal and financial outcomes: a field experiment', Journal of Applied Psychology, 81(6), 827-32. Berelson, B. (1952) Content Analysis in Communication Research. Mew York: Free Press. Berson, Y. and Avolio, B.J. (2004) 'Transformational leadership and the dissemination of organizational goals: a case study of a telecommunications firm', The Leadership Quarterly, 15(5), 625-46. Bligh, M.C., Kohles, J.C. and Meindl, J.R. (2004) 'Charisma under crisis: presidential leadership, rhetoric, and media responses before and after the September 11th terrorist attacks', The Leadership Quarterly, 15(2), 211-39. Bryman, A. (1986) Leadership and Organizations, london: Routledge. Bryman, A. (1995) Disney and His Worlds. London: Routledge. Bryman, A. (2004) 'Qualitative research on leadership; a critical but appreciative review', The Leadership Quarterly, 15(6), 729-69. Bryman, A. (2006) 'Integrating quantitative and qualitative research: How is it done?', Qualitative Research, 6(1), 97-113. Bryman, A. (2007) 'Barriers to integrating quantitative and qualitative research', Journal of Mixed Methods Research, 1(1), 8-22. Bryman, A. (2009) 'Mixed methods in organizational research'. In D.A. Buchanan and A. Bryman (eds). Handbook of Organizational Research Methods. London: 5age, pp. 516-31. Bryman, A. (2011) 'Mission accomplished? Research methods in the first five years of Leadership', Leadership, 7(1). Bryman, A. and Bell, E. (2007) Business Research Methods, revised ed. Oxford: Oxford University Press. Collinson, D. and Grint, K. (2005) 'Editorial: the leadership agenda', Leadership, 1(1), 5-9. Conger, J.A. (1998) 'Qualitative research as a cornerstone methodology for understanding leadership', The Leadership Quarterly, 9(1), 107-21. DeGroot, T, Kiker, D.S. and Cross, T.C. (2000) 'A metaanalysis to review organizational outcomes related to charismatic leadership', Canadian Review of Administrative Sciences. 17(41,356-71 Dooley, K.J. and Lichtenstein, B, (2008) 'Research methods for studying the dynamics of leadership'. In M. Uhl-Bien and R. Marion (eds), Complexity Leadership. Fart i: Conceptual Foundations. Charlotte, NC: Information Age, pp. 269-90, Eliot, M. (1993) Walt Disney: Hollywood's Dark Prince. New York; Birch Lane. Friedrich, T.L, Byrne, C.L and Mumford, M.D. (2009) 'Methodological and theoretical considerations in survey research'. The Leadership Quarterly, 20(2), 57-60. Gardner, W.L., Lowe, K.B., Moss, T.W., Mahoney, K.T. and Cogliser, C.C. (Unpublished work) 'Scholarly leadership of the study of leadership: a review of The leadership Quarterly's second decade, 2000-2009', unpublished MS, Goodall, A.H. (2009) Socrates in the Boardroom: Why Research Universities Should be Led by Top Scholars. Princeton: Princeton University Press. Greene, C.N. (1975) 'The reciprocal nature of influence between leader and subordinate', Journal of Applied Psychology, 60(2), 187-93. Greene, J.C, Caracelli, V.J. and Graham, W. F. (1989) 'Toward a conceptual framework for mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3), 255-74. Holmberg, R., Fridelt, M„ Arnesson, P. and Backvall, M. (2008) 'Leadership and implementation of evidence-based practices', Leadership in HealLh Services, 21(3), 168-84. Howell, J. M. (1988) 'Two faces of charisma: Socialized and personalized leadership in organizations,' in J. Conger S R. Kanungo (eds.), Charismatic leadership: The elusive factor in organizational effectiveness (pp. 213-236). San Francisco: Jossey-Bass. Howell, J.M. and Frost, P.J. (1989), 'A laboratory study of charismatic leadership'. Organizational Behavior and Human Decision Processes, 43(2), 243-69. Hunt, J.G. and Conger, J.A. (1999) 'From where we sit: an assessment of charismatic and transformational leadership research', The Leadership Quarterly, 10(3), 335-43. Hunter, ST., Bedell-Avers, £ and Mumford, M.D. (2007) 'The typical leadership study; assumptions, implications, and potential remedies', The Leadership Quarterly, 18(5), 436-46. Insch, G.S., Moore, J.E. and Murphy, LP. (1997) 'Content analysis in leadership research: example, procedures, and suggestions for further use', The Leadership Quarterly, 8(1), 1-25. Jackson, B. and Guthey, E. (2007) 'Putting the visual into the social construction of leadership'. In B. Shamir, R. Pillai, M.C. Bligh and M. Uhl-Bien (eds), Follower-Centered Perspectives on Leadership: a Tribute to the Memory of James R. Meindl. Greenwich, CN: Information Age, pp, 167-86. Klenke, K. (2008) Qualitative Research in the Study of Leadership. Bingley: Emerald. Kroeck, K.G. Lowe, K.B. and Brown, K.W. (2004), 'The assessment of leadership'. In J. Antonakis, AT. Cianciolo and R.J. Sternberg (eds), The Nature of Leadership. Thousand Oaks, CA: Sage, pp, 71-97. Langley, A. (2009) 'Studying processes in and around organizations'. In D.A. Buchanan end A. Bryman (eds), Handbook of Organizational Research Methods. London: Sage, pp, 409-29. Lord, R.G. De Vader, C.L. and Alliger, G.M. (1986) 'A metaanalysis of the relation between personality traits and leadership perceptions: an application of validity generalization procedures', Journal of Applied Psychology, 71(3), 402-10. — 38 THE SAGE HANDBOOK OF LEADERSHIP Lowe, K.B. and Gardner, W. L. (2000) 'Ten years of The Leadership Quarterly, contributions and challenges for the future', The Leadership Quarterly, 11(4), 459-514, Lowin, A. and Craig, C.R. (1968) 'The influence of performance on managerial style: an experimental object lesson in the ambiguity of correlational data', Organizational Behavior and Human Performance, 3, 440-58 Luthans, F, and Lockwood, D.L. (1984) 'Toward an observation system for measuring leader behavior'. In J.G. Hunt, DM. Hosking, C.A. Schriesheim and R. Stewart (eds), Leaders and Managers: International Perspectives on Managerial Behavior and Leadership. New York: Pergamon. Luthans, F., Rosenkrantz, S.A. and Hennessey, H.W. (1985) 'What do successful managers really do? An observation study of managerial activities', Journal of Applied Behavioral Science, 21: 255-70, Maitlis, S. and Lawrence, T.B. (2007) 'Triggers and enablers of sensegiving in organizations', Academy of Management Journal, 50(1), 57-84. Mann, R.D. (1959) 'A review of the relationship between personality and performance in small groups'. Psychological Bulletin, 56(4), 241-70, Meindl, J.R., Ehrlich, 5.B. and Dukeridt, J.M. (1985) 'The romance of leadership', Administrative Science Quarterly, 30, 78-102. Mizrahi, T. and Rosenthal, B.B. (2001) 'Complexities of coalition building: leaders' successes, strategies, struggles, and solutions', Social Work, 46(1), 63-78. Mosley, L, (1985) Disney's World. Briarcliff Manor, NY: Stein and Day, Mumford, M.D. (2006) Pathways to Outstanding Leadership: A Comparative Analysis of Charismatic, Ideological and Pragmatic Leaders. Mahwah, NJ: Erlbaum. Mumford, M.D. and Van Doom, J.R. (2001) 'The leadership of pragmatism: reconsidering Franklin in the age of charisma', The Leadership Quarterly, 12(3), 279-309. Mumford, M.D., Friedrich, T.L., Caughron, J.J. and Antes, A.L. (2009) 'Leadership research: traditions, developments, and current directions'. In D.A. Buchanan and A. Bryman (eds), Handbook of Organizational Research Methods. London: Sage, pp. 111-27. Phillips, N. and Di Domenico, M,L (2009) 'Discourse analysis in organizational research: methods and debates'. In D.A. Buchanan and A. Bryman (eds), Handbook of Organizational Research Methods. London: Sage, pp. 549-65, Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y. and Podsakoff, N.P. (2003) 'Common method biases in behavioral research: a critical review of the literature and recommended remedies', Journal of Applied Psychology, 88(5), 879-903. Reed, M. (2000) 'The limits ol discourse analysis ill organizational analysis'. Organization 7(3): 524-30. Roberts, N.C. and Bradley, R.T. (1988) 'The limits of charisma'. In J,A. Conger and R.N. Kanungo (eds), Charismatic Leadership: The Elusive Factor in Organizational Effectiveness. San Francisco: Jossey-Bass, pp. 253-75. Rowland, P, and Parry, K, (2009) 'Consensual commitment: a grounded theory of the meso- level influence of organizational design on leadership and decision-making', The Leadership Quarterly, 20(4), 535-53. Rush, M.D,, Thomas, J.C. and Lord, fi.G, (1977) 'Implicit leadership theory: a potential threat to the internal validity of leader behavior questionnaires', Organizational Behavior and Human Pedormance, 20(1), 93-110. Spector, P.E. and Brannick. M.T. (2009) 'Common method variance or measurement bias? The problem and possible solutions'. In D.A. Buchanan and A. Bryman (eds), Handbook of Organizational Research Methods. London: Sage, pp. 346-62. Stogdill, R.M. (1948) 'Personal factors associated with leadership: a survey of the literature', Journal of Psychology, 25, pp. 35-71, Stone-Romero, L.I. (2009) 'Implications of research design options for the validity of inferences derived from organizational research'. In D.A, Buchanan and A. Bryman (eds), Handbook of Organizational Research Methods. London: Sage, pp. 302-27. Treviho, L.K., Brown, M. and Pincus Hartman, L. (2003) 'A qualitative investigation of perceived executive ethical leadership: perceptions from inside and outside the executive suite', Human Relations, 56(1), 5-37. Wood, M. and Ladkin, D. (2008) 'The event's the thing: brief encounters with the leaderful moment'. In K.T. Jones and J, Collins (eds). Leadership Perspectives: Knowledge into Action. Houndmills; Palgrave-Macmillan, pp. 15-28. Yammarino, F.J., Dionne, S.D., Chun, J.U. and Danserau, F. (2005) leadership and levels of analysis: a state-of-the-srience review'. The Leadership Quarterly, 16(6), 879-919. 3 The Enduring and Elusive Quest for a General Theory of Leadership: Initial Efforts and New Horizons Georgia Sorenson, George R. Goethals and Paige Haber This is my quest to follow that star, no matter how hopeless, no matter how far... 'The Impossible Dream' from Man of La Mancha (Darion, 2008) THE QUEST When tbe idea of convening a group of scholars to formulate a general theory of leadership was first proposed, one of those who eventually became a key member remarked that the idea of such a project was 'quixotic' Professor Toanne Ciulla used the term exactly as it is defined - as the American Heritage Dictionary (2009) has it, 'idealistic or romantic without regard to practicality.' What a charming, silly idea. And in the end, the quest and idealism endures but the goal of a general theory remains elusive. However, as Ciulla herself documents, we went far, and learned a great deal along the way. The Quest for a General Theory of Leadership (GTOL) involves a story about process, and another about product. The story begins with Tames MacGregor Burns. Burns is a restless scholar who began thinking more generally about the phenomenon of leadership after his Pulitzer Prize and National Book Award-winning classic Roosevelt: The Soldier of Freedom was published in 1970. Fie believed that be would need to expand beyond his familiar disciplines of political science and history in order to fully comprehend the subject. Accordingly, he immersed himself in the fields of philosophy and psychology and in 1978, at age 60, published one of his most influential works. Leadership. Burns then became increasingly interested in fostering tbe study of leadership. He laid the groundwork for the Program in Leadership Studies at Williams College, his alma mater, and, in the early 1990s, became closely involved with shaping and establishing the Jepson School of Leadership Studies at the University of Richmond and the Center for Political Leadership and Participation at Ihe University of Maryland. In 1997 the center's name was changed to The James MacGregor Burns Academy of Leadership to honor his lasting contribution to the field. Writing about leadership and promoting its study was not enough for Bums. He perceived a need for greater intellectual coherence in an extremely wide-ranging field of study and practice. Tn an interview with Sorenson and Goethals on July 5, 2009, Professor Burns stated that studying leadership in Ihe early years was liberating and took him beyond a focus on biography and politics. He added that the study of leadership demanded intellectual creativity and reach. After 36 THE SAGE HANDBOOK OF LEADERSHIP working in an interdisciplinary intellectual environment, and the GTOL product is a serious and successful attempt at bridging and integrating these disciplinary silos. From this, continued interdisciplinary conversations and explorations of leadership can take place and a clearer picture of disciplinary overlap and divergence can emeTge. While GTOl. covered a great deal of intellectual ground, there are a few areas that could be explored in more depth. The discussion of power could be expanded to greater emphasis on motiva-lion and influence, and the leader-follower relationship discussion can be furthered through exploring relationships between group members (within group) and between leaders (inteTgroup). Additionally, greater focus on Ihe purpose of the leadership process can be expanded. The inclusion of group relations work may provide some insight into these areas as well as contribute to continued exploration of various levels of the leadership system: intrapersonal, interpersonal, group, and system as a whole. In reflecting back on the process, one of the authors shared that complexity science, cognitive studies, and human development studies may also provide insight. GTOL has broadened the leadership studies field, and there is much potential for "nature growth and development of the conversations it brought forward. Leadership studies must continue to be challenged to move beyond the leader—follower-shared goal conversation. To embrace (he complex and adaptive nature of leadership studies and societal leadership challenges, there is a call for more organic, systemic, and integrative ideas and approaches. It will not be easy, but to end at the beginning. Burns concludes in Ihe Quest for A General Theory Let me leave you with a challenge and a question. The amazing events that unfolded in Montgomery and the state and nation are that the people in action embraced every major aspect of leadership and integrated it: individual leadership, collective leadership, intra-group and inter-group conflict, conflict of strongly held values, power aspects, etc. - and ultimately produced a real change leading to more change. They made our country a better country. If those activists could integrate the complex processes and elements of leading in practice, in reality, should we not be able to do so in theory? REFERENCES American Heritage Dictionary of the English Language, 4th edn (2009), Boston, MA: Houghton Mifflin. Banks, S.P. (2008) Dissent and the Failure of Leadership. Cheltenham, UK: Edward Elgar, p. 19. Burns, J. (2009) Personal Communication, 1 June 2009 Couto, R. (2007) Reflections on Leadership. Lanham, MO: University Press of America, p. 1. Darion, J. (2008) 'The Impossible Dream, Man of La Manchi. Gill, R. (2006) Theory and Practice of Leadership, london: Sage, p. 91. Goethals, G. and Sorenson, G. (eds) (2004) The Encyclopedia of Leadership. Thousand Oaks, CA: Sage. Goethals, G. and Sorenson, G. (eds) (2006) The Quest for a General Theory of Leadership. Cheltenham; UK: Edward Elgar, pp. 39-153. Grandstaff, M. and Sorenson, G. (2008) Strategic Leadership: A General's An. Vienna, VA: Management Concepts. Hickman, G. Leading Change in Multiple Contexts: Concepts and Practices in Organizational, Community, Political, Social, and Global Change Settings, July 2009, Sage books, Mangan, K. (2002) 'Leading the way in leadership: the unending quest of the discipline's founding father, James MacGregor Burns', Chronicle of Higher Education, May 31, p. 39. Morrill, R. (2007) Strategic Leadership: Integrating Strategy and Leadership in Colleges and Universities. Washington, DC: American Council on Education, p. 8, 14. Reyatt, K. 2009. GTOL II - General Theory of Leadership - the Neat Phase. Online, Available at http://www.iliaspace.org/ groun/feadershipscholarshipmig/forurn/topics/gtol-ii-general-theory-of. Accessed June 11 2009. Sorenson, G, (2007) 'An intellectual history of leadership studies: the role of James MacGregor Burns'. In R. Couto (ed.), Reflections on Leadership, Lanham, MD: University Press of America, pp. 19-30. Wilbur, K. (2000) A Theory of Everything. Boston, MA: Shambhaia Publications. 4 Leadership Development INTRODUCTION Every year organizations invest considerably in developing their leaders. Annual estimates range from $16.5 billion (Fulmer & Goldsmith, 2001) to over $45 billion (Lamoureux, 2007) for leadership development programs and other supporting management/executive education activities in the United States alone. A 2004 survey indicated that respondent companies reported spending on average approximately $7 million annually on executive development, including formal classroom offerings and other related concerns such as coaching and developmental job experiences (Bolt, 2007). Whereas this is a substantial estimate of executive development investments, the author of the report goes on to say that, '.. .many felt this amount was understated' (p. 20). From these estimates it is clear that leadership development is big business. A question that will be addressed in this chapter is: What is the evidence lhat it is also good science and therefore also good business? A decade ago Day (2000) reviewed the literature on leadership development and concluded that interest in the field 'appears to be at its zenith' (p. 581) especially among those with more applied interests (e.g., HR practitioners, consultants); however, there was a good deal less scholarly interest in the topic, as evidenced by the lack of published research in the area. This led to several appeals for building a science of leadership development (Day & O'Connor, 2003; Day & Zaccaro, 2004). Thus, one goal of this chapter is to examine what has transpired subsequently in terms of relevant research, what conlributions have been made toward establishing a more evidence-based approach (Rousseau, Manning, & Denyer. 2008) David V. Day to leadership development, and what are the major areas of research thai have yet to be undertaken. Specifically, how much (or little) progress has been made in terms of theory, research, and practice in the first decade of the new millennium, and where should we turn our scholarly attention in the future? One emerging development in the field is that recent global survey data collected by researchers at Developmental Dimensions International (DDI; Howard & Wellins, 2008) suggest that leaders are increasingly dissatisfied with (heir organization's development offerings. In addition, perceived program quality has declined, developmental programs are seen as poorly executed, and confidence in leaders continues to decline steadily (among other noted concerns). This has led the authors of the DDI report to adopt a pessimistic tone in commenting that despite its criticality to long-term organizational sustainability, '...leadership development is going nowhere fast' (p. 4). Given the considerable financial investments lhat organizations appear to make in leadership development initiatives, this apparent state of affairs is especially distressing. One possible explanation for this critical assessment of the leadership development Held on the part of survey respondents is that expectations have risen as developmental initiatives have become more widely used. But as will be elaborated on in more detail in the chapter, there is still the erroneous belief that leadership develops mainly in leadership development programs. In evaluating this limited and limiting belief about where and how development occurs, several facets of leadership development will be examined, including recent theory, empirically based research on the topic, practice-based advancements, as well THE SAGE HANDBOOK OF LEADERSHIP LEADERSHIP DEVELOPMENT 19 as future directions in the field. The overall goal is to use these various lenses to better determine where the field of leadership development is heading and to identify what are the most pressing challenges now and for the future. THEORETICAL PERSPECTIVES AND LEADER/LEADERSHIP DEVELOPMENT A basic distinction has been drawn between leader development and leadership development, focusing on the development of individuals (leaders) as compared to the development of social structures and processes (Day, 2000). They are not synonymous, but are often treated and discussed that way. The former is the more common and traditional approach that is focused on building individual capabilities, whereas the latter is moving more towards teambuilding and organizational development. The distinction between leader and leadership development was further elaborated on in the revised edition of The Center for Creative Leadership (CCL) Handbook of Leadership (McCauley & Van Velsor, 2004) in which leader development was defined as 'the expansion of a person's capacity to be effective in leadership roles and processes' (Van Velsor & McCauley, 2004, p. 2), What is thought to develop in leader development includes individual self-management capabilities (e.g., self-awareness, balancing conflicting demands, ability to learn, and leadership values), social capabilities (e.g., ability to build and maintain relationships, building effective work groups, communication skills, and ability to develop others), and work facilitation capabilities (e.g., management skills, ability to think/act strategically and creatively, and ability to initiate and implement change) (Van Velsor & McCauley. 