Maslow's pyramid, McGregor's X and Y, Schein's iceberg, Edmondson's psychological safety, Tushman's ambidextrous org, and the long argument about what makes people work well together.
Organizational behaviour is the study of how individuals, groups, and structures act inside organizations. It sits where psychology, sociology, anthropology, and management research overlap — and it tries to answer the obvious but stubborn questions: why do people work, what makes a group cohere, why do firms produce predictable patterns of pathology, and how do you design a place that doesn't.
The field as a research discipline is a child of the postwar American university. The Hawthorne studies (1924–32) provided the founding empirical paradox; Maslow (1943), McGregor (1960), and Herzberg (1959) provided the founding theoretical claims; the journals — Administrative Science Quarterly (1956), the Academy of Management Journal (1958) — gave it institutional form.
This deck covers the canonical theorists (Maslow, McGregor, Schein), the Hawthorne studies and their long shadow, group dynamics (Dunbar, Tuckman), the contemporary high-performance findings (Edmondson, Tushman/O'Reilly), and the structural shift remote work has imposed.
The founding empirical episode of OB. Between 1924 and 1932, researchers from the Western Electric Company and (after 1927) Harvard Business School ran a series of experiments at the Hawthorne Works plant in Cicero, Illinois. The questions were Taylorist: what level of factory illumination, rest breaks, and incentive pay maximised worker productivity?
The results were peculiar. Productivity rose when illumination was increased; it also rose when illumination was decreased. It rose when rest breaks were added, then continued to rise when they were removed. The control group's productivity also rose. Something other than the manipulated variables was driving the change.
The interpretation, due primarily to Elton Mayo and his collaborators (Fritz Roethlisberger, William Dickson, T.N. Whitehead), was published in Management and the Worker (1939). Their claim: workers responded to the attention of the researchers, to changes in their social grouping, and to the meaning the experimental context gave their work. The "Hawthorne effect" became shorthand for the contamination of any field experiment by the participants' awareness of being studied.
Subsequent re-analyses (Stephen Jones, 1990s; Steven Levitt, 2009) have argued the original Hawthorne data are messier than Mayo claimed and the "effect" was overstated. What survives intact is the deeper insight: workers are social creatures whose behaviour cannot be predicted from job design and incentives alone. The human-relations school of management was born here.
Abraham Maslow (1908–1970) published "A Theory of Human Motivation" in Psychological Review in 1943. The paper proposed a hierarchy of human needs — physiological, safety, love and belonging, esteem, self-actualisation — that organised motivation as a layered structure. Higher needs become motivating only as lower ones are reasonably satisfied.
Maslow himself never drew a pyramid. The triangular diagram now universally associated with him was added by management textbooks in the 1960s and 70s. Maslow used the hierarchy to make a humanistic-psychology argument against behaviourism and Freudianism — humans had positive growth needs, not just deficits to be remedied.
OB and management adopted Maslow enthusiastically because the hierarchy implied a managerial agenda: pay covered the lower levels, but to motivate skilled work, organisations had to offer belonging (team membership), esteem (recognition), and self-actualisation (challenging, meaningful work).
The empirical support for the strict hierarchy is weak. Cross-cultural studies (Hofstede; Wahba & Bridwell, 1976) found the level ordering didn't replicate. Tay and Diener's 2011 large-sample study found the needs are simultaneously rather than sequentially motivating — people in poverty also pursue belonging and esteem; well-fed people still need security.
What survives is the soft claim — there are multiple kinds of need and pay alone doesn't motivate. The hard pyramid is mostly a teaching device.
Douglas McGregor's The Human Side of Enterprise (1960) is the most-cited single book in management psychology. McGregor (MIT Sloan, 1937–1964; Antioch College president, 1948–1954) argued that managers operate from one of two implicit theories of human nature, and that the theory determines what they do.
McGregor was clear: Theory X is the view most American managers actually hold; Theory Y is the view he believed was empirically more accurate. The book was a sustained argument that managerial assumptions are self-fulfilling — Theory X management produces Theory X workers, and the system reproduces its own evidence.
The intellectual reach of the X/Y distinction is hard to overstate. Every subsequent participative-management theory — Likert's System 4, Argyris on double-loop learning, Senge's learning organisation, the Lean management tradition — sits in McGregor's Theory Y lineage.
