Transformational, Transactional and Servant Leadership: Three Models of Influence, and Why the Most Popular One Is the Hardest to Measure

James MacGregor Burns was a historian and political scientist, not an organizational researcher, and his 1978 book Leadership drew its central cases from American presidents and global political movements rather than from firms. That context matters, because the distinction Burns introduced carries political and moral weight that later management writing often strips away. Burns described transactional leadership as an exchange: one person takes the initiative in making contact with another for the purpose of trading valued things, a legislator trading a vote for a favor, a manager trading pay for output. Nothing about a transactional exchange is inherently inferior in Burns's account; it is simply bounded. Once the exchange is completed, the relationship between the parties has not fundamentally changed, and each side's interests remain separate from the other's.

Burns set this against what he called transforming leadership (the field later settled on "transformational," largely following Bass's usage), which he described as a relationship in which leader and follower raise one another to higher levels of motivation and, in his words, morality. For Burns, this was not simply about inspiring people to work harder; it required the leader to engage followers' higher-order needs and convert followers' separate interests into a shared purpose, changing what followers actually wanted, not just how hard they pursued what they already wanted. Burns built an explicit moral test into this: transforming leadership had to serve followers' genuine, considered interests, not simply move them toward whatever the leader wanted. A leader who manipulated followers toward ends that did not serve them, however inspiring the method looked from outside, was not exercising transforming leadership by Burns's own definition.

Burns treated transactional and transforming leadership as two ends of a single continuum, not as separate, independently varying dimensions; a leader became more transformational by becoming less transactional. That turned out to be one of the first things Bass changed when he adapted Burns's distinction for organizational research, and it is worth flagging as an assumption rather than a settled fact, because the empirical work discussed in Section 2 rejected the continuum model in favor of treating the two as separate dimensions a leader could score highly or poorly on independently. Burns's framework also had no measurement instrument attached to it. It was an interpretive, narrative argument built on historical case material, in the tradition of political biography rather than psychometrics, which gave it genuine explanatory richness and no way, on its own terms, to test whether a given leader was really transforming followers or simply appeared so from a sympathetic vantage point, a problem Section 3 returns to in a more technical form.

Bernard Bass took Burns's distinction and did something Burns had not attempted: he built a model specifically for organizational leadership, broke each side of Burns's continuum into separate, named components, and set out to measure all of them with a standardized instrument. In Leadership and Performance Beyond Expectations (Bass, 1985) and in a shorter, more widely read statement published five years later (Bass, 1990), he described transformational leadership through four components the literature still calls "the four I's": idealized influence, a leader behaving in ways that earn identification and emulation, acting as a model followers want to resemble; inspirational motivation, articulating an appealing vision and communicating confidence that it is achievable; intellectual stimulation, prompting followers to question their own assumptions and approach familiar problems in unfamiliar ways; and individualized consideration, attending to each follower's specific developmental needs rather than treating the team as a uniform unit.

On the transactional side, Bass specified three further components rather than leaving transactional leadership as one undifferentiated category. Contingent reward is the conventional exchange Burns described: clear performance expectations tied to a real consequence, delivered reliably. Management-by-exception active involves a leader watching for deviations from a standard and intervening before they become serious. Management-by-exception passive involves waiting until a standard has already been violated before stepping in, a more reactive variant that performed worse than the active form in most of Bass's own data. A seventh, residual category, laissez-faire leadership, described the practical absence of leadership: a leader who avoids decisions, delays responses, and does not engage with followers' needs at all. Bass treated laissez-faire not as a mild or neutral style but as the category associated with the worst outcomes in his research, consistently below even passive management-by-exception.

The change Bass made to Burns's model that matters most for how the theory gets used in practice was replacing the single continuum with two independent dimensions. In the 1990 Organizational Dynamics paper, Bass argued explicitly that transformational and transactional leadership are not opposite ends of one scale; a leader can score high or low on each independently, and the most effective leaders in his data were not purely transformational but scored highly on both. He called the pattern behind this the augmentation effect: transformational behaviors added measurable variance in follower effort and performance on top of what contingent reward already explained, rather than substituting for a working transactional relationship. The practical implication is specific, not generic: a manager whose basic contingent-reward practices are inconsistent, promises linked to performance that do not reliably materialize, gains comparatively little from inspirational rhetoric layered over that inconsistency, because the augmentation effect describes addition to a functioning base, not compensation for a broken one.