2004). This type of personal leader development is characterized by several features, including that it unfolds over time; is maximized by a variety of experiences that provide feedback, challenge, and support; and is also contingent on an individual's ability and willingness to learn from experience. Finally, the CCL perspective maintains that the most effective leader development initiatives are those that integrate various experiences and embed them in the organization's context. In this manner the CCL approach acknowledges that unless leader development can be made part of the everyday business of an organization it will fall short of optimal effectiveness (Vicere & Fulmer, 1998). It is telling thai The CCL Handbook of Leadership devotes 13 chapters to various topics associated with leader development, yet only two chapters are devoted to leadership development (e.g., O'Connor & Quinn, 2004; Palus & Horth, 2004). This is not a limitation of the CCL approach; rather, it highlights that much more is known about the practice of leader development than leadership development. And as noted, the ideal approach looks for ways to connect and integrate across these domains instead of adopting an either/or perspective (Day, 2000). One context in which leadership and leader development is critically important is the military. It is generally recognized that a military branch such as the US Army is involved every day in training soldiers and growing leaders. In an attempt to compile and disseminate what the US military - and in particular the Army Research Institute - was doing in the way of theory-building and research around leader development, an edited book was published on the topic of leader development for transforming organizations (Day, Zaccaro, & Halpin, 2004), The overall approach of the book was based on the premise that developing leaders at all organizational levels is an effective means of transforming organizations. In other words, transforming individuals through leader development also transforms organizations. Among the various book sections, there were foci on cognitive skill development, developing practical and emotional intelligence, and enhancing team skills. The book concluded with a chapter examining challenges to developing a science of leader development (Day & Zaccaro, 2004), which included (a) conceptualizing and measuring change (since change is at the heart of development), (b) criterion development (moving beyond job performance to conceptualizing and measuring actual development), (c) incorporating new and diverse research methods, (d) incorporating more developmental theory info the approaches to leader development, and (e) addressing the role of context in development among other potential challenges. Although this book is helpful in drawing together a number of different perspectives on leader development, it falls short of offering much in the way of a comprehensive and integrative theory - or substantive theorizing (Weick, 1995) -about leader development. In an attempt to develop more rigorous theory on the topic, Lord and Hall (2005) proposed a model of leadership skill development based on research relating leadership to social identity and values as well as the acquisition of domain-specific expertise (e.g., Ericsson, Charness, Feltovich, & Hoffman, 2006; Ericsson, Krampe, & Tesch-RSmer, 1993; Ericsson & Lehmann, 1996). At the core of their theoretical approach is that skill development involves changes in a leader's information processing activities, moving from micro-level skills into higher levels of organizations (e.g., competencies) that guide behavior, knowledge, and social perceptions. Important to the development of leadership capacity from novice to intermediate to expert levels are individuals' self-views of leadership. Specifically, as leaders develop, there is an expected shift from individual- to collective-level identities, which represents a movement from surface-level to deep-level structures. One implication of this approach to leader development is that it is thought to occur over an extended time period. As noted in this piece, a classic finding from the expert performance literature is that it takes a minimum of 10 years or 10,000 hours of dedicated practice to achieve expert status in a given domain (Ericsson et al., 1993). Although this has yet to be examined empirically, it is expected that a similar time frame is required to develop the expert leader. Working in the domain space of student leadership development, Komives and colleagues developed a grounded theory approach to identity-based leader development (Komives, Owen, Longerbeam, Mainelia, & Osteen, 2005). The researchers studied a small (N = 13) but diverse sample of college students over several time periods throughout their undergraduate studies. One of their general findings was a gradual shift from a heroic leader-centric view of leadership to one that considered leadership as a collaborative and relational process. This shift is best personified in comparing statements about leadership at early stages in the developmental process ('I am not a leader') to those later on (T can be a leader even when I am not the leader," p. 605, italics in original). The Leadership Identity Development (LID) model that resulted from this grounded theory study was further elaborated and discussed in terms of practice applications (Komives, Longerbeam, Owen, Mainelia, & Osteen, 2006). The work of Komives et al. illustrates how the development of one's leadership skills and identity commences in late adolescence and early adulthood, if not earlier (Schneider, Ehrhart, & Bhrhart, 2002; Schneider, Paul, White, & Holcombe, 1999). Considered in conjunction with the 10-ycar rule to developing expert performance, this work also suggests that an adult lifespan perspective is recommended when considering the development of leadership expertise and expert leaders. Day and Harrison (2007) further developed the themes of multilevel leader identity in describing a proposed development approach incorporating organizational levels as well as levels of development (i.e., leader and leadership development). Their approach postulates that as leaders move up an organization's hierarchy, there is a need to move from an individual to relational and then collective identity (Lord & Hall, 2005). What distinguishes this approach from most others that focus solely on leader development is that it recommends incorporating processes that involve participants in engaging across boundaries (functional, hierarchical, geographical) as a way to both develop collective leader identities as well as to engage in leadership development. A particular feature of this developmental model is that it recognizes that the fundamental needs of leader development change as individuals take on greater-leadership role responsibilities. In the most ambitious theoretical approach to leader development to date, Day, Harrison, and Halpin (2009) proposed an integrative framework linking leadership expertise (or the expert leader) at the most visible level, supported by leader identity and self-regulation processes at a meso level, with adult development at the foundation. Specifically, the selection-optirnization-compen-sation orchestrating process of successful and healthy aging (Baltes, 1997; Baltes & Baltes, 1990) were proposed as the adult lifespan building blocks in terms of setting and sticking to overall life goals (selection), using resources effectively in pursuing life goals (optimization), and responding in effectively adaptive ways when goals are blocked or resources are unavailable (compensation). This latter point in particular is relevant to leader development from an expert performance domain because research has shown that across a wide variety of domains (e.g., chess, medicine, computer programming, physics, sports, and music among others) experts demonstrate maximal adaptations to domain-specific constraints (Ericsson & Lehmann, 1996). Thus, a key process in healthy adult development may also be critical in achieving an expert level of performance in a related domain of leader development. The Day et al. (2009) approach also reinforces the notion of leader development as a lifelong journey that is part of ongoing adult development processes. The integrative approach includes 13 general theoretical propositions and over 90 specific, testable hypotheses that are offered in hopes of motivating researchers to adopt a theoretically grounded approach to future investigations of the leader development process. Empirical tests of aspects of the theory are ongoing (e.g., Day & Sin, hi press). Another approach taken by Luthans, Avolin, Gardner, and colleagues has focused on processes involved in what is termed authentic leadership development (Luthans & Avolio, 2003; also see Chapter 26 in this volume). The definition of authenticity lhat is used to ground their approach involves someone 'owning' their experiences and acting in accordance with those inner thoughts 40 THE SAGE HANDBOOK OF LEADERSHIP and feelings (Avolio & Gardner, 2005). Within this approach, a distinction is drawn between Ihe development of authentic leaders (Shamir & Eilam, 2005) and authenticleadershipdevelopmenr. The latter goes beyond focusing on just the leader to addressing the development of an authentic relationship between leaders and followers, which requires a focus on the shared relationships between leaders and followers rather than each entity separately. At the core of authentic leadership development is positive modeling of authentic leadership lo help create authentic followership (Gardner, Avolio, Luthans, May, & Walumbwa, 2005). Key processes within each of these authentic components (leadership/followership) include self-awareness and self-regulation. Like most theoretically oriented approaches to leadership development in the current literature, there is little in Ihe way of empirically based tests of authentic leadership development or its components. Nonetheless, it is a worthy contribution to the leadership literature arising from the recent interest in positive psychology (Snyder & Lopez, 2002) and positive organizational scholarship (Cameron, Dutton, & Quinn, 2003). Most of what has been discussed thus far has focused mainly on understanding leader development from various theoretical perspectives. There are also recent initiatives designed to understand the leadership development process theoretically. In one such approach, Day, Gronn, & Salas (2004) proposed a model of developing team-based leadership capacity - conceptualized in terms of the amount of shared, distributed, and connected ways of working together - with regard to collectively addressing leadership challenges. The framework developed in this paper flows from an updated IMOI (input-mediator-outptit-input) model of team processes that explicitly recognizes the role of feedback in how teams adapt and perform (Ilgen, Hollenbeck, Johnson, & Jundt, 2005). Refer to Chapter 25 of this volume for an extensive review of Leadership in Teams. The team leadership capacity model begins with an accounting of individual team member resources (i.e., knowledge, skills, and abilities of members) that shapes the amount of teamwork that develops as a function of the formal leader's resources (leadership skills, leadership knowledge, and leadership abilities) and formal developmental interventions that are used. Teamwork serves as a mediator for team learning, which in turn shapes the level of team leadership capacity that develops. This capacity can be used as a resource (input) for the next team performance cycle. From this perspective, team leadership capacity was conceptualized as an emergent state in teams (Marks, Mathieu, & Zacearo, 2001) that provides a vitally important resource, especially when complex adaptive challenges are faced. In theory at least, team leadership capacity can provide necessary resources lo help teams be resilient and adaptive even under the most challenging circumstances. In a related program of theory-building and research, Kozlowski and colleagues have advanced a comprehensive and integrative framework for understanding team leadership and team development. The approach is grounded in the observation that it is difficult to apply prescriptions from existing leadership research to teams operating in complex and dynamic decision-making environments (Kozlowski, Gully, McHugh, Salas, & Cannon-Bowers, 1996). Beginning with the initial Iheory and guidelines for application (Kozlowski et al., 1996), the research team has provided additional conceptual insight regarding the processes associated with team development (Kozlowski, Gully, Nason, & Smith, 1999), empirically examined the effects of feedback on the regulation of individual and team performance across mulliple goals and multiple levels (DeShon, Kozlowski. Schmidt, Milncr, & Wiechmann, 2004), and further integrated team development and adaptation with team learning as emergent group phenomena (Kozlowski & Bell, 2008). Their most recent contribution has elaborated more specifically on the role of the leader in the team development process (Kozlowski, Watola, Jensen, Kim, & Botero, 2009) in terms of helping the team move from relatively novice to expert status and beyond to building adaptive capability in the team. In these latter stages of team development, the team takes on more responsibility for its learning, leadership, and performance. Taken together, this work provides an impressive theoretical and empirical foundation for understanding team leadership and, in particular, how something like adaptive capability - or leadership capacity - develops in teams. Overall, there appears to be a number of promising advances in the theoretical understanding of leader and leadership development. Some of the consistent themes across these various approaches include a focus on developing leadership expertise (Day et al., 2009; Lord & Hall, 2005), various perspectives on the role of leader identity (e.g., Day & Harrison, 2007; Day et al., 2009; Komives et al„ 2005, 2006; Shamir & Eilam, 2005), as well as the development of adaptive leadership capacity in teams (e.g.. Day et al., 2004; Kozlowski et al., 2009). These are all encouraging signs that the field of leadership development is moving beyond a 'best practices' approach to taking a scientific stance in developing theory and theoretically grounded research propositions and hypotheses. It is a sound (and necessary) step in developing a rigorous science for leader and leadership development. LEADERSHIP DEVELOPMENT 41 The next section will examine research published in the firsl decade of Ihe twenty-first century lo review whal empirically based contributions exist to further support the scientific development of the field. Whal is (he evidence that leader and leadership development have begun lo emerge as local topics of scientific research? RESEARCH ON LEADER/LEADERSHIP DEVELOPMENT There are two general approaches that are adopted in studies of leadership development that is broadly stated lo include leader development. The first approach is that of training, in which a particular set of knowledge, skills, or abilities are targeted for intervention and improvement. Included in this training focus is management 'development' lhal primarily emphasizes managerial education (Latham & Seijts, 1998; Wexley & Baldwin, 1986). There are also different approaches to fostering executive development in terms of helping new incumbents 'get up to speed' in their respective positions (Bauer, Erdogan, Liden, & Wayne, 2006). But it should be noted that the latter approach is something else entirely than what is typically considered in traditional management development. Also, although many sources refer lo training and development together (e.g., planned effort by an organization to facilitate the learning of job-related behavior on the part of its employees; Wexley & Latham, 1991), the concept of development has a much different focus than traditional training. Rather than skills or abilities training, or a classroom education focus, development initiatives focus on the more hazy and far-reaching goal of building individual and collective capacity in preparation to meel unforeseen challenges (Day, 2000). In short, training provides proven solutions lo known problems, whereas development helps people to better learn their way out of problems that could not be predicted (Dixon, 1993). An exemplar of a training approach in enhancing Ihe abilities of leaders can be found in the longitudinal, randomized field experiment reported by Dvir, Eden. Avolio, and Shamir (2002). In particular, the researchers were interested to examine whether enhancing transformational leadership ihrough training would impact follower development and performance. Given that the research design was a true field experiment, stronger claims can be made with regard to causality relative lo most other study designs. The focal sample was infantry cadets undergoing officer training in the Israeli Defense Forces and the training intervention consisted of leadership workshops designed lo enhance their leadership skills before becoming platoon leaders. The experimental leaders received transformational leadership training, whereas those in the control group were exposed to 'roulme eclectic leadership training' (p. 737). Another interesting feature of the Dvir et al (2002) study was that the outcomes included both follower performance and follower development. In addition, Ihe researchers examined die effects of the leadership interventions on 'direct' followers as well as 'indirect' followers (i.e., those two levels below Ihe platoon leader). Indirect follower performance was opcrationalized in terms of written and practical performance in areas such as light weapons, physical fitness, and marksmanship (direct follower performance was not measured). Direct and indirect follower development was operationalized in terms of variables such as self-efficacy, collectivistic orientation, extra effort, active engagement, and internalization of moral values. In general, results indicated that those leaders receiving transformational leadership training had a more positive impact on direct followers' development and indirect followers' performance than did leaders in the control group. As noted previously, because of Ihe rigorous experimental design as well as the focus on follower performance and development, this study provides an especially noteworthy example of leadership training research. Compared to Ihe focused and structured approach that is the hallmark of training, development initiatives can seem relatively nebulous. For example, it is taken almost as gospel that experience is the most effective way to develop leadership, at least as reported by managers themselves (McCall & Hollenbeck, 2002; McCall, Lombardo, & Morrison, 1988). The interview-based approach that McCall and colleagues used with diverse samples of executives revealed thai challenging work experiences incorporating novel responsibilities (e.g., 'stretch' assignments) were perceived to be more developmental than classroom experiences or those on-the-job experiences that were more routine and less challenging. This has led these researchers to propose lessons of experience as a way lo understand executive development, with one outcome being an assessment tool lo quantify the developmental components of managerial jobs (McCauley, Rudennan, Ohlotl, & Morrow, 1994). Despile the apparent value of experience -and in particular the motivation and ability lo learn from experience - there is the implicit assumption that the more challenging an experience is the more developmental value il holds. This runs counter lo Iheory and research from the adult learning and dcvelopmenl literatures, which suggest some experiences might be loo 42 THE SAGE HANDBOOK OF LEADERSHIP challenging in terms of putting people 'in over their heads' (Kegan, 1904) such that the ability to learn from the experience is compromised. The notion of potentially diminishing returns of challenging work experiences on leadership skills development was examined recently by DeRue and Wellman (2009). Specifically, the researchers hypothesized that a learning goal orientation would help offset the diminishing returns of developmental challenges by helping individuals to reframe the challenge and the corresponding mistakes as learning opportunities rather than as failed attempts to prove competence (Dweck. 1986; Elliott & Dweck, 1988). Access to feedback was also hypothesized as a factor (hat might offset diminishing returns associated with high developmental challenge, because feedback is essential to learning. Examining 225 on-the-job experiences across 60 managers, DeRue and Wellman indeed found evidence of diminishing returns in the relationship between developmental challenge and leadership skill development. In addition, access to feedback - but not a learning orientation - dampened these curvilinear effects. The implications of these findings highlight the practical importance of deploying individuals to work experiences where there is an optimal amount of developmental challenge. Of course, it is not quite so easy to know where that inflection point is on an individual-by-individual basis in terms of when a challenging assignment becomes too challenging. This is why it is important to maximize feedback availability in general, but especially in job assignments where developmental challenges are expected. Although the hypothesized role of adopting a learning goal orientation was not supported, the results were in the appropriate direction, suggesting that this individual motivational difference is worth considering in future research on leader/leadership development. Yet another research study examined the personal trajectories of development (Raudenbush, 2001) in a student sample (N = 1,315) engaged in service learning projects. The particular approach to leader development in this case was action learning in which self-awareness and skills development occur as part of engaging in meaningful project-based work (Marquardt, 2004; Rcvans, 1980). The overarching goal of action learning is to use project work as a means of enhancing development rather than taking individuals away from their work in order to develop. These projects can be considered as a type of stretch job assignment; furthermore, the action learning approach to development has become increasingly more common in organizations of all kinds (Conger & Xin, 2000). The research approach in question was longitudinal and multilevel in nature. studying within- and between-person changes over the course of approximately 15 weeks (Day & Sin, in press). The underlying assumption was that leader development is a highly individualized process. Therefore, there are individual difference factors that can serve to enhance leader development. This assumption was tested empirically by the authors. The theoretical foundation of the research was based on the integrative model of leader development proposed by Day et al. (2009) that was discussed briefly previously in the chapter. The researchers conceptualized leader identity as a time-varying covariate and modeled it as a longitudinal, within-person variable. Results across four separate time periods suggested that leadership effectiveness (independently rated by the team coach) was positively related to the self-rated strength of leader identity (i.e., the extent that 1 see myself, or identify, as a leader). It is important to remember that this is a within-person effect, so the relevant comparison is with those time periods in which that same individual was less likely to identify as a leader. In short, it acknowledges that individual leader identity is dynamic, just as arc overall developmental processes. At a between-persons level, the researchers conceptualized goal orientation (learning- and performance-oriented) as something that would differentiate among leader development trajectories. As hypothesized, holding a stronger learning goal orientation was associated with higher initial levels of rated leader effectiveness (i.e., intercepts) and also positively related to more effective developmental trajectories (i.e., slopes). Performance goal orientation was related to initial effectiveness levels (positive for a 'prove' performance orientation and negative for an 'avoid' performance orientation) but unrelated