The 1969 third Japanese theory — Theory Z — was added by William Ouchi (1981) for the Japanese consensus-management style. Less influential but historically resonant.
Frederick Herzberg's 1959 book The Motivation to Work (with Mausner and Snyderman) introduced the two-factor theory: the things that make people satisfied at work and the things that make them dissatisfied are different things, not opposite ends of a single scale.
Hygiene factors — pay, working conditions, company policy, supervisory style, security — when inadequate, produce dissatisfaction. When adequate, they don't produce satisfaction; they just remove dissatisfaction.
Motivators — achievement, recognition, the work itself, responsibility, advancement, growth — when present, produce satisfaction. When absent, they don't produce dissatisfaction; they just produce neutral indifference.
The implication: throwing money or perks at a demotivated workforce mostly raises the floor without raising the ceiling. Real motivation comes from the design of the work — what Herzberg called job enrichment.
Herzberg's 1968 HBR article "One More Time: How Do You Motivate Employees?" — still the highest-selling reprint in HBR history — distilled the argument. The article introduced the famous "KITA" (kick in the ass) typology: positive KITA (rewards), negative KITA (threats), and the (in Herzberg's view) trivial difference between them. Real motivation was a different mechanism entirely.
Empirically, the two-factor structure is contested. Vroom and others found that the same factor (e.g. recognition) can produce both satisfaction and dissatisfaction depending on direction. The clean two-factor split is methodologically suspect. The underlying insight — that work design matters more than perks — has held up.
Edgar Schein (1928–2023, MIT Sloan, Bell Labs trained) was the dominant theorist of organizational culture. Organizational Culture and Leadership (1985, with successive editions through 2017) is the canonical text.
Schein's three-level model:
The model's usefulness is that it explains why culture-change efforts focused on Level 1 (re-decorate the office, change the mission statement) fail to alter Level 3 behaviour. The deep assumptions persist; the surface signals shift slightly; the organization reverts.
Schein's deeper insight: culture is the solution a group has accumulated to its problems of external adaptation and internal integration. It is what has worked in the past. To change culture is to ask the group to abandon what has historically kept it alive — which it will resist for entirely sensible reasons.
His later work on humble inquiry (2013) and helping (2009) extended the framework into practical methods for diagnosis and change.
Robin Dunbar's 1992 paper "Neocortex size as a constraint on group size in primates" proposed that the size of stable social groups in a primate species is predicted by the size of its neocortex. The corresponding human number — calculated from human neocortex ratio — was approximately 150.
The number has been observed in: Neolithic village sizes (typically 100–200), Hutterite agricultural community split thresholds (averaging 110–180), military unit cohesion (the Roman maniple, the modern company-sized unit, ~150), Christmas-card list sizes, hunter-gatherer band size aggregations.
The OB application: organizations above ~150 people develop qualitatively different communication patterns. Below 150, you can know everyone personally; above 150, you cannot. The transition produces predictable problems — emergence of factions, decline of informal coordination, need for explicit hierarchy and process.
Companies that have built explicitly around the Dunbar number: Gore (W.L. Gore, the GORE-TEX maker, splits buildings at ~150 employees); Spotify's squad/tribe model with tribes capped near Dunbar size; Valve's flat structure (with reported coordination problems above the threshold); historically, monasteries and the British Army company.
The modern critique (Lindenfors et al., 2021; Bernard, 2021) is that Dunbar's calculation rests on shaky neocortex-ratio extrapolation and the empirical data don't support a sharp threshold. The number is probably better thought of as a soft heuristic than a constant. The organizational implication — that group size affects coordination cost — is robust regardless.
Bruce Tuckman's 1965 paper "Developmental Sequence in Small Groups" proposed a four-stage model of team formation that has since become near-universal in management training. (Tuckman added a fifth stage, adjourning, in 1977.)
Forming. Initial polite uncertainty. Members orient themselves to the group, the task, and each other. Productivity is low; behaviour is cautious.
Storming. Conflict emerges. Members push back on roles, leadership, and method. Many groups fail at this stage. Those that survive learn each other's styles and develop mechanisms for surfacing and resolving disagreement.
Norming. Agreement on roles and norms emerges. Behaviour stabilises; trust develops; the group develops its own working rhythm.
Performing. The group operates effectively. Members are interdependent and oriented to the task. Productivity is high. Most teams that last long enough reach this stage.