Bass and his long-time collaborator Bruce Avolio built the Multifactor Leadership Questionnaire, most widely used in its "5X" short form, to measure all seven components plus laissez-faire from two vantage points: a leader's self-rating and, usually weighted more heavily in research use, ratings from the leader's own subordinates. The ambition behind the MLQ was a genuine advance over Burns's narrative method: instead of arguing from a handful of political biographies, Bass wanted a scale that could be administered across organizations, scored consistently, and tested against outcome measures in a replicable way. For roughly three decades the MLQ became the dominant quantitative instrument in leadership research, and a large share of what the management literature now reports as evidence for transformational leadership is, more precisely, evidence based on MLQ scores correlating with other measures. That dependence on a single instrument, and what happens when the instrument itself is scrutinized closely, is the subject of Section 3.

Transformational leadership is, by a reasonable case, one of the most over-praised constructs in popular management writing, and the empirical literature on the MLQ gives that judgment real support rather than just rhetorical force. Daan van Knippenberg and Sim Sitkin's 2013 review in the Academy of Management Annals, one of the most thorough critical assessments of the charismatic-transformational leadership literature to date, identified four specific problems rather than one general complaint. First, the construct lacks a clear conceptual definition: successive versions of the model add or rearrange dimensions (idealized influence alone has been split into "attributed" and "behavioral" variants in some later work) without a principled account of how the dimensions combine into the larger construct, or why these particular behaviors, rather than others a leader might display, were chosen to represent it.

Second, and most directly relevant to anyone using MLQ-style instruments to evaluate real leaders, is a discriminant validity problem: in many samples, the four transformational subscales correlate with each other so highly that treating idealized influence, inspirational motivation, intellectual stimulation and individualized consideration as four separable psychological processes is difficult to sustain statistically. In practice, a single underlying impression, something close to general approval of the leader, may be driving scores on all four subscales at once, which would mean the MLQ's four-factor structure is measuring one thing under four labels rather than four distinct behaviors. Contingent reward, nominally transactional, has also been found correlating almost as strongly with the transformational subscales as the transformational subscales correlate with each other in some studies, further blurring the categorical line Bass drew between the two sides of his model.

Third, van Knippenberg and Sitkin flagged a causal-direction problem built into how most of the research is actually conducted. The typical study asks followers to rate their leader's transformational behaviors and, often in the same survey, to report on outcomes such as their own engagement, commitment or perceived unit performance. Because both measures come from the same follower at the same time, a positive correlation is consistent with the leader's behavior causing the follower's favorable state, but it is equally consistent with the follower's already-favorable state coloring how they describe the leader. This is not a hypothetical worry. James Meindl, Sanford Ehrlich and Janet Dukerich's 1985 study of what they called the "romance of leadership" found a specific, demonstrated bias: when research participants were given identical organizational performance outcomes and asked to evaluate the quality of leadership behind them, outcomes attributed to leadership were rated more favorably than the same outcomes attributed to other causes. Applied to MLQ research, a unit that happens to have had a good quarter may give its leader more generous ratings on inspirational motivation and idealized influence than an otherwise identical leader whose unit had an ordinary quarter, independent of any actual difference in what either leader did. Causality, in a meaningful share of the published evidence, may run substantially from outcome to perceived style rather than, as popular management writing usually assumes, the other way around.

Fourth, and following from the first three, van Knippenberg and Sitkin argued that transformational leadership as commonly operationalized risks collapsing into something close to a general measure of leadership a rater approves of, rather than a specific, falsifiable account of a distinct behavioral mechanism. None of this means the construct is worthless, or that the correlational evidence linking MLQ scores to desirable outcomes, replicated across a large number of meta-analyses, should be dismissed outright. It means the evidential basis for treating transformational leadership as a validated, precisely measured behavioral prescription, rather than a broad and somewhat fuzzy descriptive label correlated with good outcomes for reasons only partly understood, is thinner than the volume of management writing built on top of it suggests. A manager reading a leadership assessment built around the four I's is reading something closer to a structured impression than an audited behavioral record.