to slope differences. These results are consistent with those of DeRue and Wellman (2009) with regard to the individualized nature of leader development and support the notion that there are identifiable factors that can enhance the process (e.g., feedback accessibility, leader identification, learning goal orientation). An interesting aspect of the Day and Sin (in press) study was that the overall developmental trend across all subjects was generally negative, with a slight positive upturn near the end of the project (i.e., curvilinear). Although this may seem inconsistent with implicit notion of development involving positive change over lime, it is entirely consistent with developmental theory. It was fust articulated by the eminent developmental psychologist Paul Baltes in this way: '...any process of development entails an inherent dynamic between gains and losses ... no process of development consists only of growth or progression' LEADERSHIP DEVELOPMENT 43 (Baltes, 1987, p. 611, italics in original). The goal of successful development, therefore, is the maximization of gains and the minimization of losses. Given the overall negative developmental trend in these data, it suggests that action learning was an extremely challenging developmental experience and might have been loo challenging for many of the participants to develop in a consistently positive manner. In further examining this possibility. Day and Sin (in press) used growth mixture modeling (Wang & Bodner, 2007) to determine whether multiple unobserved subpopulatious: with different developmental trajectories existed. Indeed, they found two such subpopulatious: The majority (approximately 90% of the sample) demonstrated the previously described negative and curvilinear effect; however, approximately 10% of them sample demonstrated a positive linear developmental trend. Furthermore, there were significant differences between these subpopulatious on several individual difference measures (e.g., selection-optimization-compensation, core self evaluations, performance goal orientation). In short, there was a small but significant sub-population who apparently did not find the action learning initiative to be so challenging as to interfere with their learning and positive development. These individuals were better able to maximize developmental gains while minimizing corresponding losses. Although it is loo premature to say with any certainty what might be the robust set of factors that can be used to potentially identify those individuals who arc better equipped to take on extreme developmental challenges, further research of this sort holds greal promise for being better able to predict those 'high potential' leaders whose development might be accelerated through especially challenging job assignments or through action learning. Another area of potential relevance to leader development is that of leadership efficacy, defined as a specific form of self-efficacy associated with the level of confidence someone feels as a leader in relevant situations calling for leadership (Hannah, Avolio, Luthans, & Harms, 2008). The underlying thinking with regard to leadership efficacy is that individual differences in factors such as personality and values, along with previous leadership experiences, shape the level of leadership efficacy that is internalized. Having greater leadership efficacy is thought to lead to greater willingness to engage as a leader when the situation calls for it, and is also likely to motivate an individual to practice leadership or seek out challenging leadership assignments (Day et al., 2009). In a test and extension of this general model, Chan and Drasgow (2001) hypothesized that leadership self-efficacy would serve as a mediator of the relationship between individual differences in personality, values, and experience, and the outcome of motivation to lead (MTL). They defined MTL as an individual differences construct 'that affects a leader's or leadcr's-to-be decisions to assume leadership training, roles, and responsibilities, and that affect his or her intensity of efforts al leading and persisting as a leader' (p. 482). From this perspective, it is assumed that individual levels of MTL can change as a function of leadership experience (amount and quality). II was further assumed that MTL is a multidimensional construct consisting of the following components: (a) affective-identity (person likes to lead and sees it as part of personal identity), (b) noncal-culative (person will lead only if not calculative about the costs involved with it), and (c) social normative (lead for reasons of perceived duty or responsibility). In a test of these assertions using military recruits and junior college students, the researchers constructed a measure of MTL and tested the proposed mediational model. Results suggested that those who like to lead and see it as an important part of their identity lend to be extraverted, value competition and achievement, and have more previous leadership experience and higher leadership self-efficacy than their peers. Individuals high in noncalculative MTL do not expect rewards or privileges for leading, but are motivated to lead because they are agreeable in terms of personality and value harmony in the group regardless of their respective levels of leadership experience or self-efficacy. Finally, individuals high in social-normative MTL are motivated by social obligation and a sense of duty, but also tend to be more accepting of social hierarchies and more rejecting of social equably than their peers (as well as having more leadership experience and higher leadership self-efficacy). It is also interesting to note that study results indicated that general mental ability was unrelated to MTL, supporting the notion that social and cognitive abilities arc separate components of a leader's personal resources. In summary, the authors concluded that their approach provides preliminary evidence that personality, values, and experience may be linked to leadership performance through the process of leader development, which is at least partly attributable to individual differences in the type anil degree of motivation to lead. Although this is an impressive research study, one limitation of the proposed model is that it did not consider the possible reciprocal relationships between leadership efficacy and experience, choosing to focus on the direct relationship between past leadership experience and leadership efficacy. It is entirely likely thai future leadership 44 THE SAGE HANDBOOK OF LEADERSHIP LEADERSHIP DEVELOPMENT 45 experience will be a function of the level of leadership efficacy (Day et a!., 2009). Individuals actively choose and shape their experiences and are not merely passive participants. In moving from a focus on leader development to one that is concerned primary with leadership development in teams, Carson, Tesluk, and Marrone (2007) recently examined both internal and external antecedents of the emergence of leadership influence across team members (i.e., shared leadership). Specifically, it was hypothesized that an internal team climate of shared purpose, social support, and voice would be positively related to the level of team leadership. It was also hypothesized that supportive externa] coaching by a team manager would be related to the level of shared leadership in a team, and that external coaching would interact with internal team climate in predicting shared leadership levels (coaching was thought to he more strongly related to shared leadership when the internal team climate is unsupportive). Finally, the level of shared leadership in a team was hypothesized to be positively related to team performance. An interesting measurement feature of the Carson et al. (2007) study is that rather than asking team members or their manager about the level of shared leadership in the team (e.g., Hiller, Day, & Vance, 2006; Pcarce & Sims, 2002), they adopted a social network approach involving a measure of density or the perceived amount of leadership displayed by team members as perceived by all others on the team. Using a sample of 59 MBA consulting teams (N = 348), they found support for all hypotheses about the antecedents of leadership development as well as for the predicted positive relationship between shared leadership and team performance. The theoretical and practical implications of these findings are important in suggesting that both internal and external factors to the team matter' in the development of shared leadership, and that developing leadership, and not just individual leaders, is of concern in developing high-performing teams. Taken together, the research reviewed in this section suggests that significant advances have been made in advancing the empirical science of both leader and leadership development. As expected in such early stages of building a leadership development science, there are more conceptual and theoretical publications than empirical studies in the recent literature. Indeed, this is probably a good thing in terms of having available theoretical foundations to help guide research. Given these (and other) advancements, there is reason for optimism in the field. Attention is next focused briefly on a few important practice-oriented concerns before a final section looking ahead to recommended future directions in leader and leadership development. PRACTICE CONCERNS IN LEADER/ LEADERSHIP DEVELOPMENT One of the most endemic practical issues in the field is the tendency to take an episodic view to development: that is, there is an (implicit) assumption that development occurs only as part of a discrete program or a challenging job experience. What this fails to capture is the more important point that what is learned from the program or experience and how it changes behavior or decision-making in future leadership situations is what really mailers. It is not the experience but the learning from experience that is most important for development. So what will be highlighted in this section are primarily issues that are associated with learning - both individual learning and learning about the impact of the leadership development initiative (i.e., evaluation). Feedback A key finding from the previously mentioned DDI report (Howard & Wellins, 2008) is that participants reported that there sue not enough opportunities to learn on the job. This is tragic because learning should he a daily, ongoing process regardless of the job. Judging from the comments provided in the report to support this finding, it appears that respondents saw learning as closely tied to having a mentor or having access to interesting and challenging job assignments. This perception is not wrong, but it is limited, because one of the most basic tools needed to promote learning is feedback. It is an extremely valuable resource that is underused in many organizational settings. A basic principle in both goal setting and learning theories is that actions devoid of feedback are not as potent as actions with feedback in terms of learning. Given the importance of feedback, it is troubling to consider how many opportunities are missed every day for either giving or receiving feedback. If developing the expert leader requires a minimum of 10,000 hours of intensive, dedicated practice, then the only way that will happen is practice - with feedback - occurs in a daily, continuous, and ongoing manner on the job. And perhaps even more important to learning than negative feedback is positive feedback, because it provides information as to what has been done appropriately in addition to providing reinforcement to motivate a repeat of the behavior in the future. The relative lack of feedback, despite the wealth of feedback opportunities, points to another key finding from the DDI report: managers do not know how to help then reports develop, or they know how but refuse to do it. From a practice perspective, one concrete recommendation is to work with leaders to help them understand the importance of feedback in developing themselves and others, and to also develop the skills and confidence to deliver (and receive) feedback on a regular ongoing basis. Sustainability Just as with any other business initiative, successful leadership development efforts require more than a brilliant plan - they require diligent execution and follow-through. Unfortunately, a majority of such initiatives fail because of weak execution and not because of poor strategy or a weak idea (Howard & Wellins, 2008). The underlying issue concerns the need to make leadership development sustainable and not to rely on an episodic or progTam-focused approach to development. Most leaders acknowledge that the most profound development and learning occurs on-the-job and not in the classroom; however, managers are typically left on their own to try and integrate learning from leadership development programs into a persona] development plan. Strategic leadership development, on the other hand, takes the perspective that leadership development is an ongoing process (Vicere & Fulmer, 1998), which has the distinct advantage of having learning and development occur every day rather than only when programs or other kinds of events or interventions are scheduled. Succession planning To be effective, succession planning needs to go beyond merely identifying potential future leaders to also understanding the developmental needs of these individuals and arranging the appropriate experiences to help them learn and develop. In this manner, succession planning and leadership development are inherently intertwined in the identification and development of leadership talent (Day, 2007). Despite the value of such endeavors, the reality is that most organizations do not have a succession plan in place and those that do lend to have ineffective plans. The primary reason that such plans are ineffective is because they tend to focus mainly on the identification of high-potential talent and ignore the need for ongoing development of these individuals. Without a sound link to ongoing leader development, at best such initiatives constitute only replacement planning (Berke, 2005), which limits their effectiveness. This is because it leads to the likelihood of putting people into promotion positions for which they are ill-prepared. In short, it can turn into a classic case of someone 'in over1 their heads' that contributes to eventual derailment. High potentials A high-potential leader is typically someone who has been identified as possessing the potential to move eventually into a senior leadership position in the organization. In theory, being identified as high potential puts an individual into a special pool of candidates to receive accelerated developmental experiences. But as noted in recent surveys, less than half of all organizations actually had a program to accelerate the development of high-potential leaders (Howard & Wellins, 2008). Again, this can result in at best only a partially successful implementation of succession planning and ineffectual learning and development of high-potential mean leaders. Another common problem is failing to establish a commonly shared understanding of what being a high-potential leader means. Specifically, how is potential conceptualized and defined, and what are the behavioral criteria used to identify a high-potenlial leader? In most succession planning exercises, senior managers discuss and plot the job performance and perceived leadership potential of candidates for eventual promotion to an executive position. Those demonstrating high performance and high potential are considered prime candidates for accelerated development, but there is little evidence that the meaning of potential is shared among senior management (Day, 2009). As a result, past job performance has inordinate influence on who is identified as being high potential, leading to cases in which an individual is not ready for accelerated development. This scenario not only risks the candidate's career through possible derailment but also wastes the financial resources that are invested in development. Evaluation Although it is considered a hallmark of an effective development initiative (Howard & Wellins, 2008), efforts to evaluate the results of such initiatives are often forgotten or ignored. There is a well-known taxonomy of training 46 THE SAGE HANDBOOK OF LEADERSHIP LEADERSHIP DEVELOPMENT 47 outcomes thai includes reactions, learning, behavior, and results (Kirkpatrick, 1975). Unfortunately most of the evaluation efforts are focused on participants' reactions to the developmental program (i.e., 'smile sheets'), with little attention to understanding whether the leader's developmental experience had an impact on his behavior or the organization. Hannum, Martineau, and Reinell (2007) note that the questions key stakeholders (e.g., funding agencies, designers, sponsors, and participants) often have about leadership development include: • Is the investment in leadership development worthwhile? • What difference does leadership development make? • What strategies work best to develop leaders and leadership? • How can developmental initiatives be sustained? The authors concluded: 'The complexity of leadership development requires innovative models and approaches to evaluation ... to answer those questions' (p. xiv). Calling for innovative evaluation models is needed but may be too ambitious, given that leadership development evaluation of any kind is the exception rather than the rule. But an interesting statistic to keep in mind is that programs rated very high in quality were 20 times more likely to evaluate the results of their leadership development initiatives than those rated as very low quality (Howard & Wellins, 2008). Any leadership development initiative that aspires to high-quality status should have comprehensive evaluation designed into it from the beginning. The issues highlighted above represent just a few of the most pressing concerns with regard to the practice and delivery of leadership development initiatives. Overall, it appears that there have been important theoretical and research advances in the development of a science of leader and leadership development, but the practical side of the field appears to have deteriorated since publication of the Day (2000) review. It is unclear whether this is due to rising expectations on the part of participants, a worsening quality in programs and initiatives to develop leadership, or a combination of factors. But whatever the causes, these are serious issues that can undermine any scientifically grounded and well-designed initiative. But one thing is certain and lhat is leadership development is one part of a larger organizational system (Day, 2009) and piecemeal approaches to address these and other practice-oriented issues are unlikely to meet with much success. FUTURE DIRECTIONS IN THE SCIENCE AND PRACTICE OF LEADERSHIP DEVELOPMENT If the field of leadership development is truly going nowhere fast, as argued by Howard and Wellins (2008), then where should it be heading instead? This section takes a brief look at the three interrelated areas of theory, research, and practice in identifying future direction in the Held of leader and leadership development. Theory There has been a recent call in the literature for promoting more integrative theory-building strategies in the general field of leadership, and this certainly applies to leadership development as well. Some have proposed that leadership theory has reached a developmental plaleau and that it needs to move to the next level of integration (Avolio, 2007). One way to do this is by more fully considering the dynamic interplay between leaders and followers as well as taking more fully into account the context in which these interactions occur (Avolio &. Gardner, 2005; Gardner et af, 2005). Another way of thinking about this proposed integrative strategy is in terms of inclu-siveness. For far too long, leadership theory has been mainly about the leader. More integrative theories recognize that the leadership landscape includes leaders, followers, and the situational context as essential ingredients in this dynamic interaction. Along these lines, critical perspectives on leader-centric leadership theory have begun to emerge and gain acceptance in the field. The critical perspectives 'challenge the traditional orthodoxies of leadership and following ... and make the claim that leadership is a process thai goes on between all people and that all people can be involved in leadership' (Jackson & Parry, 2008, p. 83). In this way the critical perspective is entirely consistent with the kind of integrative and inclusive leadership theory being argued for here. Another area of future theoretical interest is in moving towards more integrative and inclusive leadership development of a different kind. Leadership is a dynamic, evolving process and as such it incorporates behaviors, perceptions, decision-making, and a whole host of other constructs. Thus, leadership by nature is an eclectic phenomenon and attempting to conceptualize and study its development from any one theoretical perspective (e.g., motivational, emotional, behavioral) will yield at best limited results. What are needed are more inclusive and integrative perspectives that cut across any number of theoretical domains. One example is the integrative theoretical approach to leader development lhat links the otherwise apparently disparate domains of adult development, identity and self-regulation, and expertise acquisition (Day et al., 2009). Leadership development theory will advance by integrating across multiple domains and disciplines in a more eclectic fashion. Research There are some unique aspects of leader and leadership development that need to be considered in moving forward with a research agenda. One such consideration is that of levels. As noted by Avolio (2004), '[Leadership development is always a multilevel development process' (p. 94, italics in original). Relevant levels to consider are within-pcrson and between-person; the next higher dyadic level, involving relationships with followers, peers, and subordinates; and team and organizational levels. Researchers will need to be very clear as to the correct level in which they are working and to choose the type of research design, measures, and analyses that are most suitable for the respective level(s). And, in particular, it would appear thai cross-level approaches (e.g., individuals within teams, teams within organizations) hold great promise for furthering our understanding of developmental processes. Leadership development is a dynamic and longitudinal process, which inherently involves the consideration of time. It has been argued that we need better theory and more research that explicitly address time and the specification of when things occur (Mitchell & James, 2001). In no area of research is this truer than leadership development. When it comes to something like leader development, it can be conceptualized as a process occurring across the entire adult life span (Day et al., 2009). Clearly there are limits in terms of what any one research study can tackle in terms of time frame; however, acknowledging this feature of leader development will ideally push researchers to give careful attention to when they measure things, how many times they measure, and linking measurement with an explicit framework that lays out when (and how) developmental changes are thought lo take place. This is indeed a high standard for researchers, but il is one that is likely lo reap huge dividends in terms of better understanding leader development and ultimately devising ways to accelerate the process. A final research recommendation is to take into consideration the individualized nature of development. Leaders do not develop in the same way following identical growth patterns. People learn different things from the same experience and some learn the key lessons of experience more readily than others. Methodological and analytical approaches that take a more individualized approach to leader development will likely yield more insighl than those that try to model average trends across a given sample. Raudenbush (2001) has proposed a personal trajectory approach to developmental research. Although it may get a bit messy to conceptualize and model a unique trajectory for every developing leader, there arc other individualized approaches such as growth mixture modeling thai can identify and predict different latent trajectory classes and lhat also allow within-class variation of individuals (Wang & Bodner, 2007). These are powerful techniques that can help researchers better understand the individualized nature of leader development, especially when used in conjunction with informed decisions about lime and the timing of key processes as well as the various levels on which development takes place. Practice The observation that leaders are ill-prepared to handle future challenges is not new. Peter Dracker (1995) noted some time ago: 'At most one-third of such [executive selection] decisions turn out right; one-third are minimally effective; and one-third are outright failures' (p. 22). Thus, even though leadership development is a strategic human capital concern of many organizations, current and past data suggest that it is not being done very effectively. An issue that has challenged the effectiveness of leadership development initiatives is the focus on relatively short-term, episodic-based thinking in terms of how development occurs. Traditional thinking about leadership development has viewed it as a series of unconnected, discrete programs with little assistance in integrating across these developmental episodes (Vicere & Fulmer, 1998). Contemporary thinking about leadership development views it as continuous and ongoing throughout the adult life span (Day et al., 2009). In short, just about any experience has the potential to contribute to learning and development to the extent that it includes aspects of assessment, challenge, and support (Van Velsor & McCauley. 