Locus Of Control
agency
The degree to which a person attributes outcomes to their own action rather than to luck, fate or other people. Julian Rotter introduced it in 1966 as a generalised expectancy — a default assumption carried across situations, not a belief about any particular one.
In practice
Two people lose the same account. One reconstructs the three calls where it went wrong; the other concludes procurement was always going to pick the incumbent. Both may be right about this account. Only one of them can be given a different result next quarter.
The common mistake
Treating it as a fixed personality trait and a moral grade. It is a generalised expectancy learned from a history of outcomes, it is domain-specific in practice rather than global, and in genuinely constrained circumstances an external attribution is often the accurate one.
Ask someone why something turned out as it did and the answer tends to fall consistently on one side of a line: because of what I did, or because of what was done to me. Rotter (1966) named that consistency and built the first instrument to measure it.
Rotter's finding
Julian Rotter's monograph introduced internal versus external control of reinforcement as a generalised expectancyRotter, J. B. (1966). 'Generalized Expectancies for Internal Versus External Control of Reinforcement.' Psychological Monographs: General and Applied 80(1). The 29-item forced-choice I-E Scale, developed within social learning theory: expectancy of reinforcement generalises across situations that a person perceives as similar. within social learning theory. The crucial word is generalised. Rotter was not describing a belief about one situation but a default that transfers to new ones, formed from a history of occasions on which action did or did not change the outcome. That origin story matters: an external locus is a reasonable inference from an environment in which effort genuinely did not pay.
Martin Seligman and Steven Maier (1967) demonstrated the mechanism directly. Animals given inescapable shock stopped attempting escape even once escape became available — learned helplessness. The generalisation outlives the conditions that produced it, which is the entire practical problem: the belief was accurate when it formed and is wrong now, and nothing about the experience of holding it distinguishes the two cases.
Two corrections to Rotter
Hanna Levenson (1973) showed the scale conflated two different externals. Believing outcomes are controlled by powerful others is not the same as believing they are controlled by chance: the first implies a strategy — find out who decides — and the second implies none. A single internal-external dimension hides the distinction that most affects what someone should do next.
Albert Bandura (1977) raised the sharper objection. Locus of control asks whether outcomes respond to action in general; self-efficacy asks whether you can perform the specific actionBandura, A. (1977). 'Self-efficacy: Toward a Unifying Theory of Behavioral Change.' Psychological Review 84(2). Bandura distinguished outcome expectancy — that a behaviour produces a result — from efficacy expectancy — that one can execute the behaviour. Task-specific efficacy predicts behaviour substantially better than a global locus score.. Those come apart constantly: someone can be certain that deliberate practice produces skill and equally certain they personally will not sustain it. Bandura's task-specific measure predicts behaviour better than a global score, and where the two frameworks compete the evidence has generally favoured his.
The objections
Causality runs both ways, and the correlational literature cannot separate them. Internal locus is associated with better outcomes across health, academic and occupational measures, but a history of effort paying off produces an internal locus just as readily as an internal locus produces effort. Bernard Weiner (1985) refined the picture by splitting attributions along three dimensions rather than one — locus, stability and controllability — on the grounds that stable-uncontrollable attributions produce very different behaviour from unstable-controllable ones with the same locus.
The ethical objection is the serious one. Where constraints are real — capital, discrimination, illness, an employer who will not promote — an external attribution is accurate, and coaching someone toward an internal one is training them to misdescribe their situation and blame themselves for it. Rotter's own framing protects against this: the expectancy is learned from the environment, so the question of whether it currently fits the environment is always open, and sometimes it fits.
What it rules out
It rules out reading a single attribution as evidence of disposition — the construct is about the default across situations, not about being right on one occasion. It rules out treating an internal locus as a virtue, since it is an expectancy formed by circumstance. And it rules out a global score as a prediction about a specific task, which is Bandura's point.
It does not rule out external explanations. It rules out the case where the same external explanation arrives regardless of what changed — which is the diagnostic, and the only one the construct actually supports.
Sources
Bandura, A. (1977). 'Self-efficacy: Toward a Unifying Theory of Behavioral Change.' Psychological Review 84(2). · Levenson, H. (1973). 'Multidimensional Locus of Control in Psychiatric Patients.' Journal of Consulting and Clinical Psychology 41(3). · Rotter, J. B. (1966). 'Generalized Expectancies for Internal Versus External Control of Reinforcement.' Psychological Monographs 80(1). · Seligman, M. E. P. & Maier, S. F. (1967). 'Failure to Escape Traumatic Shock.' Journal of Experimental Psychology 74(1). · Weiner, B. (1985). 'An Attributional Theory of Achievement Motivation and Emotion.' Psychological Review 92(4).
Concept web
Open the full webQuestions
What is locus of control?
The degree to which someone attributes outcomes to their own action rather than to luck, fate or other people. Rotter (1966) defined it as a generalised expectancy — a default carried across situations, learned from a history of whether action changed the result.
Who developed locus of control?
Julian Rotter, in a 1966 monograph in Psychological Monographs, within social learning theory. His 29-item forced-choice I-E Scale was the first instrument to measure it.
What is the difference between locus of control and self-efficacy?
Locus of control asks whether outcomes respond to action at all; self-efficacy, in Bandura's sense (1977), asks whether you can perform the specific action. They come apart routinely, and the task-specific measure predicts behaviour better than a global score.
Is an internal locus of control always better?
No. Where constraints are genuinely external, an external attribution is accurate, and pressing someone toward an internal one trains them to misdescribe their situation. Causation also runs both ways: effort that pays off produces an internal locus as readily as the reverse.
Can locus of control change?
Yes — it is learned, not fixed. Seligman and Maier (1967) showed the expectancy outlives the conditions that formed it, which is the practical problem: it was accurate when it formed and may be wrong now, and holding it feels identical in both cases.