Robert Greenleaf spent most of his career in management research and development at AT&T before retiring and, in 1970, publishing a short essay titled "The Servant as Leader." He credited the idea's spark to Hermann Hesse's 1932 novel Journey to the East, whose central figure, Leo, accompanies a group on a long journey as a servant, doing menial work and sustaining the group's spirit, and who turns out, once Leo disappears and the journey falls apart, to have been its actual, indispensable leader all along, a status no one recognized while he held it. Greenleaf drew from that image a claim he stated directly rather than derived from data: the great leader is seen as servant first, and the motivation to serve has to precede, and remain primary over, the motivation to lead.

Greenleaf proposed what he called the best test of servant leadership, and it is worth stating closely because later researchers kept returning to it as a benchmark even after building very different measurement instruments around it: do those served grow as persons; do they, while being served, become healthier, wiser, freer, more autonomous, more likely themselves to become servants; and, further out, what is the effect on the least privileged in society, do they benefit, or at least are they not further deprived. That last clause, about the least privileged, is easy to drop when the idea gets summarized for a management audience, but it was integral to Greenleaf's argument: he was not simply describing an interpersonal style that happened to make direct reports feel good, but a standard with an explicit social dimension, extending past the leader's own team to people the organization affects but does not directly employ.

The contrast with Burns and Bass is worth making precise rather than leaving as a vague gesture toward "both are about elevating people." Burns's transforming leadership was already a normative claim, not a value-neutral description, but Bass's 1985 operationalization deliberately built a measurable, organizationally generic instrument on top of it, something that could be scored the same way whether the leader in question answered to shareholders, voters, or a congregation. Greenleaf's 1970 essay and his fuller 1977 book, Servant Leadership: A Journey into the Nature of Legitimate Power and Greatness, did not attempt that kind of generalized instrument at all; they are reflective essays, closer in register to theology and humanistic psychology than to organizational psychometrics, built on Greenleaf's own decades of corporate observation rather than a validated survey administered across a research sample. That origin is not a defect, but it does mean servant leadership entered the management literature roughly fifteen years before anyone tried to measure it the way Bass had already measured transformational leadership, and the field spent a long stretch after that trying to catch the construct up to the same empirical standard, not always successfully, as Section 5 describes.

The empirical turn in servant leadership research did not arrive until nearly four decades after Greenleaf's original essay, and 2008 was its most productive single year. Robert Liden, Sandy Wayne, Hao Zhao and David Henderson (2008) built and validated a 28-item measure organized around seven dimensions: emotional healing, helping followers recover from personal and professional setbacks; creating value for the community, orienting the organization's activity toward benefiting the wider community, not only its own performance; conceptual skills, understanding the organization and the follower's work well enough to provide genuinely useful support rather than generic encouragement; empowering; helping subordinates grow and succeed; putting subordinates first; and behaving ethically. Tested against existing, already-validated measures in the same study, servant leadership as they measured it predicted community citizenship behavior, in-role performance and organizational commitment with an increment beyond what transformational leadership and leader-member exchange, both already covered earlier in this module, had already explained. That incremental finding is the strongest single piece of evidence that servant leadership, measured this way, is not simply a repackaging of transformational leadership or LMX under a different name.

Dirk van Dierendonck's 2011 review in the Journal of Management took stock of what had by then become, in his own account, a crowded field: several competing servant leadership measures had been published in a short span, each built around a different number of dimensions and a somewhat different emphasis, Liden et al.'s seven-dimension scale among them but not the only one. Rather than adopting any single existing instrument as definitive, van Dierendonck proposed a consolidating model built around characteristics that recurred across the competing measures: empowering and developing people, humility, authenticity, interpersonal acceptance, providing direction, and stewardship, acting as a caretaker of the organization's resources and purpose on behalf of others rather than for personal gain. The review's own framing of this proliferation as a problem, not merely an interesting diversity of approaches, is worth taking seriously: when different research teams measure servant leadership with non-equivalent instruments, findings from one study do not cleanly accumulate with findings from another, a genuine obstacle to the kind of cumulative evidence base the MLQ, for all its own measurement problems, had already built for transformational leadership by 2011.