2004). The primary focus in Ihe field is on developing individual leader skills: however, there is no certainty that better leadership will result. After all, leadership involves a dynamic social interaction within a given situational context and that effective followers are needed along with effective 48 THE SAGE HANDBOOK OF LEADERSHIP LEADERSHIP DEVELOPMENT 43 leaders (Hollander, 2009). In addition, leadership development will likely require intervention at a more macro group, team, or organizational level. Brit it is not an either/or proposition; rather, state-of-the-art practices involve determining how to link leader development with more aggregate leadership development to enhance the overall leadership capacity in a collective (Day, 2000; Day lit aL 2004). CONCLUSIONS There is reason to be sanguine about the direction that the field of leader/leadership development is heading, especially on the scientific side of the scientist-practitioner equation. Over the past decade there has been increasing attention paid to theorizing about the leadership development process, especially in terms of moving beyond any single, bounded theoretical approach to conceptualizing leadership. The process is inherently dynamic and eclectic, so it makes sense to build theoretical frameworks that reflect these features. Although the sheer number of research-related publications on leader and leadership development is still relatively small, it is a growing area that is already contributing to a better empirically based understanding of some impoiTant aspects of the leadership development process. It is a daunting task going forward because of the lengthy time frame involved in the development of leaders and because of all the interrelated issues that can potentially affect development. But rather than posing as a threat, these issues present a wealth of opportunities for researchers. There are any number of issues to investigate, but one thing is certain: singie-shot, survey-based research designs are unlikely to add much value to this nascent leadership development science. Research designs that incorporate multiple measurement perspectives, mixed methods, and a longitudinal component are more likely to be well-suited to providing scientific insight to the leadership development process. Given the recent evidence that the practice of leadership development is slipping in terms of perceived quality and value that is added in organizations (Howard & Wellins, 2008), it may be time to take a step back and rethink what is needed to better support an evidence-based approach to leadership development. What may be most needed to help motivate this cause is not only continuing interest in the field theoretically and empirically but also efforts devoted to translating ideas into action and science into sound practice. REFERENCES Avolio, B. J. (2004). Examining the Full Range Model oi leadership: Looking back to transform forward. In D. V. Day, S. J Zaccaro 8 S. M. Halpin (eds), Leader development for transforming organizations: Growing leaders for tomorrow (pp. 71-98). Mahwah, NJ: Erlbaum. Avolio, B. J. (2007). Promoting more integrative strategies for leadership theory-building. American Psychologist, 62, 25-33. Avolio, B. I. & Gardner, W, L. (2005). Authentic leadership development: Getting to the roots of positive forms of leadership. The Leadership Quarterly, 16, 315-338. Baltes, P. B. (1987). Theoretical propositions of life-span developmental psychology: On the dynamics between growth and decline. Developmental Psychology, 23, 611-626. Baltes, P. B. (1997). On the incomplete architecture of human ontogeny: Selection, optimization, and compensation as foundation of developmental theory. American Psychologist, 52, 366-380. Baltes, P. B. S Baltes, M. M. (1990). Psychological perspectives on successful aging: The model of selective optimization with compensation. In P. B. Baltes & M. M. Baltes (eds), Successful aging: Perspectives from the behavioral sciences (pp. 1-34). New York: Cambridge University Press. Bauer, f, A„ Erdogan, B, Liden, R. C, & Wayne, S. J. (2006). A longitudinal study of the moderating role of extraversion: Leader-member exchange, performance, and turnover during new executive development. Journal of Applied Psychology, 91, 298-310, Berke, 0, (2005), Succession planning and management: A guide to organizational systems and practices. Greensboro, NC: Center for Creative Leadership. Bolt, J. F. (2007). Mapping the future of leadership development. In J. F. Bolt (ed.), The 2007 Pfeiffer Annual: Leadership development (pp. 3-23). San Francisco, CA: John Wiley & Sons. Cameron, K. S„ Dutton, J. E., & Quinn, R, E. (eds). (2003). Positive organizational scholarship: Foundations of a new discipline. San Francisco: Berrett-Koehler. Carson, J. B„ Tesluk, P. E„ & Marrone, J. A. (2007). Shared leadership in teams: An investigation of antecedent conditions and performance. Academy of Management Journal, 50, 1217-1234. Chan, K.-Y. & Drasgow, F. (2001). Toward a theory of individual differences and leadership: Understanding motivation to lead. Journal of Applied Psychology, 86, 481^198. Conger, J. A. SXin, K. (2000). Executive education in the 21st century. Journal of Management Education, 24, 73-101, Day, D. V. (2000). Leadership development: A review in context. The Leadership Quarterly. 11, 581-613. Day, D. V. (2007). Developing leadership talent: A guide to succession planning and leadership development. Alexandria, VA: Society for Human Resource Management Foundation. Day, D. V. (2009). Executive selection is a process not a decision. Industrial and Organizational Psychology: Perspectives on Science and Practice, 2, 159-162. Day, D. V., Gronn, P., & Salas, E. (2004). Leadership capacity in teams. The Leadership Quarterly, 15, 857-880. Day, D. V. & Harrison, M. M. (2007). A multilevel, identity-based approach to leadership development. Human Resource Management Review, 17, 360-373. Day, D. V., Harrison, M. M., & Halpin, S. M. (2009). An integrative approach to leader development: Connecting adult development, identity, and expertise. Mew York: Routledge. Day, D. V. 8 O'Connor, P. M. G. (2003). Leadership development: Understanding the process. In S. Murphy & R. Riggio (eds), The future of leadership development (pp. 11-28). Mahwah, NJ: Erlbaum. Day. D. V. & Sin, H.-P. (in press). Longitudinal tests of an integrative model of leader development: Charting and understanding developmental trajectories. The Leadership Quarterly. Day, D. V. & Zaccaro, 5. J. (2004). Toward a science of leader development. In 0. V. Day, S. J. Zaccaro & S. M. Halpin (eds), Leader development for transforming organizations: Growing leaders for tomorrow (pp. 383-399). Mahwah, NJ: Erlbaum. Day, D. V., Zaccaro, 5.1., & Halpin, S. M. (eds). (2004). Leader development for transforming organizations: Growing leaders for tomorrow. Mahwah, NJ: Erlbaum. DeRue, D. S. 8 Welliran, N. (2009). Developing leaders via experience: The role of developmental challenges, learning orientation, and feedback availability. Journal of Applied Psychology, 94, 859-875. DeShon, R. P., Kozlowski, S. W, J„ Schmidt, A. M., Milner, K. R,, Si Wiechmann, D. (2004). A multiple-goal, multilevel model of feedback effects on the regulation of individual and team performance. Journal of Applied Psychology, 89, 1035-1056. Dixon, M. M. (1993). Developing managers for the learning organization. Human Resource Management Review, 3, 243-254. Drucker, P. F. (1995). Managing in a time of great change. New York: Truman Talley Books/Dutton. Dvir, T, Eden, D, Avolio, B. J„ & Shamir, B. (2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735-744, Dweck. C. S, (1986). Motivational processes affecting learning. American Psychologist, 41, 1040-1048. Elliott, E. S. & Dweck, C. S. (1988). Goals: An approach to motivation and achievement. Journal of Personality and Social Psychology, 54, 5-12. Ericsson, K. A., Charness, N„ Feltovich, P. J., a Hoffman, R. R. (eds). (2006). The Cambridge handbook of expertise and expert performance. New York: Cambridge University. Ericsson, K. A., Krampe, R. T, Si Tesch-Romer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100, 363-406. Ericsson, K. A. Si Lehmann, A. C. (1996). Expert and exceptional performance: Evidence of maximal adaptation to task constraints. Annual Review of Psychology, 47, 273-305 Fuimer, R. M. & Goldsmith, M. (2001). The leadership investment: How the world's best organizations gain strategic advantage through leadership development. New York: American Management Association. Gardner, W. L, Avolio, B, J., Luthans, F, May, D. R.; 8 Walumbwa, F. (2005). 'Can you see the real me?' A self-based model of authentic leader and follower development The Leadership Quarterly, 16, 343-372. Hannah, S. T., Avolio, B. J., Luthans, F, & Harms, P. D. (2008), Leadership efficacy: Review and future directions. The Leadership Quarterly, 19, 669-692 Hannum, K. M., Marlineau, J. W„ & Reinelt, C. (eds). (2007), The handbook of leadership development evaluation. San Francisco, CA; Jossey-Bass. Hiller, N. J., Day, D. V., Si Vance, R. J. (2006). Collective enactment of leadership roles and team effectiveness: A field study. The Leadership Quarterly, (7,387-397. Hollander, E. P. (2009). Inclusive leadership: The essential leader-follower relationship. New York: Routledge. Howard, A. S Wellins, R. 5. (2008). Global leadership forecast 2008J2009: Overcoming the shortfalls in developing leaders. Pittsburgh, PA: Development Dimensions International. llgen, D. R., Hollenbeck, 1. R„ Johnson, M., & Jundt, J. (2005). Teams in organizations: From l-P-0 models to IMOI models. Annua! Review of Psychology, 56, 517-543. lackson, B. & Parry, K. (2008). A very short, fairly interesting and reasonably cheap book about studying leadership. London: Sage. Kegan, R, (1994). In over our beads: The mental demands of modem life. Cambridge, MA: Harvard University Press. Kirkpatrick, D. L. (1975). Evaluating training programs: A collection of articles from the Journal of the American Society for Training and Development. Madison, Wl: American Society for Training and Development. Komives, S. R., LDngerbeam, 5. D., Owen, J. E., Mainella, F. C, S Osteen, L (2006). A leadership identity development model: Applications from a grounded theory. Journal of College Student Development, 47, 401-418. Komives, S. R., Owen, J. E., Longerbeam, S. D., Mainella, F. C, S Osteen, L, (2005). Developing a leadership identity: A grounded theory. Journal of College Student Development, 46, 593-611. Kozlowski, S. W. J. & Bell, B. S. (2008). Team learning, development, and adaptation. In V. I. Sessa & M. London (eds), Work group learning: Understanding, improving & assessing how groups learn in organizations (pp. 15-44). New York: Erlbaum. Kozlowski, S. W. J., Gully, 5. M„ McHugh, P. P., Salas, E„ & Cannon-Bowers, J. A. (1996). A dynamic theory of leadership and team effectiveness: Developmental and task contingent leader roles. Research in Personnel and Human Resources Management, 4, 253-305. Kozlowski, S. W. J., Gully, S. M., Nason, E. R„ & Smith, E. M. (1999). Developing adaptive teams: A theory of compilation and performance across levels and time. In D. R. llgen a E. D. Pulakos (eds). The changing nature of work performance: Implications for staffing, personnel actions, and development ipp. 240-292). San Francisco. CA: Jossey-Bass. Kozlowski, S. W. J„ Watola, D. J„ Jensen, J. M„ Kim, B. H., & Botero, I, C. (2009). Developing adaptive teams: A theory 48 THE SAGE HANDBOOK OF LEADERSHIP leaders {V\ develop; 'j. more r'^j ^ lir . Salas, G. F. Goodwin & ss in complex organiza-■es and approaches adership develop-j and industry solutions. ,1. (I99S), Management development^ H. Thierry & Associates (eds), _.r work and organizational psychology Vol. 3, pp. 257-272). Hove, UK: Psychology ..■ss/Erlbaum. LorrJ, R. G. & Hall, R. J. (2005). Identity, deep structure, and the development of leadership skill. The Leadership Quarterly, 16, 591-615. Luthans, F. 8 Avolio, B. (2003). Authentic leadership development. In K. S. Cameron, J. E. Button & R. E. Quinn (eds), Positive organizational scholarship: Foundations of a new discipline (pp. 241-258). San Francisco, CA: Berrett-Koehler. McCall, M. W., Jr & Hollenbeck, G, P. (2002). Developing global executives: The lessons of international experience. Boston, MA: Harvard Business School. McCall, M. W., Jr., Lombardo, M. M.( S Morrison, A. M. (1988). The lessons of experience: How successful executives develop on the job. Lexington, MA: Lexington Books. McCauley, C. D„ Ruderman, M. N., Ohlott, P. J., & Morrow, J. E. (1994), Assessing the developmental components of managerial jobs. Journal of Applied Psychology, 79, 544-560. McCauley, C. D. & Van Velsor, E, (eds). (2004), The Center for Creative leadership handbook of leadership development (2nd ed.), San Francisco: Jossey-Bass. Marks, M. A., Mathieu, J, E., S Zaccaro, 5.1. (2001). A temporally based framework and taxonomy of team processes. Academy of Management Review, 26, 356-375. Marquardt, M. J. (2004). Action learning: Solving problems and building leaders in real time. Palo Aito, CA: Davies-Black. Mitchell, T. R. & James, L. R. (2001). Building better theory: Time and the specification of when things happen. Academy of Management Review, 26, 530-547. O'Connor, P. M. G. & Quinn, L. (2004). Organizational capacity for leadership. In C. D. McCauley & E. Van Velsor (eds), The Center for Creative Leadership handbook of leadership development (2nd ed., pp. 417-437). San Francisco: Jossey-Bass. Palus, C. J. & Horth, 0. M. (2004). Exploration for development. In C D. McCauley & E. Van Velsor (eds), The Center for Creative Leadership handbook of leadership development (2nd ed., pp. 438-464), San Francisco: Jossey-Bass. Pearce, C. L, & Sims, H. P., Jr (2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, directive, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6, 172-197. Raudenbush, S. W. (2001), Comparing personal trajectories and drawing causal inferences from longitudinal data. Annual Review of Psychology, 52, 501-525. Revans, R. W. (1980). Action learning. London: Blond & Briggs. Rousseau, D. M, Manning, J,, & Denver, D, (2008). Evidence in management and organizational science: Assembling the field's full weight of scientific knowledge through syntheses. Academy of Management Annals, 2, 475-515. Schneider, B., Ehrhart, K. H., & Ehrhart, M, G. (2002). Understanding high school leaders: II. Peer nominations of leaders and their correlates. The Leadership Quarterly, 13, 275-299. Schneider, B., Paul, M, C, White, S. S., & Holcombe, K. M. (1999). Understanding high school student leaders: I. Predicting teacher ratings of leader behavior. The Leadership Quarterly, 10, 609-636. Shamir, B. & Eilam, G. (2005). 'What's your story?' A life-stories approach to authentic leadership development. The Leadership Quarterly, 16, 395-418. Snyder, C R. & Lopez, S, J, (eds). (2002), Handbook of positive psychology. New York: Oxford University. Van Velsor, E. & McCauley, C, D. (2004). Our view of leadership development. In C. D. McCauley & E. Van Velsor (eds), 77ie Center for Creative Leadership handbook of leadership development (2nd ed., pp. 1-22). San Francisco: Jossey-Bass. Vicere, A. A. & Fulraer, R. M. (1998). Leadership by design. Boston, MA: Harvard Business School. Wang, M. & Bodner, T. E. (2007). Growth mixture modeling: Identifying and predicting unobserved subpopulations with longitudinal data. Organizational Research Methods, 10, 635-656. Weick, K. E. (1995), What theory is nor, theorizing is. Administrative Science Quarterly, 40,385-390. Wexley, K, H. & Baldwin, T. T. (1986). Management development. Journal of Management, 12, Ill-ISA. Wexley, K. N. S Latham, G. P. (1991). Developing and training human resources in organizations (2nd ed.}. New York: HarperCollins. PART II Macro and Sociologica Perspectives m so THE SAGE HANDBOOK OF LEADERSHIP of dynamic team leadership. In E. Salas, G. F. Goodwin & C. S. Burke (eds), Team effectiveness in complex organizations: Cross-disciplinary perspectives and approaches (pp. 113-155). New York: Routledge. Lamoureux, K. (2007, July). High-impact leadership development Best practices, vendor profiles and industry solutions. Oakland, CA: Bersin S Associates. Latham, G. P. & Seijts, G. H. (1998). Management development. In P. J. D. Drenth, H. Thierry & Associates (eds), Handbook of work and organizational psychology (2nd ed„ Vol. 3, pp. 257-272). Hove, UK: Psychology Press/Erlbaum. Lord, R. G. 8 Hall, R. J. (2005). Identity, deep structure, and the development of leadership skill. The Leadership Quarterly, 16, 591-615. tuthans, F. 8 Avolio, B. (2003). Authentic leadership development. In K. S. Cameron, J. E. Dutton & R. E. Quinn (eds), Positive organizational scholarship: Foundations of a new discipline (pp. 241-258). San Francisco, CA: Berrett-Koehler. McCall, M. W„ Jr & Hollenbeck, G. P. (2002). Developing global executives: The lessons of international experience. Boston, MA: Harvard Business School. McCall, M. W., Jr., Lombardo, M. M., & Morrison, A. M. (1988). The lessons of experience: How successful executives develop on the job. Lexington, MA: Lexington Books. McCauley, C. D., Ruderman, M. IM., Ohlott, P, J., 8 Morrow, J. E. (1994). Assessing the developmental components of managerial jobs. Journal of Applied Psychology, 79, 544-560. McCauley, C. D. 8 Van Velsor, E. (eds). (2004). The Center for Creative Leadership handbook of leadership development (2nd ed.). San Francisco: Jossey-Bass. Marks, M. A„ Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and taxonomy of team processes. Academy of Management Review, 26, 356-376. Marquardt, M. J. (2004). 4cf/on learning: Solving problems and building leaders in real time. Palo Alto, CA: Davies-Black. Mitchell, T. R. S James, L. R. (2001). Building better theory: Time and the specification of when things happen. Academy of Management Review, 26, 530-547. O'Connor, P. M. G. & Quinn, L. (2004). Organizational capacity for leadership. In C. D. McCauley S E. Van Velsor (eds), The Center for Creative Leadership handbook of leadership development (2nd ed., pp. 417-437). San Francisco: Jossey-Bass. Palus, C. J. & Horth, D. M. (2004). Exploration for development. In C. D. McCauley & E. Van Velsor (eds). The Center for Creative Leadership handbook of leadership development (2nd ed., pp. 438-464). San Francisco: Jossey-Bass. Pearce, C. L, & Sims, H. P., Jr (2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, directive, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6, 172-197. Raudenbush, S. W. (2001). Comparing personal trajectories and drawing causal inferences from longitudinal data. Annual Review of Psychology, 52, 501-525. Revans, R. W. (1980). Action learning. London: Blond & Briggs. Rousseau, D. M„ Manning, J., & Denyer, D. (2008). Evidence in management and organizational science: Assembling the field's full weight of scientific knowledge through syntheses. Academy of Management Annals, 2, 475-515. Schneider, B„ Ehrhart, K. H., & Ehrhart, M. G. (2002). Understanding high school leaders: II. Peer nominations of leaders and their correlates. J7ie leadership Quarterly, 13, 275-299, Schneider, B„ Paul, M. C, White, S. S„ & Holcombe, K. M. (1999). Understanding high school student leaders: I Predicting teacher ratings of leader behavior. The Leadership Quarterly, 10, 609-636. Shamir, B. & Eilam, G. (2005). 