The model is a stage theory and like all stage theories has been criticised as too linear and too universal. Connie Gersick's 1988 punctuated-equilibrium research on project teams found groups didn't develop linearly through stages; they spent the first half of their time slowly establishing a working approach, then at the midpoint had a sharp transition into delivery mode.
The synthesis: Tuckman's stages happen, but not always in order, and not always cleanly. The categories remain useful as diagnostics: a "storming" team needs different intervention than a "norming" one.
The most influential OB construct of the last 25 years. Amy Edmondson (Harvard Business School) introduced the term in her 1999 Administrative Science Quarterly paper "Psychological Safety and Learning Behavior in Work Teams."
The definition: psychological safety is "a shared belief held by members of a team that the team is safe for interpersonal risk-taking." More specifically: a team is psychologically safe when its members feel they can raise problems, admit mistakes, ask questions, propose ideas, and disagree without humiliation, retaliation, or social penalty.
Edmondson's original empirical paradox: in her hospital-team study, the better teams reported more medication errors. The interpretation: better teams weren't making more errors — they were reporting them. Worse teams were burying errors out of fear of blame. Psychological safety predicted not error rate but error visibility, which over time predicted learning.
The construct hit the mainstream after Google's 2012-2015 Project Aristotle study of 180 internal teams found psychological safety to be the strongest single predictor of team effectiveness — outranking cohesion, dependability, structure, meaning, and impact. The 2015 New York Times Magazine article on Aristotle ("What Google Learned from Its Quest to Build the Perfect Team") brought the term to general business readership.
Edmondson's The Fearless Organization (2018) and Right Kind of Wrong (2023) extended the construct. The empirical literature on psychological safety has grown rapidly — by 2023 there were over 1,000 published studies, with meta-analyses (Frazier et al., 2017) confirming moderate-to-large effects on team learning, performance, creativity, and engagement.
Michael Tushman (Harvard) and Charles O'Reilly (Stanford) introduced the concept in their 1996 California Management Review article and 2004 HBR follow-up. The argument: long-lived firms must simultaneously exploit existing capabilities (efficiency, refinement, incremental improvement) and explore new ones (innovation, experimentation, disruption-adjacent).
The two activities require different cultures, structures, processes, and reward systems. Trying to do both within a single business unit produces the worst of both — exploration is starved by exploitation's metrics; exploitation is destabilised by exploration's permissiveness.
The Tushman-O'Reilly solution: structural ambidexterity. House the explore activity in a separately organised unit, with its own people, metrics, culture, and budget; protect it from the exploit unit's processes; reconnect at the senior-leadership level to share resources and absorb exploration's results.
Case studies: IBM's Emerging Business Opportunity (EBO) program; Ciba Vision's separated R&D for daily-disposable contact lenses; USA Today's separated digital unit; Intel's long-running separation of new-process and existing-fab operations.
The contrasting view is contextual ambidexterity (Birkinshaw and Gibson, 2004) — the same individuals making explore-exploit trade-offs based on context, supported by an organizational climate that rewards both. Empirical results are mixed; both forms appear to work under different circumstances.
Tushman's broader framework — punctuated equilibrium in organizational change, "winning through innovation" — sits in the tradition of Schumpeter's creative destruction translated into operational management.
James G. March (1928–2018) was OB's most subtle theorist. With Herbert Simon (his collaborator on Organizations, 1958) and Richard Cyert (A Behavioral Theory of the Firm, 1963), he founded the behavioural-organization theory that displaced rational-actor models.
The Simon-March-Cyert framework: organizations are not unitary rational actors. They are coalitions of agents with imperfect information, bounded rationality, conflicting goals, and operational routines that drift over time. Decisions are made by satisficing — accepting the first sufficiently-good option rather than searching for the optimum (Simon's Nobel-winning 1955 concept).
March's most famous later contribution: the "garbage can model" of organizational choice (Cohen, March, Olsen 1972). In ambiguous organizations — universities, government bureaus, project teams under uncertainty — choice opportunities, problems, solutions, and participants flow independently and combine accidentally. Solutions get attached to problems they were not designed for. Problems get attached to choice opportunities by historical drift. The result is decision-making that looks irrational from outside but is the ordinary product of the ambiguity.
March's 1991 paper "Exploration and Exploitation in Organizational Learning" — the conceptual ancestor of Tushman-O'Reilly ambidexterity — formalised the trade-off between refining what you know and discovering what you don't.