Servant leadership's empirical limitations are not identical to transformational leadership's, but they rhyme with them in at least two places. Much of the supporting evidence, including Liden et al.'s own validation study, relies on follower-rated survey data collected cross-sectionally, which carries the same causal-direction ambiguity Section 3 described for the MLQ: a follower who already has a good relationship with a leader may rate that leader highly on putting subordinates first partly because the relationship already feels good, independent of specific, identifiable servant behaviors the leader displayed. And the dimensions that make up servant leadership, particularly putting subordinates first, behaving ethically and empowering, overlap substantially with content found in authentic leadership and ethical leadership research, constructs this series takes up directly in Lesson 30, raising a discriminant validity question that parallels the one van Knippenberg and Sitkin raised about the four I's.

A further, more practical limitation deserves honest statement rather than diplomatic silence. Greenleaf's own test asks whether those served grow, become more autonomous and more capable; it does not specify what a servant leader should do when a fast, unpopular, directive decision is genuinely required and there is no time to build consensus or develop anyone through the process. Neither Liden et al.'s measure nor van Dierendonck's synthesis resolves this tension, and it is not obvious that it can be resolved at the level of a general leadership construct at all, since the answer plausibly depends on the specific decision rather than on how servant-oriented the leader generally is. A manager who treats servant leadership as a style that should apply uniformly, regardless of the decision in front of them, is making the same universal-prescription error Section 2 warned against for the augmentation effect, applied to a different framework.

Placed side by side, Burns, Bass and Greenleaf answer three different questions rather than offering three competing answers to the same question, and collapsing them into a single "which leadership style is best" comparison obscures more than it reveals. Transactional leadership, in Bass's specific sense of contingent reward and management-by-exception, answers a narrow and mechanically checkable question: did the leader set clear expectations, link them to a real and reliably delivered consequence, and monitor performance against a standard. It is the least glamorous of the three frameworks and the one most easily audited from the outside, because it describes observable administrative behavior rather than an impression of inspiration or service. Bass's own augmentation-effect finding gives it a specific, testable priority: whatever value the other two frameworks add, in his data, it adds on top of a functioning transactional relationship, not as a substitute for a broken one.

Transformational leadership, measured through the MLQ, answers a broader question about what followers experience from a leader and whether that experience correlates with effort, engagement and performance; the correlational record across a large number of meta-analyses is real, and dismissing the construct outright would go beyond what Section 3's critique actually supports. What a manager should resist is treating an MLQ-style profile, their own or a colleague's, as a validated audit of specific behavioral change rather than what it mostly is: a structured follower impression, entangled with how well the unit has recently performed, built on subscales that do not clearly separate from each other or from a general halo of approval. Used for self-reflection, prompting a leader to ask where their own idealized influence shades into mere likability, the construct is genuinely useful. Used as a certified diagnosis in a promotion file, it is carrying more evidentiary weight than the measurement literature behind it can support.

Servant leadership asks a different question again, closer to Greenleaf's original framing than to either of Bass's two dimensions: not whether the leadership style raises performance, but whom the leadership exists to benefit, and what happens to the people served, and in Greenleaf's fuller version, to people outside the immediate team entirely. Its research base, through Liden et al. and van Dierendonck, is younger and still less settled on a single instrument than the three-decade MLQ literature, which should temper how confidently its correlational findings get used in a high-stakes decision. Greenleaf's best test nonetheless remains a usable question a manager can put to their own conduct directly, without waiting for a validated scale to administer it.