'What's your story?' A life-stories approach to authentic leadership development. The leadership Quarterly 16, 395^118. Snyder, C. R. 8 Lopez, S. J. (eds). (2002). Handbook of positive psychology. New York: Oxford University. Van Velsor, E. 8 McCauley, C. D. (2004). Our view of leadership development. In C. D. McCauley 8 E. Van Velsor (eds). The Center for Creative Leadetship handbook of leadership development (2nd ed„ pp. 1-22). San Francisco: Jossey-Bass. Vicere, A. A. 8i Fulmer, R. M. (1998). Leadership by design. Boston, MA: Harvard Business School. Wang, M. 8 Bodner, T, E. (2007). Growth mixture modeling: Identifying and predicting unobserved subpopulations with longitudinal data. Organizational Research Methods, 10, 635-656. Weick, K. E. (1995), What theory is not, theorizing is. Administrative Science Quarterly, 40, 385-390, Wexley, K. N. 8 Baldwin, T. T. (1986). Management development. Journal of Management, 12, 277-294. Wexley, K. N. 8 Latham, G. P. (1991). Developing and training human resources in organizations (2nd ed.). New York: HarperCollins. PART II Macro and Sociological Perspectives CHARISMATIC LEADERSHIP tt7 Charismatic Leadership INTRODUCTION The term charisma comes from a Hellenistic word ^apiCFucc or kharisma, meaning a 'gift' or 'divine favor' or 'supernatural power'. In ancient times, it was believed that certain individuals such as prophets or religious leaders or healers were given gifts from the gods to help them in their earthly tasks. These were called charismata. The term was adopted by the Christians of the New Testament period to similarly describe Godly gifts given to the faithful. Most commonly referenced among the gifts were notions of prophecy connected with visionary experiences and the ability of prophets to speak in the person of God (or the Holy Spirit). Among the oldest known literary references to charismata are those found in the Bible: Now there are varieties of gifts (charismata).... But to each one (individual) is given the manifestation of the Spirit for the common good. For to one is given the word of wisdom through the Spirit, and to another the word of knowledge, to another faith, and to another gifts of healing, and to another the effecting of miracles, and to another prophecy.... But one and the same Spirit (God) works all these things, distributing to each one individually as He wills. (1 Corinthians, 12, 4-11) Despite the term's long history, it was not used to describe a categoiy of secular leadership until the writings of the German sociologist Max Weber (1864-1920). He was the first to apply the term 'charismatic' to leaders in the secular as well as religious world. His typology of three types of authority in society (the traditional, the rational-legal, and the charismatic) established charismatic leadership as an important term to describe forms of authority based on perceptions of an Jay A. Conger extraordinary individual. In contrast to authority derived from traditions or rules which conferred legitimacy on individuals, the holder of charisma was 'set apart from ordinary men and is heated as endowed with ... exceptional powers and qualities ... [which] are not accessible to the ordinary person but are regarded as of divine origin or as exemplary, and on the basis of them the individual concerned is treated as a leader' (Weber, 1947, pp. 358-359). As the reader can discern, Weber preserved the essence of the earliest meaning of the term — an individual in a leadership role imbued with extraordinary powers. While Weber did not provide a comprehensive theoretical model of this form of leadership, his writings (Willner, 1984) do provide us with elements of the character and the course of charismatic leadership: (1) the condition under which it typically arises (distress); (2) one requirement for its maintenance (mission successes); (3) its likely outcome over time (institutionalization); and (4) some of the means by which charismatic leaders exercise their authority (powers of mind and speech, heroism, magical abilities). Because of Weber's sociological perspective, however, his exposition of the personal attributes and relational dynamics between the leader and followers was minima!. Sometime later, organizational theorists would focus much of their research on these particular gaps. BEYOND WEBER: CONTRIBUTIONS FROM POLITICAL SCIENCE AND SOCIOLOGY It was not until the 1960s that political scientists and sociologists began to explore Weber's ideas on charismatic leadership seriously. There was • particular interest in applying his ideas to understanding the influence of historical figures such as Churchill and Hitler as well as popularly elected leaders from the newly emerging democracies on the African continent. These explorations focused on answering three central questions: • Could and should charisma be extended beyond its original religious context? • Was there a universal 'charismatic personality' ot were there differing attributes among charismatic leaders? • Where was the locus of 'charisma' - was it to be found in the leader's extraordinary qualities, in the larger social context, or in the relationship with followers? Addressing the first question, two political scientists, Karl Loewenstein (1966) and C.I. Friedrich (1961). argued against extending the concept beyond its religious antecedents. Loewenstein felt that true forms of charismatic leadership were not to be found in the modem world but only in cultures with 'magico-rehgious' or primitive ambiance. Friedrich stressed that the term centered on a transcendental call by a divine being. Charismatic authority, he argued, had to remain linked to this original meaning. Their point of view never gai ned momentum and was resolved by the widespread acceptance by both political scientists and sociologists that the term should include both secular as well as religious leaders. The second point of debate concerned the charismatic leader's 'personality'. One camp suggested that a universal set of characteristics could be identified for all charismatic leaders. The other - in particular, political scientists Dow (1969) and Winner (1984) - argued that the search for a universal set of qualities common to charismatic political and religious leaders would not yield decisive results. They pointed to variations in individual personalities that were so great (comparing Gandhi to Hitler to Churchill to Kennedy, for example) that a single charismatic personality type seemed highly improbable. Instead, Willner (1984) argued that charismatic leadership was more effectively explained as a relational and perceptual phenomenon: 'It is not what the leader is but what people see the leader as that count in generating the charismatic relationship' (pp. 14—15). Because societies and groups differ in their definitions of what constitutes extraordinary qualities, the content of leadership images, projected and perceived, would necessarily differ from group to group. It was therefore impossible, Willner contended, to construct a universal 'charismatic personality'. This line of thinking became the dominant position in the field. Regarding the third question of where the locus of charisma resided, sonic (Blau, 1963; Chinoy, 1961; Friedland, 1964) believed that the social and historical context was the critical determinant in the emergence of charismatic leadership. They fell strongly that times of turmoil and revolution were needed to precipitate charismatic leadership. Others (Dow, 1969; Marcus, 1961) argued that charisma resided within the attributes of the charismatic leader -for example, with their visions or ideologies. As the leading proponent of this point of view, Willner's research (1984) showed that charismatic leadership did not need to be the product of a turbulent environment. From an in-depth review of six case studies of charismatic political leaders, she concluded 'Only two. Hitler and Roosevelt, seem to conform sufficiently closely to the preconditions of crisis and psychic distress specified in the conventional formula' (p. 46). From her research, Willner identified four factors that, aided by personality, appear to be catalytic in the attribution of charisma to a leader; (1) the invocation of important cultural myths by the leader; (2) performance of what are perceived as heroic or extraordinary feats; (3) projection of attributes 'with an uncanny or a powerful aura'; and (4) outstanding rhetorical skills (1984, p, 61), From the field's perspective, Willner's research was pivotal in understanding charismatic leadership, for it highlighted (he relational and perceptual dynamics with followers. While context retained the potential to influence these dynamics significantly, it was not the casual factor or a necessary catalyst". In addition to these three areas of debate, the political scientist lames McGregor Burns was examining charismatic leadership through another lens that would become highly influential within the field and beyond. He wanted to explain the follower relationships and their outcomes. In his 1978 book Leadership, Burns had concluded that leaders could be separated generally into two types: the 'transformational' and the 'transactional' (see Diaz-Saenz, Chapter 22, this volume). The transformational leaders were the same leaders described as charismatic by fellow academics. Since most readers will he familiar with Burns' basic ideas, we include only a brief summary here. For Burns, leadership at its essence could be boiled down to the notion of an exchange. Both the leader and the follower had something to offer one another. It was in the nature of what was exchanged, however, that his model came into play. For Burns, transformational or charismatic leaders offered a transcendent purpose as their mission - one which addressed the higher-order needs of their followers. In the process of achieving this mission, both SS THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 09 the leaders and the led were literally transformed or actualized as individuals - hence, the term 'transforming'. Bums (1978) explained: 'The result of transforming leadership is a relationship of mutual stimulation and elevation that converts followers into leaders and may convert leaders into moral agents' (p. 4). At the other end of the spectrum was transactional leadership. Significantly more common of the two forms, transactional leadership was based on a relationship with followers which consisted of mundane and instrumental exchanges: 'The relations of most leaders and followers are transactional -leaders approach followers with an eye to exchanging one thing for another: jobs for votes, or subsidies for campaign contributions. Such [instrumental] transactions comprise the bulk of the relationships...' (p, 4). Burn's conceptualization would later influence the thinking of many scholars in the organizational leadership field. For example, Bernard Bass (1985) built much of his model of 'transformational leadership' around Burn's ideas. Interestingly, the central idea of leadership as an exchange was already present in the organizational and psychology literature. For example, it is central to the leader-member exchange (e.g. Graen & Scandura 1986), operant conditioning (Podsakoff et al., 1982) and path-goal models (House & Mitchell, 1974). In each, the relationship belween leader and led is dependent upon a series of trades or bargains that are mutually beneficial and are maintained so long as the benefits to both parties exceed the costs (Bass, 1970) (In Burns' terms, these exchanges would be 'transactional' not 'transformational'). Missing is the element of higher-order needs being met and die elevation of both the leader and led to a more evolved state of being. This was the critical contribution that Burns brought to the field. Up to that moment in time, the notion of leaders who manage meaning, infuse ideological values, construct lofty goals and visions, and inspire was missing entirely from this literature of leadership exchange. What is intriguing about the influence of Burns then is not so much the notion of leadership as an exchange but the idea that certain forms of leadership create a cycle of rising aspirations which ultimately transform both leaders and their followers. As we will see, Burns' ideas would have great appeal to organizational theorists grappling with the twin issues of organizational change and empowerment in the 1980s. The 'transformational leader' model spoke to both these issues. After all, these were leaders concerned about transforming the existing order of things as well as directly addressing their followers' needs for meaning and personal growth. CONTRIBUTIONS FROM ORGANIZATIONAL BEHAVIOR Interest in research on charismatic leadership among political scientists waned by the late 1970s. A decade later, another group of scholars became intrigued by the subject. These were social psychologists and organizational behavior faculty who resided primarily in business schools. They would undertake the most extensive attempts at investigating charismatic leadership. Several major theories were proposed along with dozens and dozens of empirical investigations of charismatic leadership in organizations. These studies involved a wide range of samples such as middle and lower-level managers (Bass & Yammarino, 1988; Conger & Kanungo, 1994; Hater & Bass, 1988; Koene etal., 1991), senior executives (Agle & Sonnenfeld, 1994; Conger, 1989), US Presidents (House et al., 1991), educational administrators (Koh et al., 1991; Roberts & Bradley, 1988; Sashkin, 1988), military leaders (Koene et a!., 1991; Howell & Avolio, 1993), and students who were laboratory subjects (Howell & Frost, 1989; Kirkpatrick, 1992; Puffer, 1990; Shamir, 1992). In addition, the subject was explored using a wide variety of research methods. For example, there have been field surveys (Conger & Kanungo 1992, 1994, 1.997; Hater & Bass 1988; Podsakoff, et al., 1990), laboratory experiments (Howell & Frost, 1989; Kirkpatrick, 1992), content analyses of interviews and observation (Conger, 1989; Howell & Higgins, 1990), and analyses of historical archival information (House et al., 1991). What is more remarkable than this flowering of research is the relative uniformity of findings despite some differences in theoretical approaches. As Shamir et al. (1993) noted, findings across the board demonstrate that leaders who engage in the behaviors that are theorized to be charismatic actually produce the charismatic effects that the theory predicts. In addition, many of these studies have shown repeatedly that leaders who are perceived as charismatic receive higher performance ratings, are seen as more effective leaders than others holding leadership positions, and have more highly motivated and more satisfied followers than others in similar positions (e.g. see Agle & Sonnenfeld, 1994), The research of organizational theorists can be organized into distinct topic areas of charismatic leadership: (I) the leader's behavior; (2) the followers' behavior and motives; (3) the leader's and followers' psychological profiles; (4) contextual influences; (5) forces that institutionalize various outcomes of the leader-follower relationship; and (6) liabilities of the relationship with charismatic leaders. Leader behaviors Both the greatest amount of theoretical development as well as empirical research to date have been in the area of leader behaviors. This is due in large part to the backgrounds of the majority of researchers. Most have a strong behavioral orientation. Essentially, there arc three groups of researchers who have carved out then- own models—though each has a measure of overlap with the others in the attributes they identify. They are also the ones who have built the most comprehensive theories as well as conducted the greatest amount of empirical research in the field. They are (1) Bernard Bass, Bruce Avolio, and their colleagues; (2) Robert House, Boas Shamir, and their colleagues; and (3) Jay Conger and Rabindra Kanungo. Bass and Avolio As noted earlier, Bass and his colleague Avolio would borrow from Burns the notion of 'transformational leadership' and develop a similar model for organizational leaders. As Bryman (1992, pp. 97-98) has pointed out, their model goes further conceptually than the Burns' original model. Bass conceptualized the transactional and transformational dimensions as separate, whereas Burns defined them as two ends of a spectrum. For Bass, therefore, a Jeader could be both transformational and transactional. In addition, Bass was determined to more precisely identify the actual behaviors that these leaders demonstrated, whereas Burns was content with more of a 'big picture' overview. At the heart of Bass's model of transformational leader is the notion that these leaders are able to motivate subordinates to performance levels that exceed their own and their leader's expectations. Transformational leaders accomplish this by raising the importance of certain goals, by demonstrating the means to achieve them, and by inducing subordinates to transcend their self-interests for the goals' achievement. In the process, they are also stimulating and meeting subordinates' higher-order needs, which in turn generates commitment, effort, and ultimately greater performance. Bass and Avolio (Bass, 1985; Bass & Avolio, 1993) built their model of transformational leadership around four behavioral components of the leader: (1) charisma or idealized influence; (2) inspiration; (3) intellectual stimulation; and (4) individualized consideration. Charisma is defined in terms of both leader's behavior (such as articulating a mission) and followers' reactions (such as trust in the leader's ability) (Bass & Avolio, 1993). However, the emphasis is on charisma's role in enabling the leader to influence followers by arousing strong emotions and identification with the leader. Identifying with the leader reduces follower resistance to change, while emotional arousal creates a sense of excitement about the mission. Bass (1985) argues, however, that charisma alone is insufficient for- transformational leadership: 'Charisma is a necessary ingredient of transformational leadership, but by itself h is not sufficient to account for the transformational process.' (p. 31). While Bass originally treated inspiration as a Subfactor within charismatic leadership, his more recent writings describe it as a separate component designed to motivate. Much of this dimension centers on communication, in that the transformational leader: 'Communicates high expectations, uses symbols to focus efforts, and expresses important purposes in simple ways.' (1990, p. 22) The component of individualized consideration is similar to the early Ohio Stale notions of consideration. It includes providing encouragement and support to followers, assisting their development by promoting growth opportunities, and showing trust and respect for them as individuals. Its role is to bond the leader and the led and to build follower self-confidence and heighten personal development. Intellectual stimulation, the final dimension, is a process whereby the leader increases follower awareness of problems by challenging them with new ideas and perspectives and by influencing followers to creatively rethink their traditional ways of approaching organizational tasks. In Bryman's work Charisma and Leadership in Organizations (1992), the methodological shortcomings of the Bass model have been well highlighted. Since both of the measures to capture transformational leadership (the LBDQ and the MLQ) are based on subordinate ratings, there are potential problems of contamination by implicit leadership theories. Bass and Avolio, for example, discovered that descriptions of the transformational leader are significantly closer to subordinates' images of the ideal leader than transactional leadership. There are also issues about whether respondent's ratings of their leader's behavior are affected by their knowledge of the leader's effectiveness. In addition, there is little appreciation for situational variables or differences. For example, while the research findings show considerable similarity across studies, there is some variance. Yet the implication for situational differences remains unexplored (Bryman, 1992, pp. 128-129). Bryman (1992) also points out that Bass' measure of charisma itself may be a bit flawed. For example, vision is treated as a component of inspirational, rather than charismatic leadership. Yet the bulk of the literature in the field sees 90 THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 91 vision as a component of charismatic leadership. In addition. Max Weber believed that the basis for charismatic leadership was a perception by followers that their leader was extraordinary. At best, only two of Bass' 10 items could be considered to convey this quality. As one might imagine, there has also been some confusion as to the essential differences between the Bass transformational leadership model and other models of charismatic leadership. For one, the role of charisma in the Bass model is very important. In their empirical studies (e.g. Avolio & Yammarino, 1990; Bass, 1985), the component of charisma generally has the strongest correlation of any of the model's dimensions with subordinates' ratings of leadership effectiveness and their own satisfaction. It is clearly the most influential. In addition, as Bryman (1992) notes, the Bass model is built around the leader articulating a vision that excites followers and engaging in behaviors that build intense loyalty and trust. These dimensions overlap considerably with those postulated by charismatic leadership theories. A comparison of the Bass model with other charismatic theories is presented in the next chapter. Such a comparison reveals that, in essence, there is little real difference in behavioral components. In the literature itself, we also see the two terms used interchangeably and sometimes authors describe them as one or even use the label 'charismatic/transformational leadership' (e.g. House & Shamir, 1993; Hunt, 1991). Conger and Kanungo This model builds on the idea that charismatic leadership is an attribution based on followers' perceptions of their leader's behavior. For example, most social psychological theories consider leadership to be a by-product of the interaction between members of a group. The leadership role behaviors displayed by a person make that individual (in the eyes of followers) not only a task leader or a social leader and a participative or directive leader but also a charismatic or non-charismatic leader. The Conger and Kanungo (1999) framework for charismatic leadership is built around a three-stage model of leadership which involves moving organizational members from an existing present state toward some future state. This dynamic might also be described as a movement away from the status quo toward (he achievement of desired longer-term goals. In the initial stage, the leader must critically evaluate the existing situation or status quo. Deficiencies in the status quo or poorly exploited opportunities in the environment lead to formulalions of future goals. But before devising appropriate organizational goals, the leader must assess what resources are available and what constraints stand in [he way of realizing future goals. In addition, the leader musl assess the inclinations, the abilities, the needs, and level of satisfaction experienced by followers. This evaluation leads to a second stage: the actual formulation and conveyance of goals. Finally, in stage three, the leader demonstrates how these goals can be achieved by the organization. It