The thread connecting March's work is a sustained scepticism about the rationality myth — and a generous attention to how organizations actually work, with all their mess. He is the OB theorist most worth reading in the original.
Henry Mintzberg (McGill) wrote his PhD by following five chief executives around for a week each, recording every activity in five-minute increments. The 1973 book The Nature of Managerial Work demolished the planning-organising-controlling textbook account of management.
What managers actually do: deal with brief, fragmented, oral interactions (managers averaged 9 minutes between interruptions); spend their time on reactive rather than reflective work; rely heavily on informal information networks rather than formal reporting systems; engage in roles Mintzberg classified as interpersonal (figurehead, leader, liaison), informational (monitor, disseminator, spokesperson), and decisional (entrepreneur, disturbance handler, resource allocator, negotiator).
Mintzberg's 1990 HBR article "The Manager's Job: Folklore and Fact" is the compact version. His 2004 book Managers Not MBAs extended the argument into a critique of MBA programs that, in his view, train people to plan and analyse without ever managing.
His Structure in Fives (1979) and the 1989 expansion into the Mintzberg on Management typology — Simple Structure, Machine Bureaucracy, Professional Bureaucracy, Divisionalised Form, Adhocracy — remain the most-used framework for diagnosing organizational design.
Mintzberg's critique of strategy as a planning exercise (The Rise and Fall of Strategic Planning, 1994) places him as the field's most consistent voice for emergent, learning-based, practitioner-grounded management.
Chris Argyris (1923–2013, Yale, Harvard Business School) and Donald Schön introduced the distinction in Organizational Learning (1978).
Single-loop learning. Detect a problem and adjust to fix it within the existing operating frame. The thermostat metaphor: the room is too cold, so it turns on the heat. The mental model is unchanged.
Double-loop learning. Detect a problem and question the underlying assumptions, goals, or values that produced it. The thermostat asks whether 22°C is the right target temperature in the first place.
Argyris's claim: organizations and individuals overwhelmingly default to single-loop learning. Double-loop learning requires examining one's own assumptions, which is psychologically threatening. Most organizations have defensive routines — practices that protect members from embarrassment but prevent the deeper learning that would resolve persistent problems.
Argyris's Theories of Action framework distinguishes espoused theory (what people say they do) from theory-in-use (what they actually do). The gap is rarely visible to the actor. Argyris developed action-research methods to surface the gap and produce double-loop change.
His later work — Knowledge for Action (1993), Reasoning, Learning, and Action (1982) — provided the methodological apparatus. Senge's The Fifth Discipline (1990) on the learning organization sits squarely in the Argyris-Schön tradition and has been the most widely-read popularisation.
The motivational theory most empirically supported in OB. Edward Deci and Richard Ryan (Rochester) developed self-determination theory (SDT) over 50 years, beginning with Deci's 1971 PhD work on intrinsic motivation.
The core claim: humans have three universal psychological needs — autonomy (sense of self-direction), competence (sense of effectiveness), and relatedness (sense of connection to others). Workplaces that satisfy these needs produce intrinsic motivation, sustained engagement, and well-being. Workplaces that thwart them produce compliance at best and resentment at worst.
The most-cited finding: extrinsic rewards can undermine intrinsic motivation when they are perceived as controlling. Deci's classic experiment had subjects solve interesting puzzles either for pay or for free. The paid subjects solved fewer puzzles in the post-experiment free-choice period than the unpaid subjects — being paid had transformed the puzzles from intrinsically interesting into a job.
The OB implications: pay matters for fairness and as hygiene, but pay-for-performance schemes designed as control mechanisms can undermine the intrinsic motivation that produces the best work. This is consistent with Herzberg's earlier finding and inconsistent with the strong agency-theory account of incentives.
SDT's empirical literature is unusually deep — over 1,000 studies across cultures and contexts. The 2017 Deci-Ryan synthesis in Self-Determination Theory: Basic Psychological Needs in Motivation, Development, and Wellness is the comprehensive reference.
Daniel Pink's Drive (2009) is the popular-management version, foregrounding autonomy, mastery, and purpose.
The most empirically defensible personality framework in OB is the Five-Factor Model (the "Big Five"): Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism. The OCEAN model emerged from converging factor-analytic research in the 1980s (Goldberg; Costa & McCrae; John).