None of the three frameworks tells a manager what to do when the three pull in different directions at once, and that gap is worth naming rather than smoothing over. A follower's growth, Greenleaf's test, might be best served by a stretch assignment that the organization's current incentive structure, the transactional layer, does not reward, and that no amount of inspirational framing, the transformational layer, can make financially rational for the leader to assign. Bass's model explains why a leader needs a working transactional base before transformational behavior adds much; it does not explain what a leader should do when the transactional system itself is the obstacle to the kind of follower development Greenleaf's test asks about. That is not a gap this lesson can close, because it depends on the specific incentive structure, the specific follower and the specific organization in front of a specific manager, which is exactly the kind of context-dependent judgment no single leadership framework, however well validated, can substitute for.

Select a leader you can study through verifiable, citable material: a published profile, a documented organizational turnaround, an authorized biography, a well-documented case study, or, if drawing on your own professional observation, first-hand experience clearly labeled as such rather than presented as an external citation. Do not invent a case.

Write a report of 1,500 to 2,000 words that:

1. identifies at least two specific, observable behaviors by this leader and classifies each one using Bass's own definitions (contingent reward, management-by-exception active or passive, or one of the four I's), rather than a general impression of "good" or "inspiring" leadership;
2. for any behavior classified under the four I's, explicitly tests it against van Knippenberg and Sitkin's (2013) discriminant-validity critique: could the same evidence just as easily be redescribed as general charisma or likability rather than a distinct transformational process, and why or why not;
3. assesses whether a romance-of-leadership bias (Meindl, Ehrlich & Dukerich, 1985) might be shaping your own judgment of this leader: would you rate the same behaviors as positively if the organization's results had been mediocre rather than good, or vice versa;
4. evaluates the same leader, or a different leader if the evidence fits better, against Greenleaf's "best test" for servant leadership and against at least three of Liden et al.'s (2008) seven dimensions, with concrete, cited evidence for each dimension assessed, not an assertion;
5. states explicitly what evidence is missing that would make your analysis more conclusive, rather than resolving the gap with an unsupported assumption.

Cite at least four academic sources, including at least three from this article's reference list.

Assessment rubric (100 marks)

1. Burns, J. M. (1978). Leadership. Harper & Row.
2. Bass, B. M. (1985). Leadership and performance beyond expectations. Free Press.
3. Bass, B. M. (1990). From transactional to transformational leadership: Learning to share the vision. Organizational Dynamics, 18(4), 19–31. https://doi.org/10.1016/0090-2616(90)90061-S
4. Bass, B. M., & Avolio, B. J. (1995). MLQ, Multifactor Leadership Questionnaire: Technical report. Mind Garden.
5. van Knippenberg, D., & Sitkin, S. B. (2013). A critical assessment of charismatic-transformational leadership research: Back to the drawing board? Academy of Management Annals, 7(1), 1–60.
6. Meindl, J. R., Ehrlich, S. B., & Dukerich, J. M. (1985). The romance of leadership. Administrative Science Quarterly, 30(1), 78–102.
7. Greenleaf, R. K. (1970). The servant as leader. Greenleaf Center for Servant Leadership.
8. Greenleaf, R. K. (1977). Servant leadership: A journey into the nature of legitimate power and greatness. Paulist Press.
9. Liden, R. C., Wayne, S. J., Zhao, H., & Henderson, D. (2008). Servant leadership: Development of a multidimensional measure and multi-level assessment. The Leadership Quarterly, 19(2), 161–177. https://doi.org/10.1016/j.leaqua.2008.01.006
10. van Dierendonck, D. (2011). Servant leadership: A review and synthesis. Journal of Management, 37(4), 1228–1261.

Verification notes: all ten sources were checked via live web search against publisher or secondary citation records (title, journal or book, year, and volume/page/DOI details where stated) immediately before drafting; none were read in full. The division head's speech example in the opening section, the sales director scenario in the decision block, and the plant manager and regional manager examples elsewhere in the article are original, fictional teaching material constructed for this lesson and are explicitly labeled as such; none describes a real company, person, or dataset.

Next in The Human Capital MBA series: Lesson 28 examines emotional intelligence and its role in managerial decision-making.