is along these three stages that behavioral components unique lo charismatic leaders can be identified. Conger and Kanungo note that in reality the stages rarely follow such a simple linear flow. Instead, most organizations face ever-changing environments, and their leadership must constantly be revising existing goals and tactics in response to unexpected opportunities or other environmental changes. In terms of the actual behaviors of charismatic leaders, Conger and Kanungo distinguish charismatic leaders from non-charismatic leaders in stage one by their sensitivity to environmental constraints and by their ability to identify deficiencies and poorly exploited opportunities in the status quo. In addition, they are sensitive to follower abilities and needs. In stage two, it is their formulation of an idealized future vision and their extensive use of articulation and impression management skills that sets them apart from other leaders. Finally, in stage three, it is [heir deployment of innovative and unconventional means to achieve their vision and Iheir use of personal power lo influence followers that are distinguishing characterisiics, A more detailed explanation of each stage follows. Charismatic leaders are very critical of the slatus quo. They tend lo be highly sensitive to both the social and physical environments in which they operate. When leaders fail to assess properly constraints in the environment or the availability of resources, their strategies and actions may not achieve organizational objectives. They will be labeled ineffective. For this reason, it is important that leaders are able to make realistic assessments of the environmental constraints and resources needed to bring about change within their organizations. In the assessment stage, what distinguishes charismatic from non-charismatic leaders is the charismatic leaders' ability to recognize deficiencies in the present context. In other words, they actively search out existing or potential shortcomings in the status quo. For example, the failure of firms to exploit new technologies or new markets might be highlighted as a strategic or tactical opportunity by a charismatic leader. Likewise, a charismatic entrepreneur might more readily perceive marketplace needs and transform them into opportunities for new products or services. In addition, internal organizational deficiencies may be perceived by die charismatic leader as platforms for advocating radical change. Because of Iheir emphasis on shortcomings in the system and their high levels of intolerance for them, charismatic leaders are always seen as organizational reformers or entrepreneurs. In other words, they act as agents of innovative and radical change. However, the attribution of charisma is dependent not on the outcome of change but simply on the actions taken to bring about change or reform. After assessing the environment, charismatic leaders can be distinguished from others by the nature of their goals and by the manner in which they articulate them. They are characterized by a sense of strategic vision. Here the word vision refers to an idealized goal that the leader wants the organization to achieve in the future. In order to be perceived as charismatic, leaders not only need to have visions and plans for achieving them but also they mus! be able to articulate their visions and strategies for action in ways so as lo influence their followers. Here articulation involves two separate processes: articulation of the context and articulation of the leader's motivation to lead. First, a charismatic leader must effectively articulate for followers the following scenarios representing the context: (1) the nature of the slatus quo and its shortcomings; (2) a future vision; (3) how the future vision, when realized, will remove existing deficiencies and fulfill the hopes of followers; and (4) the leaders' plan of action for realizing the vision. In articulating the context, the charismalic's verbal messages construct reality such that only the positive features of the future vision and only ihe negative features of the status quo are emphasized. The status quo is usually presented as intolerable, and the vision is presented in clear specific terms as the most attractive and attainable alternative. In articulating these elements for subordinates, the leader often constructs several scenarios representing the status quo, goals for Ihe future, needed changes, and the ease or difficulty of achieving goals depending on available resources and constraints. In their scenarios, the charismatic leaders attempt to create among followers a sense of disenchantment or discontentment with the status quo, a strong identification with future goals, and a compelling desire to be led in the direction of the goal in spite of environmental hurdles. Besides verbally describing the status quo, future goals, and the means to achieve them, charismatic leaders must also articulate them own motivation for leading their followers. Using expressive modes of action, bolh verbal and nonverbal, the3' manifest their convictions, self-confidence, and dedication to materialize whal Ihey advocate. Charismatic leaders' use of rhetoric, high energy, persistence, unconventional and risky behavior, heroic deeds, and personal sacrifices all serve to articulate [heir high motivation and enthusiasm, which then become contagious among their followers. These behaviors form part of a charismatic leader's impression management. In the final stage of the three stage leadership process, effective leaders build in followers a sense of trust in their abilities and expertise. The charismatic leader does this by building trust through personal example and risk taking and through unconventional expertise. Generally, leaders are perceived as trustworthy when they advocate their position in a disinterested manner and demonstrate a concern for followers' needs rather than Iheir own self-inleresl. However, in order to be charismatic, leaders must make these qualities appear extraordinary. They must transform their concern for followers' needs into a total dedication and commitment to a common cause they share and express ihem in a disinterested and selfless manner. They must engage in exemplary acts that are perceived by followers as involving great personal risk, cost, and energy (Friedland, 1964). In this case, personal risk might include the possible loss of personal finances, the possibility of being fired or demoted, and the potential loss of formal or informal status, power, authority, and credibility. The higher the manifest personal cost or sacrifice for the common goal, the greater is the trustworthiness of a leader. The more leaders are able to demonstrate that they are indefatigable workers prepared to take on high personal risks or incur high personal costs in order to achieve their shared vision, the more they reflect charisma in the sense of being worthy of complete trust. Finally, charismatic leaders must appear to be knowledgeable and experts in their areas of influence. Some degree of demonstrated expertise, such as reflected in successes in the past, may be a necessary condition for the attribution of charisma (Weber, [1924] 1947). They demonstrate an expertise in devising effective but unconventional strategies and plans of action (Conger. 1985). House, Shamir et al, In one of the field's earliest writings on charismatic leadership in organizations, Robert House (1977) published a book chapter entitled 'A 1976 Theory of Charismatic Leadership'. It outlined not only the leader behaviors that were possibly associated with charismatic leadership but also certain personal traits and situational variables. In it. House argued that these leaders could be distinguished 92 THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 93 From others by their tendency to dominate, a strong conviction in their own beliefs and ideals, a need to influence others, and high self-confidence. Through emotionally appealing goals and the demonstration of behaviors that aroused followers' own needs for achievement, affiliation, and power, the charismatic leader was able to motivate high levels of task accomplishment. In addition. House theorized that these leaders simultaneously communicated high-performance expectations as well as confidence in their followers' ability to meet such expectations. These actions, in turn, enhanced follower expectations that their efforts would lead to accomplishments. Through role-modeling, charismatic leaders demonstrated the values and beliefs they wished for followers to endorse in order for the mission to be successful. Like most models in the early stages of theory development, it had several important shortcomings. As Yukl (1994) notes, House's description of the influence process was rudimentary, especially in light of the profound influence he argued that these leaders had over their followers. Secondly, his theory was based largely around dyads - the leader and 'the follower' - rather than collectives, which are the basis of organizations. Finally, absent from bis theory were certain components that would appear in later theories such as the notion of self-sacrifice, unconventional behavior, and the use of non-traditional strategies and tactics (Conger, 1989; Conger and Kanungo. 1987). Since that time, House along with several colleagues (House & Shamir, 1993; House et al.. 1991; Shamir et al., 1993) have made revisions to his earlier theory. The most important and significant revision was by Shamir et al., (1993) in an article entitled 'The Motivation Effects of Charismatic Leadership: A Self-Concept Based Theory'. Focused on explaining the profound levels of motivation typically associated with charismatic leadership, they postulated that these motivational effects could best be explained by focusing on the self-concept of the followers. Citing supporting research (e.g. Prentice, 1987). they point out that as human beings we behave in ways that seek to establish and affirm a sense of identity for ourselves (known as the self-concept). What charismatic leaders do is to tie these self-concepts of followers to the goals and collective experiences associated with their missions so that they become valued aspects of the followers' self-concept. In terms of details, their theory hypothesizes that charismatic leadership achieves its motivational outcomes through at least four mechanisms: (1) changing follower perceptions of the nature of work itself; (2) offering an appealing future vision; (3) developing a deep collective identity among followers; and (4) heightening both individual and collective self-efficacy. The processes that Shamir et al., (1993) describe as producing these effects follow in the paragraphs below. Charismatic leaders transform the nature of work (in this case, work meant to achieve the organization's vision) by making it appear more heroic, morally correct, and meaningful. They in essence de-emphasize the extrinsic rewards of work and focus instead on the intrinsic side. Work becomes an opportunity for self- and collective-expression. The reward for individual followers as they accomplish mission tasks is one of enhanced self-expression, self-efficacy, self-worth, and self-consistency. The idea is that eventually followers will come to see their organizational tasks as inseparable from their own self-concepts - that 'action is not merely a means of doing but a way of being' (Yukl, 1994). To accomplish this change in perceptions of work, the charismatic leader uses several means. One of the most important mechanisms, as described by Shamir et al. (1993), is the leader's vision, which serves to enhance follower self-concepts in three ways. First, by offering an optimistic and appealing future, the vision heightens the meaningftilness of the organization's goals. Secondly, the vision is articulated as a shared one, which promotes a strong sense of collective identity. Presumably the vision is also unique and, by stressing that the vision is the basis for the group's identity, the charismatic leader distinguishes his followers from others and further encourages followers to transcend then individual self-interests for those of the collectives. Thirdly, the leader's expression of confidence in followers' abilities tu achieve the vision heightens their sense of self-efficacy. They feel capable of creating a reality out of what is currently a lofty and Utopian set of ambitions. Integral to Shamir et al.'s motivational theory is the charismatic leader's ability to create a deep collective identity. Asjust noted, the shared vision is one of the principal means. In addition, the charismatic leader actively promotes perceptions that only by banding together can group members accomplish exceptional feats. Furthermore, the leader uses his own behavior to increase identification with the collective through the deployment of rituals, ceremonies, slogans, symbols, and stories that reinforce the importance of a group identity. The significance of creating this collective identity is in the follower outcomes thai it is able to produce. Specifically, the authors cite research (Mcindl & Lerner, 1983) indicating that a shared identity among individuals increases the 'heroic motive' and the probability that individual self-interests will be abandoned voluntarily for collective and altruistic undertakings. As a result, as charismatic leaders cultivate a collective identity in their followers' self-concepts, they are heightening the chances that followers will engage in self-sacrificial, collective-oriented behavior. The group identification in essence strengthens the shared behavioral norms, values, and beliefs among the members. All of this ensures a concerted and unified effort on the part of followers to achieve the mission's goals. Finally, Shamir et al. argue that charismatic leaders achieve their extraordinary levels of follower motivation by focusing their efforts on building follower self-esteem and self-worth. They accomplish this by expressing high expectations of their followers and simultaneously great confidence in the followers' abilities to meet these expectations (Yukl, 1994). This in turn enhances the perceived self-efficacy of followers. From the research of Bandura (1986), we know that the sense of self-efficacy can be a source of strong motivation. For example, it has been shown that individuals with high self-efficacy are more willing to expend greater work effort and to demonstrate persistence in overcoming obstacles to achieve their goals. By also fueling a collective sense of self-efficacy, the charismatic leader feeds perceptions of the group that they together accomplish tremendous feats. In addition, when collective self-efficacy is high, members of an organization are more willing to cooperate with each other in joint efforts to realize their shared aims (Yukl, 1994). In this revised theory, what we see is a shift from House's earlier conceptualization where charismatic leadership was viewed more as a dyadic process to one that is a collective process. As Yukl (1994) has noted, the recent theory also places more emphasis on the reciprocal nature of the influence process under charismatic leadership. For example, charismatic leadership is likely to be far more motivational when the leader chooses a vision that is congruent with the followers' own values and identities. Likewise, followers are more likely to select as their leader an individual who espouses their core values, beliefs, and aspirations despite the fact that these may not always be clearly articulated by followers themselves. Finally, a charismatic leadership model proposed by Sashkin (1988) under the label of 'Visionary Leadership' was presented in our book on Charismatic Leadership in 1988. Although his model has received little research attention, it does highlight the importance of visioning behavior, a core element in charismatic leadership. Besides visioning behavior, Sashkin idciiliiied five other behaviors: causing attention of others on key issues through unconventional and creative actions; effective interpersonal communication; demonstrating trustworthiness; showing self-respect and respect toward others; and taking personal risk. Follower dynamics Earlier research on charismatic leaders by political scientists and psychoanalysts (e.g. Downlon, 1973; Kets de Vries, 1988; see Gabriel, Chapter 29, this volume) proposed that the followers of charismatic leaders were more likely to be those who were easily molded and persuaded by such dynamic leaders because of an essentially dependent character. Followers were drawn to a charismatic leader who exudes what they lack: self-confidence and conviction. For example, in a study of the charismatic, religious leader Reverend Sun Moon, Lodahl (1982) found thai followers had greater feelings of helplessness, cynicism, distrust of political action, and less confidence in then sexual identity than a sample of college students. Other studies (e.g. Freemesser & Kaplan, 1976; Galanter, 1982) found followers of charismatic political and religious leaders to have lower self-esteem, a higher intolerance for indecision and crisis, and more experiences of psychological distress than others (see Tourish, Chapter 16, this volume). But these studies were almost entirely conducted on populations of individuals disaffected by society or in contexts of crisis where individuals are needy by definition. In the corporate world, the situation is likely to be quite different. For example, in a large corporation, the subordinate of a charismatic leader may not have chosen voluntarily to belong to that leader's unit. More commonly, bosses are hired or promoted into positions, and the subordinates are already in place. So for subordinates, there is often lillle freedom to select who will lead Ibem. Likewise, a leader may find himself inheriting a staff of confident, assertive employees. In the case of entrepreneurial companies founded by charismatic leaders, followers may be drawn to such contexts because of the challenge and opportunity. They may be seekers of the risk and uncertainty associated with a new venture—quite in contrast to followers who are dependent seekers of certainty. Conger (1989) noted that there have been two popular explanalions for why followers arc attracted to charismatic leaders. The first centered on psychoanalytic notions of the ego. Essentially, the argument goes that followers are attempting to resolve a conflict between who they are and what they wish to become. They accomplish this by substituting their leader as their ideal, or in psychoanalytic terms, their ego ideal. Some 94 THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 35 psychoanalysts (e.g. Downton, 1973) trace this type of need hack to an individual's failure to mature in adolescence and young adulthood. Because of absent, oppressive, or weak parents, individuals may develop a state of identity confusion. Associating emotionally with the charismatic leader is a means of coping with this confusion and achieving maturity. Given that the leader is in essence a substitute parent and model, a powerful emotional attachment is naturally formed by followers. Wishing to garner the leader's attention and affection, followers enthusiastically comply with his wishes. The assumption underlying this scenario of follower—leader dynamics is that followers are fulfilling a pathological need rather than a healthy desire for role models from whom to learn and grow. The second school of thought is that followers are attracted to the charismatic leader because of a more constructive identification with the leader's abilities, a desire to learn from them, a quest for personal challenge and growth, and the attractiveness of the mission. This, of course, is what the theories in the previous section have largely argued. With Bass (1985), it is the opportunity to fulfill higher-order needs. In the Shamir et al. theory (1993), it is an opportunity to have one's self-esteem, self-worth, and self-efficacy enhanced. Conger (1989) found in his study of charismatic leaders in business that subordinates often described the importance of an attraction to (heir leader's self-confidence, their strong convictions in the mission, their willingness to undertake personal risks, and their history of prior accomplishment. As a result, subordinates often felt a sense of fulfilling their own potential as they met their leader's high expectations. In addition, as ouhers have found (e.g. Bass, 1985), the leader's vision offered attractive outcomes that were motivating in themselves. But Conger felt that simple identification and an attractive vision did not fully explain the commitment and motivation that followers demonstrated for their charismatic leaders. Instead, Conger- discovered that the personal approval of the charismatic leader became a principal measure of a subordinate's self-worth. A dependency then developed to the point that the leader largely defined one's level of performance and ability. As Shamir et al. (1993) have also noted, the leader's expression of high expectations set standards of performance and approval while a continual sense of urgency and the capacity to make subordinates feel unique further heightened motivation. Taken together, these actions promoted a sense of obligation in followers to continually live up to their leader's expectations. As the relationship deepens, this sense of obligation grows. The leader's expression of confidence in a subordinate ability in essence creates a sense of duty and responsibility. Subordinates can only validate the leader's trust in them through exceptional accomplishments. Over the long term, a dilemma naturally occurs for many followers. As the subordinate self-worth is increasingly defined in his relationship to the leader, a precarious dependence is built. Without the leader's affirmation, subordinates can feel that they are underperforming and even failing. In addition, there are fears of being ostracized. As one subordinate explained to Conger: There's a love/hate element [in our relationship]. You love him when you're focused on the same issues. You hate him when the contract falls apart. Either you're part of the team or not - there's a low tolerance for spectators. And over a career, you're in and out. A lot depends upon your effectiveness on the team. You have to build up a lot of credibility to regain any ground that you've lost. The dark-side dynamics of this dependence will be discussed further in a later section. There have also been studies of follower performance under charismatic leadership. One study (DeGroot et al., 2000) applied meta-analysis to assess the relationship between charismatic leadership style and leadership effectiveness, subordinate performance, subordinate satisfaction, subordinate effort, and subordinate commitment. Results indicate that the relationship between leader charisma and leader effectiveness is much weaker than reported in the published literature when leader effectiveness is measured at the individual level of analysis and when common method variance is controlled. Results also indicate a smaller relationship between charismatic leadership and subordinate performance when subordinate performance is measured at the individual level (r = 0.31) than when it is measured at the group level (r = 0.49 and robust across studies). The researchers found an effect size at the group level of analysis that is double in magnitude relative to the effect size at the individual level. This suggests that the effects of charismatic leadership are stronger when the leader has similar relationships with each subordinate or uses a single style to relate to each group. When the leader exhibits variable amounts of charisma to subordinates, or al least when the effect is measured at the individual level, the extent of effective leadership is reduced. These results also suggest that charismatic leadership is more effective at increasing group performance than at increasing individual performance. Other moderators were tested, but they did not account for a significant portion of variance in the observed distribution of correlations, suggesting a need for further research into other potential moderators. Mela-analysis examining the effects of charismatic leadership on subordinate effort and job satisfaction revealed lower correlations when multiple methods of measurement were used, with little convergence toward stable population estimates. If charismatic leadership behavior is to produce higher performance outcomes from subordinates, further research is needed to examine how this occurs. The role of context Until very recently, interest in the role of context and situational factors has been limited. This is due largely to the backgrounds of those researching leadership. 