The OB-relevant findings:
Conscientiousness is the strongest single Big Five predictor of job performance across virtually all jobs. The meta-analytic correlation is roughly r = 0.20–0.25 — modest but robust. (Barrick & Mount, 1991, and many replications.)
Extraversion predicts performance specifically in jobs requiring social interaction (sales, management) but not generally.
Neuroticism negatively predicts performance and positively predicts turnover, burnout, and counterproductive work behaviour.
Agreeableness predicts teamwork performance and customer-service performance but is roughly neutral or slightly negative for negotiation outcomes.
Openness predicts creative performance and adaptability to change.
The most-used commercial personality instrument in HR practice — the Myers-Briggs Type Indicator (MBTI) — is psychometrically poor. Test-retest reliability is around 50% for the "type" categorisations (i.e. the same person retakes it and gets a different type half the time). The MBTI is not validated against job performance and is not used in serious selection contexts. Its persistence is a credentialing and brand-management story, not a scientific one.
The professional consensus: use the Big Five (NEO-PI-R, IPIP scales, the HEXACO extension) for selection decisions; treat MBTI and similar typologies as light entertainment.
John French and Bertram Raven's 1959 paper "The Bases of Social Power" identified five sources of power in organizations: coercive (threat of punishment), reward (control of incentives), legitimate (formal authority), referent (admiration; charismatic), and expert (specialised knowledge). Raven later added informational (control of facts).
The OB-relevant claim: the kinds of power available to a manager strongly shape what they can accomplish. A manager whose only power is coercive will get compliance but not commitment; one with referent and expert power produces voluntary effort. Most successful managers cultivate multiple power bases in parallel.
Jeffrey Pfeffer's Power (2010) and 7 Rules of Power (2022) push the analysis further: organizational success requires explicit attention to power-acquisition, which most management theory ignores or moralises against. Pfeffer's empirical claim — that the diligent acquisition of organizational power, including by techniques the textbook frowns on, is a substantial predictor of career outcomes — is uncomfortable but well-supported.
The Stanford-Pfeffer line on power has been sharpened by political-skill researchers (Gerald Ferris and colleagues): the four-factor model of political skill (social astuteness, interpersonal influence, networking ability, apparent sincerity) reliably predicts job performance, leadership emergence, and career advancement, especially in ambiguous organizations.
The recurring finding from the Pfeffer-Ferris literature: power and politics are not pathologies of organizations; they are organizations. Treating them as pathologies prevents engagement with how the system actually works.
Karen Jehn's 1995 paper distinguished two kinds of conflict in work groups: task conflict (disagreement about the work — methods, content, decisions) and relationship conflict (interpersonal hostility — friction about personalities, values, styles).
The empirical finding, replicated many times: task conflict is moderately positive for team performance up to a point; relationship conflict is consistently negative for team performance and member satisfaction. The effects are different in direction.
The complication: task conflict and relationship conflict tend to correlate strongly in practice (~r = 0.5–0.7 in meta-analyses). Disagreements about how to do the work spill over into resentment about the colleague who keeps disagreeing. The teams that get the benefit of task conflict without the cost of relationship conflict are doing something specific — usually high-trust norms, clear roles, and the leadership skill to redirect from "you" to "the work."
The Edmondson psychological-safety construct sits next to this finding: psychologically safe teams have task conflict that does not collapse into relationship conflict. The mechanism is the absorption of disagreement into the work rather than into the people.
Practical implications: encourage explicit disagreement on substance; insist on decoupling that disagreement from interpersonal evaluation; intervene early on relationship conflict rather than letting it develop into stable enmity. Teams that fight well over ideas tend to perform well; teams that fight personally tend not to.
The Job Demands-Resources model (JD-R) — Demerouti, Bakker, Nachreiner, Schaufeli, 2001 — is the most empirically supported framework for work stress and engagement. It's structurally simple: every job has demands (workload, time pressure, emotional labour, role conflict) and resources (autonomy, feedback, social support, opportunities for development).
The two pathways:
Strain pathway. High demands + low resources = exhaustion, depersonalisation, reduced efficacy. Christina Maslach's classic three-component burnout.
Motivational pathway. High resources (independent of demand level) = engagement, satisfaction, performance.
The JD-R framework absorbs both Karasek's earlier Job Demand-Control model and Hobfoll's Conservation of Resources theory. The empirical literature — over 5,000 studies as of 2024 — supports the basic structure across industries and cultures.