'Micro-theorists' (those with a psychological or social psychological orientation) have dominated the field to date. Few researchers with a more 'macro' or sociological perspective have been active in studying leadership. As a result, environmental or contextual investigations have rarely been applied to leadership studies outside of the fields of political science and religion. As such, our knowledge in this area remains poor, and what does exist is largely theoretical and speculative. The most common speculation has been that periods of stress and turbulence are the most conducive for charismatic leadership (this argument is derived from the work of political scientists looking at charismatic leadership in political and religious contexts: see Cell, 1974; Toth, 1981). Max Weber (1968), for example, specifically focused on times of 'crisis* as facilitating environments. The basic assumption is that times of stressful change either encourage a longing among individuals for a leader who offers attractive solutions and visions of the future or that charismatic leaders have an easier time of promoting a transformational vision during times of uncertainty when the status quo appears to no longer function (Bryman, 1992). To date, the most important empirical study to examine situational factors in organizational contexts was conducted by Roberts and Bradley (1988). Using a field study, they looked at a school superintendent who was appointed a stale commissioner of education. In her- role as superintendent, she was perceived by her organization as a charismatic leader, yet as commissioner that perception failed to convey. In Roberts and Bradley's search to explain why the individual's charisma did not transfer, they discovered several essential differences between she two contexts. In terms of the larger environment, the individual's first context, a school district, was one in crisis - confirming the hypothesis that crisis may indeed facilitate the emergence of charismatic leadership. In contrast, the leader's second context, at the state level, was not in a similar state of distress. The public's perception was that their stale schools were basically sound and simply in need of incremental improvements. The individual's authority also differed in the two situations. As a superintendent, she had much mure control and autonomy. At the state level, as commissioner, quite the opposite was true. Her number one priority was political loyalty to the governor. She no longer possessed the freedom to undertake actions as she deemed necessary: instead, they had to be cleared through the governor's office. Her relationships were also different. Whereas the district organization had been small, with limited stakeholders and localized geographically, the situation at the state level was at the opposite end of the spectrum. The agency was far greater in size, complexity, and bureaucracy. The numerous committees and associations in which she had to participate meant that she had little time to build the deep, personal bonds that she had established at the district level. As a result, her impact at the state level was no longer personal and perceptions of her as a charismatic leader did not materialize. From the Roberts and Bradley study, we might conclude that context shapes charismatic leadership in at least two ways. First, an environmenl in crisis is indeed more receptive to leadership in general and is more likely to be open to proposals common to charismatic leaders for radical changes such as those embodied in the superintendent's vision. Secondly, there are structural and stakeholder characteristics of organizations which influence an individual's latitude to take initiative and to build personal relationships which determine perceptions of charismatic leadership. The position of superintendent provided structurally far more autonomy to act than that of commissioner. The less geographically dispersed and more limited number of stakeholders fostered deeper working relationships at the district level and also inspired affection and trust in her leadership. These in turn heightened perceptions of her charisma. With findings like the study of Roberts and Bradley in mind, we can think of the contexts of organizations as divided into an outer and an inner context, the outer being the environment beyond the organization and the inner including the organization's culture, structure, power distribution, and so on. Using this simple framework, it is useful to divide our discussion around these two contextual dimensions. We will start with the external environment. On the issue of whether crisis is the critical external condition, Conger (1993) hypothesized that there could actually be much more variability in environmental conditions than we might think. He argued that charismatic leadership is not 96 THE SAGE HANDBOOK OF LEADERSHIP necessarily precipitated by conditions of crisis and distress. In earlier research looking al charismatic business leaders (Conger, 1989), he found charismatic leaders who were entrepreneurs who operated in environments not so much of crisis but of great opportunity, munificence, and optimism. Instead of crisis being the sole contextual condition, there may instead be at least two conducive environments, one demanding a major reorientation of the existing order because of a perceived state of distress and the other involving the emergence or creation of a new order based on a 'munificence entrepreneurial' context. In addition, Conger argued that more of an interplay exists between the leader and the context. In other words, context is not the key determinant, but rather that the leader and the context influence one another - the relative weight of each's influence vaiying from situation to situation. For example, Willner (1984) found that while examining charismatic leaders in the political arena some were able to induce or create through their own actions the necessary contextual conditions of a crisis. We might be able to find charismatic leaders who are able similarly to foster perceptions of munificence and great entrepreneurial opportunity. Conger also went on to propose that the more conducive the contextual conditions, the less the magnitude or the fewer the number of charismatic attributes perhaps required for a leader to be perceived as charismatic. Similarly, the greater the intensity or number of 'charismatic attributes' of the leader, the need for an existing context say of extreme crisis or entrepreneurial opportunities may diminish as the leader is able to create these perceptions through bis own actions. For example, an ability at articulating unforeseen opportunities or looming problems in a credible manner may facilitate perceptions of a crisis and/or great opportunity. But this is still an area of great speculation in need of research attention. Beyond the limited efforts focusing on the external environments of charismatic leadership described above, there has been only one major theoretical work focusing on contextual conditions within organizations that may influence charismatic leadership. Pawar and Eastman (1997) proposed four factors of organizations that might affect receptivity to transformational leadership. Given our earlier discussion of the overlap between transformational and charismatic forms of leadership, it is worth examining their hypotheses as Ihey may relate to charismatic leadership. The four factors that Pawar and Eastman identified are (1) the organization's emphasis on efficiency versus adaptation, (2) the relative dominance of the organization's technical core versus its boundary-spanning units, (3) organizational structures, and (4) modes of governance. Their model is built around the central notion that transformational or charismatic leadership is essentially about leading organizational change. Organizational contexts that are more conducive to change are therefore more favorable for charismatic leadership. They begin with the notion that organizations are seeking one of two basic goals - efficiency or adaptation. The challenge is that the goals of efficiency and adaptation have conflicting purposes -the fomier requires organizational stability, while the latter is centered on change. In reality, as we know today, most business organizations attempt both simultaneously, and this highlights one of the dilemmas of Pawar and Easman's theory. It is built around idealized polarities which provide a simple elegance in terms of theory building but may not reflect the complexities of reality. Nonetheless, they hypothesize that an efficiency orientation requires goal stability and, necessarily, administrative management or transactional leadership to achieve its goals. During adaptation periods, on the other hand, the leader's role is to overcome resistance to change and to align the organization to a new environment through a dynamic vision, new goals and values. Therefore, organizations with adaptive goals are far more open to charismatic leadership. The authors caution, however, that while adaptive periods are more receptive to leadership, there must be a felt need by organizational members for transformation, otherwise they may accept more administrative management. The second contextual factor - the relative dominance of the technical core versus boundary-spanning units -refers to the fact that an organization's task systems are either more inwardly oriented or more externally oriented. In this case, Thompson (1967) had argued that organizations divide their task systems into two parts: (1) a technical core which performs the work of input processing through the operation of technology and (2) boundary-spanning functions which interface directly with the external environment. Isolated from an ever-changing external world, the technical core develops routines and stability in how it approaches its tasks (Thompson, 1967). In contrast, the boundary-spanning functions are forced to adapt continually to environmental constraints and contingencies and, as a result, can never develop highly standardized or routine approaches (Thompson, 1967). Pawar and Eastman postulate that organizations where boundary-spanning units dominate over the technical core will be more open to transformational and charismatic leadership, since they are more receptive to change. Employing Mintzberg's (1979) typology of organizational structures, Pawar and Eastman CHARISMATIC LEADERSHIP 97 propose that only certain structures will be receptive to leadership. Mintzberg's five 'ideal type' structures are (I) the adhocracy, (2) the simple structure, (3) the machine bureaucracy, (4) the professional bureaucracy, and (5) the divisional structure. Of these five, only two are hypothesized by Pawar and Eastman to be conducive to transformational or charismatic leadership. They are the simple structure and the adhocracy. Specifically, both are felt to be more receptive to organizational change through the promotion of a vision. In Ihe simple structure, the leader or entrepreneur-leader is the source of the organization's vision, and commitment is facilitated by employee loyally to the leader. In an adhocracy structure, the vision is developed through professionals who possess the power, knowledge, and willingness to work collectively (Mintzberg, 1979). It is argued that the three other structures have internal forces which mitigate against an openness to innovative leadership. For example, the machine bureaucracy is dominated by standardized tasks and work processes. Senior managers are obsessed by a control mindset, and lower-level managers are intent only on implementing operational directives from above. As such, there is little concern with innovation and change, which arc potentially threatening to a tightly orchestrated status quo. In the professional bureaucracy, professionals dominate to such an extent that management is simply a support function and marginalized to the role of facilitation. In addition, the professionals in these systems are far less convmitted to the organization than to their own work and profession. As a result, a collective vision is unlikely to be developed either by these self-centered professionals or by a marginalized group of top managers. The divisional structure is also not conducive. Built around two layers in which a headquarters operation governs quasi-autonomous divisions, the focus of the coiporate headquarters is to specify operational goals and lo monitor the divisions' accomplishment of them. The divisions Ihen are concerned with attaining operational goals. Pawar and Eastman argue that since divisional structures are concerned with operational goals, neither group is likely to show great interest in developing a vision. The final factor influencing receptivity to leadership in the Pawar and Eastman model is the mode of internal governance. They start with the assumption that membership in organizations is built around furthering individual members' self-interests (Burns & Stalker, 1961; Thompson, 1967). Yet the aim of transformational and charismatic forms of leadership is for followers to transcend their own self-interests for collective goals. Under Wilkins and Ouchi's (1983) three modes of governance (the market, the bureaucratic, and the clan), the nature of transactions between an organization and its members will differ. Under the market mode, transactions based on the exchange of commitments between Ihe organization and its members are determined by market or price mechanisms. Because an external market shapes commitments, the organization has little incentive lo socialize its members lo defer self-interests. In the bureaucratic mode, a contract for commitments is built around employees accepting organizational authority in return for wages. The organization then monitors compliance through formal monitoring and exchange mechanisms. These become the devices that curb members' self-interests. Under the clan mode, organizational members are socialized in such a way that their own interests and the organization's are aligned as one. In other words, employees still hold their self-interests, but they believe they can fulfil] them through achieving the collective's interests. As such, cultural values and norms shape self-interests. It is therefore the clan mode that is most receptive to transformational leadership since it allows for a merging of individual self-interests with the collective's goals. Institutionalization The institutionalization or routinization of charisma was an issue that intrigued Weber greatly. He believed that charisma was essentially an unstable force. It either faded or was institutionalized as the charismatic leader's mission was accomplished: If [charisma] is not to remain a purely transitory phenomenon, but to take on the character of a permanent relationship forming a stable community, it is necessary for the character of charismatic authority to be radically changed....It cannot remain stable, but becomes either traditionalized or rationalized or both. (Weber [1924] 1947, p. 364) He argued there were strong incentives on the part of charismatic leaders and their followers to transform their movements into more permanent institutions. With successes, the followers began to achieve positions of authority and material advantage. The desire naturally arose to institutionalize these, and so traditions and rules grew up to protect the gains of the mission. Institutionalization is one area where little research has been conducted in the organizational literature. We know almost nothing about the routinization of charismatic leadership. The only major study was conducted by Trice and Beyer in 1986. They examined two charismatic leaders. 98 THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 99 where in one case charisma had routinized and in anolher it had not. Their conclusions were that five key factors were largely responsible for the successful institutionalization of charisma: (1) the development of an administrative apparatus separate from the charismatic leader that put into practice the leader's mission; (2) the incorporation of the leader's mission into oral and written traditions; (3) the transfer of charisma through rites and ceremonies to other members of the organization; (4) a continued identification by organizational members to the original mission: and (5) the selection of a successor who resembles the charismatic leader and is committed to the founder's mission. In the case where charisma did not rouli-nize, these factors were largely missing. From the standpoint of the business world, however, it does appear that charisma is a relatively fragile phenomenon in terms of institutionalization. There are several examples from the management literature where succession dilemmas prevented the routinization of charismatic leadership (e.g. Bryman, 1992, 1993; Conger, 1989). The charismatic leaders in Conger's 1989 study have all since departed from their original organizations due to either promotions, moves to new organizations, retirement, or in one case, death. From informal observations, it is clear that there is little indication of any significant routinization of their charisma in their various organizations. In a 1993 article, Conger noted that one of the group—an entrepreneur—had had some success in that elements of his original mission, values, and operating procedures did institutionalize. But that individual has since left that organization, and a few years ago it was acquired by a much larger firm which has superimposed its own mission, values, and procedures. Today there is little evidence of that initial routinization of the leader's charisma. The leaders in Conger's study who were acting as change agents in large, bureaucratic organizations had practically no long-term impact in terms of institulionalizing their charisma. As Bryman (1993) argues, succession is one of the most crucial issues in routinization. When an organization possesses a charismatic leader, it creates what Wilson (1975) has called as 'charismatic demand'. The dilemma, of course, is that it is highly unlikely that a charismatic leader will be found to replace the original one. Though Bryman (1993) has found one example in a study of a transportation company, such situations appear extremely rare. Instead, what often happens is that a charismatic leader is replaced by a more mana-gerially-oriented individual. Examples of this would be Steven Jobs, who was succeeded by John Sen I ley and Michael Spindler. the succession of Lee Iacocca at Chrysler by Robert Eaton (Bryman, 1993). and Walt Disney's replacement by Roy Disney (Bryman, 1993). Biggart (1989) docs note that among direct selling organizations we often see an attempt to overcome succession problems by either promoting a national sales executive into the leadership role or to 'invest the mission in one's children' (p. 144). Looking at Amway and ShakJee, Biggart discovered that the founder's children assumed active roles in the company in turn fostering a 'charismatic presence'. But he also found that their roles were largely bureaucratic and that the companies had done little to institutionalize the founder's charisma beyond the presence of their children. Given the enormous demands for continual adaptation, owing to competition and strong needs to develop rational and formalized structures, business organizations may simply not be conducive to long-term institutionalization of a leader's charisma. Even if routinization were to be successful, it is no guarantee of continued performance success. As Conger (1993) noted, part of the dilemma is that successors may not possess the strategic skills or other abilities crucial to the firm's future leadership. For example, while the retailer Walmart has apparently institutionalized Sam Walton's values and operating beliefs, a critical issue is whether it institutionalizes his visionary insights into the world of retailing. Just as importantly. Walton's vision was most likely time-bound. So even if his strategic competence were to be institutionalized, it is the product of a specific eta in retailing and therefore may be unable to anticipate the industry's next paradigm shift. The original mission of a charismatic leader is highly unlikely to be forever adaptive. Even elements as simple as institutionalized rituals may themselves become counterproductive over time. Conger (1993) cited the example of IBM, which very effectively institutionalized many of Thomas Watson Sr's values and traditions. Several of these would prove maladaptive only decades later. For instance, Watson's original strong emphasis on sales and marketing would ensure that future company leaders were drawn from these ranks. The price, however, would be in terms of senior leaders' failure to adequately understand the strategic importance of certain new technologies and software systems. A tradition of rewarding loyalty through internal promotions added to the problem, it encouraged inbreeding around the company's worldview and simply reinforced notions of IBM's mainframe mentality and its arrogance. Even simple traditions would lose their original meaning and transform themselves into bureaucratic norms. For example, IBM's traditional corporate dress code of dark suits and white shirts is illustrative. This requirement was intended by Watson to make his salespeople feel like executives. If you dressed like an executive, you would feel like one was Watson's original thinking. Indeed, the dress code did build pride in the early days of IBM. Many decades later, however, this norm would transform into a symbol of rigidity and conformity. It bureaucratized itself as Weber would have guessed. In conclusion, we have little knowledge about this crucial area of charismatic leadership. A limited number of case studies and no systematic longitudinal research have offered us at best tantalizing tidbits of insight. LIABILITIES OF CHARISMATIC LEADERSHIP While the literature has largely been glowing about the effects of charismatic leadership in organizations, there has been some interest in the negative outcomes associated with this form of leadership. For example, Khurana (2002), in his study of the hiring and firing of CEOs at over 850 of America's largest companies, found that corporate board members and executive search consultants placed a strong emphasis on the charisma of CEO candidates. As a result, both groups artificially limited the number of candidates considered to their companies' detriment. The CEO labor market proved to be a closed ecosystem in which selection decisions were based on highly stylized criteria that more often had little to do with the problems a firm was confronting. As a result, the charismatic candidates often failed once in the CEO role. At the same time, the charismatic candidates possessed extraordinary leverage to demand high salaries and power. Since the pool of high-profile charismatic CEOs is limited, such scarcity naturally drove up wages. Jane Howell (1988) proposed a simple, dichot-omous model of socialized and personalized charisma which attempted to address the issue of (he liabilities of certain charismatic leaders. In conjunction with Robert House (Howell & House, 1993), the theory was refined (o propose a set of personality characteristics, behaviors, and effects that distinguished two forms of charismatic leadership. Specifically, socialized charismalics are described as articulating visions that serve the interests of the collective. They govern in an egalitarian, non-self-aggrandizing manner, and actively empower and develop their followers. They also work through legitimate, established channels of authority (o accomplish their goals. On the other hand, personalized charismatic leaders are authoritarian and narcissistic. They have high needs for power driven in part by low self-esteem. Their goals reflect the leader's own self-interests, and followers' needs are played upon as a means to achieve (he leader's interests. In addition, they disregard established and legitimate channels of authority as well as the rights and feelings of others. At the same time, they demand unquestioning obedience and dependence in their followers. While portraying these two forms as dichotomous, Howell and House do acknowledge that a charismatic leader might in reality exhibit some aspects of both the socialized and the personalized characteristics. This latter view is probably closer to reality than their ideal model. As such, a scaler model might be more accurate. Drawing upon actual examples of charismatic leaders, Conger (1989. 1990) examined those who had produced negative outcomes for themselves and their organizations. He found that problems could arise with charismatic leaders around (1) their visions, (2) (heir impression management, (3) their management practices, and (4) succession planning. On the dimension of vision, typical problems occurred when leaders possessed an exaggerated sense of the market opportunities for their vision or when they grossly underestimated the resources necessary for its accomplishment. In addition, visions often failed when they reflected largely the leader's needs rather than constituents or the marketplace or when the leader was unable to recognize fundamental shifts in the environment that demanded redirection. In terms of i mpression management, charismatic leaders appear prone to exaggerated self-descriptions and claims for their visions that can mislead. For example, they may present information that makes their visions appear more feasible or appealing (han they are in reality. They may screen out looming problems or else foster an illusion of control when things are actually oul of control. From the standpoint of management practices, there are examples of overly self-confident and unconventional charismatic leaders who create antagonistic relations with peers and superiors. Some such as Steven Jobs are known to create 'in' and 'out' groups within their organizations that promote dysfunctional rivalries. Others create excessive dependence on themselves and then alternate between idealizing and devaluing dependent subordinates. Many are ineffective administrators, preferring 'big picture' activities to routine work. Finally, as discussed in the section on institutionalization, charismatic leaders often have a difficult lime developing successors. They simply enjoy the limelight too much to share it. To find a replacement who is a peer may be loo threatening for leaders who tend to be so narcissistic. 100 THE SAGE HANDBOOK OF LEADERSHIP CHARISMATIC LEADERSHIP 101 Recently, Daniel Sankowsky (1995) has written about the dilemma of charismatic leaders who are prone to a pathology of narcissism (see Kels de Vries & Balazs, Chapter 28, this volume). Specifically, he has proposed a stage model showing how dark-side charismatics implicate their followers into a cycle of exploitation. First, these leaders offer a grandiose vision and confidently encourage followers to accomplish it. Followers, however, soon find themselves in an untenable position. Because of their leader's optimism, they have underestimated the constraints facing the mission as well as the resources they need but currently lack. As a result, performance inevitably falls short of the leader's high expectations. Wishing to comply with their leader's wishes, however, followers continue to strive. Soon their performance appears substandard as they fall behind. While initially the leader will blame the outside world for undermining the mission, their attention will eventually turn to the followers. Conditioned to accept their leader's viewpoint and not to challenge it, followers willingly receive the blame themselves from their leader. The reverse of the many benefits ascribed to charismatic leaders then occurs. Instead of building their followers' self-worth and self-efficacy, they gradually desU'oy it and create highly dependent individuals. CONCLUSIONS Charismatic leadership is a rich and complex phenomenon. As this chapter suggests, our understanding of the topic lias advanced significantly since Max Weber proposed the first formal theory of charismatic leadership. While political scientists and sociologists grappled with some of the more critical questions of why certain leaders are seen as charismatic, it was the field of organizational behavior that advanced theory and research to the greatest degree. That said, there arc important areas of the topic which are only partially understood to this day. Significantly more research and theory building are required, especially to deepen our understanding of the interaction effects between context and charismatic leadership, institutionalization and succession dynamics, and the liabilities of litis important form of leadership. REFERENCES Agle, B.R. & Sonnenfeld, LA. (1994) Charismatic chief executive officers: Are they more effective? An empirical test of charismatic leadership theory. Academy of Management Best Papers Proceedings, 2-6. Avolio, B. & Yamarino, F.J. (1990) Operationalizing charismatic leadership using a levels ol analysis framework. The Leadership Quarterty, 1, 193-208. Bandura, A. (1986) Social foundations of thought and action: a social-cognitive view. Englewood Cliffs, NJ: Prentice Hall, Bass, B, M. (1970) When planning for others. Journal of Applied Behavioral Science,6, 151-171. Bass, B.M. (1985) Leadership and performance beyond expectations. New York: Free Press. Bass, B.M. (1990) Bass and Stogdill's handbook of leadership: Theory, research and managerial expectations (3rd Ed.). New York: Free Press. Bass, B. M. & Avolio, B. (1993) Transformational leadership: a response to critiques. In M.M. Chemers & R. Ayman (eds), Leadership theory and research: perspectives and directions. New York: Academic Press. Bass, B.M. & Yammarino, F.J.(1988) Leadership: dispositional and situational. (ONfi Tech Rep No 1). Binghamton, NY: State University of New York, Center lor Leadership Studies. Bennis, W.G. Si Nanus, B, (1985) Leaders: rfte strategies for taking charge. New York. Harper & Row. Berlew, D.E. (1974) Leadership and organizational excitement. California Management Review, 17, 21-30 Biggart, N.W. (1989) Charismatic capitalism: direct selling organizations in America. Chicago, IL: University of Chicago Press. Blau, P.M. (1963) Exchange and power in social life. New York: Wiley. Boal, K.B. & Bryson, J.M. (1988) Charismatic leadership: a phenomenological and structural approach. In J.G. Hunt, B. R. Baliga, H. Bryman, A. (1992) Charisma and leadership in organizations. London: Sage. Bryman, A. (1993) Charismatic leadership in business organizations: Some neglected issues. The Leadership Quarterly, 4 (3/4), 289-304. Burns, J.M. (1978) Leadership. New York: Harper. Cell, CP. (1974) Charismatic heads of state: the social context. Behavioral Science Research, 4, 255-305. Chinoy, E. (1961) Society. New York: Random House. Conger, I.A. (1985) Charismatic leadership in business: an exploration study. Unpublished doctoral dissertation, School of Business Administration, Harvard University. Conger, J.A. and Kanungo, R.N. (1992) Perceived behavioral attributes of charismatic leadership. Canadian Journal of Behavioral Science, 24,86-102. Conger, J.A. (1989) The charismatic leader: beyond the mystique of exceptional leadership. San Francisco: Jossey-Bass. Conger, J.A. (1993) Max Weber's conceptualization of charismatic authority: its influence on oiganizational research. The Leadership Quarterly, 4 (3/4), 277-288. Conger, J.A, & Kanungo R.N. (1987) Towards a behavioral theory of charismatic leadership in organizational settings. Academy of Management Review, 12, 637-647. Conger, J.A. S Kanungo, R.N. (1999) Charismatic leadership in organizations. Thousand Oaks, CA: Sage. Conger, J. A. (1990) The dark side of leadership. Organizational Dynamics, 19 (2). 44-55. Dachler, P. 8t Sdiriesheim, CA. (eds), Emerging leadership vistas. Lexington, MA: Lexington Books, pp. 1-28. DeGroot, T, Kiker, D.S., & Cross, T.C. (2000) A meta-analysis to review organizational outcomes related to charismatic leadership, Canadian Journal of Administrative Sciences, 17(4), 356-372. Deluga, R.J. (1995) The relationship between attributional charismatic leadership and organizational climate behavior. Journal of Applied Psychology, 25(18), 1652-1669. Dow, T.E. (1969) A theory of charisma. Social Quarterly, 10, 306-318. Downton, l.V. (1973) Rebel leadership. New York: Free Press. Freemesser, G.F. & Kaplan, H.B. (1976) Self-attitudes and deviant behavior: the case of the charismatic religious movement. Journal of Youth and Adolescence, 5(1), 1-9. Friedland, W,H. (1964) For a sociological concept of charisma Social Forces, 43, 18-26. Friedricb, C.J, (1961) Political leadership and the problem of the charismatic power. Journal of Politics, 23, 3-24. Galanter, M. (1982) Charismatic religious sects and psychiatry: an overview. American Journal of Psychiatry, 139(2), 1539-1548. Graen, G. Si Scandura, T. (1987) Towards a psychology of dyadic organizing. In B.M, Staw & L.L. Cummings (eds), Hater, J.J. & Bass, B.M. (1988) Superiors' evaluations and subordinates' perceptions of transformational and transactional leadership. Journal of Applied Psychology, 73 695-702. House, R.l. (1977) A 1976 theory of Charismatic leadership. In l.G. Hunt & L.L. Larson (eds), leadership: the cutting edge. CarbondaleJL: Southern Illinois University Press House, R.J. and Mitchell T.R. (1974) Path-goal theory of leadership. Contemporary Business, 3, 81-98. House, R.J. & Dessler, G. (1974) The path-goal theory of leadership: some post hoc and a priori tests, In J.G, Hunt & L.L. Larson (eds), Contingency approaches to leadership. Carbondale, IL: Southern Illinois University Press. House, RJ. & Shamir, B. (1993) Toward the integration of transformational, charismatic, and visionary theories. In M. Chemmers Si R. Ayman (eds), Leadership theory and research perspectives and directions. Orlando, FL: Academic Press, pp. 577-594. House, RJ., Spangler, W.D., & Woycke, J, (1991) Personality and charisma in the US presidency: a psychological theory of leader effectiveness. Administrative Science Quarterly 36, 364-396. Howell, J.M. (1988) Two faces of charisma: Socialized and personalized leadership in organizations. In LA. Conger and R.N. Kanungo (Eds) Charismatic leadership: The elusive factor in organizational effectiveness, (pp. 213-236). San Francisco: Jossey Bass. Howell, J. M., & Frost, P. (1989) A laboratory study of charismatic leadership. Organizational Behavior and Human Decision Processes, 43,243-269. Howell, J.M. & Higgins, CA. (1990) Champions of technological innovation. Administrative Science Quarterly, 35, 317-41. Howell, J. M. and Avolio, B.J. (1993) Transformational leadership, transactional leadership, loss of control, and support for innovation: Key predictors of consolidated business unit performance. Journal of Applied Psychology, 78, 891-902 Hunt, J.G. (1991) Leadership: a new synthesis. Newbury Park, CA; Sage. Jahoda, M. (1981) Work employment and unemployment: values, theories and approaches in social research. American Psychologist, 36, 184-191. Kanter, R.M. (1967) Commitment and social organization: a study of commitment mechanisms in Utopian communities. American Sociological Review, 33(4), 499-517. Kanter, R.M. (1979) Power failures in management circuits. Harvard Business Review, 57, 65-75. Katz, J. & Kahn, R.L. (1978) The social psychology of organizations. New York: Wiley. Kets de Vries, M.F.R. (1988) Origins of charisma: ties that bind the leader and the led. In J.A. Conger & R.N. Kanungo (eds), Charismatic leadership: the elusive factor in organizational effectiveness. San Francisco; Jossey-Bass, Khurana, R. (2002) Searching for a corporate savior: the irrational quest for charismatic CEOs. Princeton, NJ: Princeton University Press. Kirkpatrick, S.A. (1992) Decomposing charismatic leadership: The effects of leader content and procession follower performance, attitudes, and perceptions. Unpublished doctoral dissertation, University of Maryland, College Park. Kirkpatrick, S.A. & Locke, E.A. (1996) Direct and indirect effects of three core charismatic leadership components on performance and attitudes. Journal of Applied Psychology, 81, 36-51. Koene, H., Pennings, H., & Schrender, M. (1991) Leadership, culture, and organizational effectiveness. Boulder, CO: Center for Creative Leadership. Koh, W.L., Terborg, J.R. & Steers, R.M, (1991) The impact of transformational leaders on organizational commitment, organizational citizenship behavior, teacher satisfaction and student performance in Singapore. Miami, FL: Academy of Management Meetings, Locke, E„ Kirkpatrick, S., Wheeler, J.K., et al„ (1991) The essence of leadership. New York: Lexington Books, Lodahl, A. (1982) Crisis in values and the success of the Unification Church. Unpublished doctoral dissertation, Cornell University, Ithaca, NY. Loewenstein, K. (1966) Max Weber's political ideas in the perspective of our time. Amherst, MA: University of Massachusetts Press. McClelland, D.C. (1985) Human motivation. Glenview, IL: Scott Foresman. Marcus, J.T. (1961) Transcendence and charisma. Western Political Quarterly. 14(March), 236-241. Meindl, J. Si Lerner, M.J. (1983) The heroic motive: some experimental demonstrations. Journal of Experimental Psychology, 19, 1-20. Mintzberg, H, (1979) The structuring of organizations. Englewood Cliffs, NJ: Prentice Hall. Neilson, E. 1986. Empowerment strategies: balancing authority and responsibility. In S. Srivastra (ed,). Executive power. San Francisco: Jossey-Bass. 102 THE SAGE HANDBOOK OF LEADERSHIP Pawar, B.S. & Eastman, K.K. (1997) The nature and Implications of contextual influences on transformational leadership: a conceptual examination. Academy of Management Review, 22; 80-109. Pettigrew, A.M. (1992). Podsakoff, Todor, & Skov. (1982) Effects of leader contingent and noncontingent reward and punishment behaviors on subordinate performance and satisfaction. Academy of Management Journal, 25, 810-821. Podsakoff, P.M., Mackenzie, S.B., Moorman, R. H. and Fetter. R. (1990) Transformational leader behaviors and their effects on followers' trust in leader, satisfaction, and organizational citizenship behaviors. The Leadership Quarterly, 1, 107-142. Prentice, DA (1987) Psychological correspondence of possessions, attitudes and values. Journal of Personality and Social Psychology, 53(6), 993-1003. Puffer, S.M. (1990) Attributions of charismatic leadership: the impact of decision style, outcome, and observer characteristics. The Leadership Quarterly, 1, 177-192. Roberts, N.C. & Bradley, R.T. (1988) Limits of charisma. In J.A. Conger & R.N. Kanungo (eds). Charismatic leadership: the elusive factor in organizational effectiveness. San Francisco: Jossey-Sass. Salancik, G.R. (1977) Commitment and the control of organizational behavior and belief. In B.M. Staw and G.R. Salancik (eds.), New directions in organizational behavior. Chicago: St. Clair, pp. 1—54. Sankowsky, D. (1995) The charismatic leader as narcissist: understanding the abuse of power. Organizational Dynamics, 57-71. Sashkin, M (1988) The visionary leader. In J.A. Conger and R.N. Kanungo (eds,), Charismatic leadership: The elusive factor in organizational effectiveness. San Francisco: Jossey-Bass. Shamir, B. (1992) Attribution of influence and charisma to the leader: the romance of leadership revisited. Journal of Applied Social Psychology, 22(5), 386-407, Shamir, B. (1995) Social distance and charisma: theoretical notes and an exploratory study. The Leadership Quarterly, 6(1), 19-47. Shamir, B., House, R. & Arthur, M.B. (1993) The motivation effects of charismatic leadership: a self-concept based theory. Organization Science, 4(4), 584. Snyder, M. & Ickes, W. (1985) Personality and social behavior. In G. Lindzey & E. Aronson (eds), Handbook of social psychology. New York: Random House, pp. 883-947. Thompson, J.D. (1967) Organizations in action. New York: McGraw-Hill. Toth, MA (1981) The theory of the two charisma. Washington, DC: University Press of America. Trice, H.M. & Beyer, J.M. (1986) Charisma and its routiniza- tion in two social movement organizations. Journal of Occupational Behavior, 7, 125-138. Walster, E„ Aronson, D., & Abrahams, D. (1966) On increasing the persuasiveness of a low prestige communicator. Journal of Experimental Social Psychology, 2,325-342. Weber, M. (1947) The theory of social and economic organizations (A.M. Henderson ä T, Parsons, trans.; T. Parsons, ed.). New York: Free Press. Weber, M. (1968) Economy and society (1925) (3 vols. ed.). New York: Bedminster. Willner, A.R. (1984) The spellbinders: charismatic political leadership. New Haven, CT: Yale University Press. Wilson, B.R. (1975) The noble savages: the primitive origins of charisma and its contemporary survival. Berkeley, CA: University of California Press. Yukl, G, (1994) Leadership in organizations, 3rd edn, Englewood Cliffs, NJ: Prentice-Hall. Zaleznik, A. & Sets de Vries, M, (1975) Power and the corporate mind. Boston, MA: Houghton Mifflin. 8 Gender and Leadership Linda L. Carli and Alice H. Eagly INTRODUCTION In this chapter we document women's undeirepresentation as leaders and examine various theoretical explanations for women's leadership disadvantage. First, we explore whether the gender gap in leadership can be explained by inherent differences between men and women that endow men with natural leadership ability. We then consider whether women's disadvantage lies in their greater domestic responsibilities and lesser investments inhuman capital in the form of paid work experience and education. We also assess whether women's advancement is obstructed by gender stereotypes and discrimination, which may result in resistance to women's influence and authority. Next, we examine research comparing the leadership styles of men and women to determine whether such differences may provide either gender wilh advantages or disadvantages as leaders, and thus potentially contribute to women's lack of access to leadership. Finally, this chapter evaluates the extent to which organizational structure and culture make it difficult for women to rise into higher-level leadership positions. In our chapter, we present many studies comparing the situations, perceptions, attitudes, and behaviors of men and women. Studies that reveal gender differences do not demonstrate dichoto-mous effects with no overlap of men and women, but rather average overall differences that occur across a variety of situational, cultural, and individual variables. Many differences and similarities have been established meta-analyrically by taking into account the results of all available studies. These effects are often moderated by cither variables, such as ethnicity, religion, country, education, organizational setting, and other factors. In these several respects, effects associated with gender resemble the effects associated with other variables studied by social scientists (e.g., personality traits, attitudes, socioeconomic status, and race). We begin by examining women's current status as leaders. To what extent have women gamed access to leadership and how has their advancement remained blocked? THE UNDERREPRESENTATION OF WOMEN LEADERS There is little doubt that the status of women has improved. Women have steadily increased their numbers in the paid labor force. In the United States, women made up only 39% of the paid workforce in 1973, but 47% by 2009 (US Bureau of Labor Statistics, 2010b, Table 2). Women's incomes have also risen: in 2009, for full-time US workers, women earned 80 cents for every dollar that men earned -up from only 62 cents in 1979 (US Bureau of Labor Statistics, 2010a). Across all organizations in the United States, women constitute 51% of those in professional and managerial positions, 37% of managers, and 25% of chief executives (US Bureau of Labor Statistics, 2010b, Table 11). Women's advancement is also apparent in politics and public sector jobs. In 2010 in the United States, women hold 17% of the Senate seats, 17% of the seats in the House, 12% of the governorships, and 23% of state executive offices (Center for American Women and Politics, 2010a). There are record numbers of women in state legislatures (Center for American Woman and Politics, 2010b) as well as in the US Congress (Center for American Woman and Politics, 2010c). Similarly, in the