The 2010s and 2020s burnout discourse has popularised the construct, sometimes losing analytical precision (burnout has become near-synonymous with "feeling stressed"). The Maslach Burnout Inventory and the more recent Burnout Assessment Tool (Schaufeli et al., 2020) provide the validated measurement.
The OB intervention design: reduce excessive demands where possible; build resources reliably (autonomy, feedback, support, mastery, growth); attend to recovery (rest, detachment, sleep). Most workplaces under-intervene on the resources side because increasing autonomy and support feels less concrete than reducing workload.
Organizations are not hierarchies on paper; they are networks in fact. The OB sub-discipline of social network analysis studies the patterns of communication, influence, and collaboration that the org chart doesn't show.
The foundational findings:
Mark Granovetter's "The Strength of Weak Ties" (1973) — weak ties (acquaintances, distant colleagues) carry more novel information than strong ties (close friends, immediate teammates). Most useful job leads, ideas, and opportunities flow through weak ties because strong ties already share your information set.
Ronald Burt's structural holes theory (1992) — the most valuable network position is one that bridges otherwise-disconnected groups. The bridge controls information flow and combines knowledge from multiple worlds. Burt's empirical work showed that managers in structural-hole positions reliably had higher pay, better performance evaluations, and more promotions than equally-skilled colleagues without the network position.
Modern organizational network analysis (ONA) — using email metadata, Slack patterns, calendar overlap — has made network mapping operational. Companies like Microsoft (Workplace Analytics), Worklytics, and TrustSphere offer ONA platforms.
The applications: identifying hidden bottlenecks (a single person mediates information flow between two halves of the org), identifying isolation (new hires not absorbed into the network), measuring the impact of reorgs on collaboration patterns, and — controversially — surveillance.
The OB literature on diversity in teams is more nuanced than the popular discussion suggests. The findings:
Demographic diversity (race, gender, age) has small and inconsistent effects on team performance in the meta-analytic literature (van Dijk et al., 2012; Bowers et al., 2000). The effects depend heavily on task type, team tenure, and organizational context.
Cognitive / functional diversity (education, expertise, perspectives) has more consistent positive effects on creative and complex problem-solving tasks (van Knippenberg & Schippers, 2007).
Diversity faultlines — when multiple demographic divides align (e.g. all women are also younger and from one department) — are particularly damaging. They produce subgroup formation and predictable conflict (Lau & Murnighan, 1998).
The mediating mechanism: diversity helps when it contributes to the team's information elaboration (genuine engagement with multiple perspectives) and hurts when it triggers social categorisation (in-group/out-group dynamics). The same diversity composition can produce either outcome depending on team norms and leadership.
The Phillips-Liljenquist-Neale 2009 paper "Is the Pain Worth the Gain?" found that diverse teams reported less satisfaction with their process but produced better decisions on complex tasks — they got better answers from a more uncomfortable conversation. Most organizations under-tolerate the discomfort and lose the gains.
The intervention design that has held up empirically: structured task focus, explicit norms about contribution, hiring leaders who can manage subgroup faultlines, training on perspective-taking and information-elaboration behaviours.
Leadership is OB's most-studied and least-resolved topic. The major lineages:
Trait theory (1900s-40s, then resurrected with Big Five). The search for personality features that predict leadership emergence and effectiveness. Modest meta-analytic effects: extraversion, conscientiousness, openness, low neuroticism. The trait approach was abandoned mid-century, then partially rehabilitated.
Behavioural / style theories (Ohio State, University of Michigan, 1950s). Two major dimensions: consideration (relationship-orientation) and initiating structure (task-orientation). Both predict different aspects of effectiveness.
Contingency theories (Fiedler, 1967; House's Path-Goal, 1971). The right leadership style depends on the situation. Empirically supported but practically unwieldy.
Transformational vs transactional leadership (Burns 1978; Bass 1985). Transformational leaders engage followers' values, identity, and aspiration; transactional leaders manage exchanges. The transformational construct (the MLQ instrument) generated thousands of studies and has been criticised for circularity — high-performance teams are described as having transformational leaders, by definition.
Authentic, servant, and ethical leadership (2000s-10s). The post-Enron generation of theories. The empirical record is variable; the constructs overlap heavily and the discriminant validity is contested.
The honest summary: there is no single profile of an effective leader; effectiveness is contingent on task, team, and context. The traits and behaviours that matter at the margin — being decisive, credible, attentive, and trustworthy — are unsexy but consistent. The popular leadership-book genre vastly overpromises.
Why organizational change is hard, and the canonical models for doing it:
Kurt Lewin (1947): three-stage model. Unfreeze (destabilise the existing equilibrium), change (introduce the new), refreeze (institutionalise it). Almost every subsequent model is a Lewin variant.
John Kotter's 8-step model (1996, Leading Change): create urgency, form a coalition, create vision, communicate it, empower action, generate short-term wins, sustain acceleration, institute change. Practical, prescriptive, widely applied. Kotter's later work (Accelerate, 2014) added the dual-operating-system idea — running a hierarchical execution organization in parallel with a network-based change organization.
The empirical record on change initiatives is sobering. The widely-cited "70% of change programs fail" figure (Beer & Nohria 2000; McKinsey research) is a heuristic; the actual evidence base is thinner than the citation suggests, but the basic claim — that most planned organizational change underdelivers — is well-supported.
The reasons change fails: leaders underestimate organizational inertia and the strength of Schein's Level 3 assumptions; the change targets only Level 1 artefacts; the urgency case is not made compellingly; the existing reward and measurement systems are not aligned with the new behaviours; middle managers are bypassed and quietly resist; the change exhaustion of the workforce (the average Fortune 500 employee experiences ~4 major change programs per year) creates cumulative cynicism.
The Beer-Nohria distinction between Theory E change (economic-value-driven, top-down, restructuring-focused) and Theory O change (organizational-capability-driven, participative, learning-focused) clarifies the trade-offs. Most successful long-term change programs combine both.
The 2020-22 pandemic produced the largest natural experiment in workplace structure in a century. Pre-2020, ~5% of US paid workdays were remote. April 2020 ~60%. 2024 settled at ~30% with hybrid the dominant pattern in white-collar work (Bloom et al., Stanford WFH research).
The OB findings emerging from the natural experiment:
Productivity. Mostly stable or slightly higher for routine knowledge work; mixed for collaborative-intensive work. Microsoft's 2022 internal study (Yang, Holtz, et al., Nature Human Behaviour) found the firm's collaboration networks became more siloed during full-remote — fewer cross-team weak ties, more time within tight clusters.
Onboarding. Substantially harder remote. New hires take longer to absorb tacit knowledge, build the weak-tie network, and develop trust. The firms that had high turnover plus high remote during 2021-22 paid the largest cost.
Career mobility. Some evidence that remote workers face slower promotion velocity than in-office colleagues at the same firm — the "proximity bias" finding.
Engagement and burnout. Mixed and population-specific. Working parents and caregivers reported higher overall well-being remote; younger workers and those with smaller homes reported lower.
Equity. The remote/hybrid divide has become a class divide. Knowledge workers can work remotely; service workers cannot. Within knowledge work, women, parents, and disability-affected workers disproportionately prefer remote; the proximity-bias risk falls on these groups.
The OB design problem of the 2020s is hybrid coordination — explicitly redesigning meetings, communication norms, performance measurement, and career systems for a workforce in multiple modes simultaneously.
Kurt Lewin (1890-1947, German-Jewish émigré, founder of MIT's Research Center for Group Dynamics) is the unsung architect of modern OB. Most of the field's central techniques — action research, T-groups, force-field analysis, the unfreeze-change-refreeze model of change — descend from Lewin.
His most-cited dictum: "There is nothing so practical as a good theory." His most consequential conceptual move: the field theory formulation that behaviour B is a function of the person P and the environment E, B = f(P, E) — not P alone and not E alone. The same person behaves differently in different fields; the same field elicits different behaviour from different people. The interaction is the unit of analysis.
Lewin's action research programme — researching organisations by intervening in them and observing the response, with the explicit goal of producing both knowledge and change — is the methodological lineage of every subsequent organisational-development practitioner. The framework is now standard in management consulting, OD interventions, and applied social science.
Lewin's National Training Laboratories (NTL, founded 1947, posthumous) developed the T-group (training group) — small-group experiential learning where participants observe and discuss their own group dynamics in real time. The T-group spawned encounter groups, group therapy traditions, and contemporary leadership-development practices.
His force-field analysis — diagnosing organisational situations as the equilibrium of driving and restraining forces — remains a basic OD intervention. The implication: change is more often achieved by reducing restraining forces than by adding driving forces. Adding more pressure to change a stuck system typically increases resistance; identifying and removing the specific restraints releases change.
Lewin died of a heart attack at 56, having transformed organisational psychology and most adjacent fields in fewer than fifteen American years. He is the field's least-celebrated foundational figure.
Erving Goffman (1922-1982, Penn) wrote about everyday social interaction with more precision than anyone before or since. The Presentation of Self in Everyday Life (1959) and Interaction Ritual (1967) are particularly relevant to organisational behaviour.
The dramaturgical framework: social interaction is a kind of theatrical performance. Each participant has a "front" — the controlled self-presentation appropriate to the role and audience — and a "back" — the off-stage self that the front conceals. Most of social life is the careful management of front-stage performance.
Applied to organisations: the meeting, the performance review, the difficult conversation, the offsite, the team retrospective — all are dramaturgical situations with stages, roles, scripts, and audiences. The Goffmanian analysis explains many phenomena that more abstract OB theories have trouble accommodating: why public meetings produce different behaviour than 1:1s; why people lie in employee-engagement surveys even when they're "anonymous"; why the change-management programme that works in one division stalls in another (different dramaturgical conventions).
Goffman's Frames Analysis (1974) introduced the concept of frames — the implicit rules participants use to make sense of the situation they are in. A workplace conversation can be framed as a problem-solving session, a status competition, a friendly catch-up, or a disciplinary episode. The frame is rarely explicit, but the participants' behaviour aligns to it. Misalignment of frames is one of the most common sources of workplace miscommunication.
The Goffmanian tradition is alive in contemporary research on emotion management at work (Arlie Hochschild's The Managed Heart, 1983, on flight attendants and the commodification of emotion), on team meetings as performance (Catherine Tinsley, others), and on the distinction between formal and informal organisational practice. The framework gives a vocabulary for what skilled managers know intuitively but rarely articulate.
An orientation to the field — what OB is, the canonical theorists, and the major contemporary findings. A useful entry before the deeper reading.
· Herbert Simon — Bounded Rationality & Satisficing — Simon's central insight that organizations are populated by agents with limited cognition, not unitary rational actors. The conceptual ground beneath all of behavioural-organization theory.
· James G. March — Intellectual Roots of the Garbage Can Model — March in his own words, on why organizational decisions look so much messier than the textbook descriptions. The most subtle voice in OB explaining the field's most subtle theory.
Read: McGregor's The Human Side of Enterprise for the foundation; Schein's Organizational Culture and Leadership for the deepest cultural model; Edmondson's The Fearless Organization for the most consequential modern construct; March and Simon's Organizations for the original behavioural framework; Pfeffer's Power for the unromanticised view of how organizations actually work.
If you read only seven things from the OB literature, read these:
1. McGregor on Theory X and Theory Y. Your beliefs about people produce the behaviour you predict.
2. Schein on culture's three levels. Surface culture is not the culture. The basic assumptions are.
3. March on bounded rationality and the garbage can. Organizations are not rational; they are coalitions of imperfectly-coordinated agents under ambiguity. Most management folklore that assumes otherwise is wrong.
4. Edmondson on psychological safety. Teams that report errors learn faster than teams that hide them. The mechanism is interpersonal; the effect is operational.
5. Argyris on single-loop and double-loop learning. Most of what looks like learning is just course correction within an unchanged frame. Real learning questions the frame.
6. Tushman and O'Reilly on ambidexterity. Long-lived firms must explore and exploit simultaneously. The two activities require structural separation; the leader integrates.
7. Mintzberg on what managers do. Brief, fragmented, oral, reactive. The textbook account of management is not a description; it is a wish.
Each of these has a substantial empirical literature behind it, has held up across replications, and is not where most management practice actually lives. The implementation gap is the field's permanent embarrassment.
Volume VII, Deck 16. The field that tries to figure out why organizations do what they do — and that has accumulated, slowly, an uncomfortable amount of useful knowledge that most organizations don't apply.
From the Hawthorne switchboards in 1928 to Edmondson's psychologically-safe teams in 2018, the through-line is that work is social, motivation is multiple, and assumptions about people produce people who fit the assumptions.
Set in Source Serif Pro and Source Sans Pro. Drafted in May 2026.