Goal Setting and Task Motivation (Locke & Latham, 2002)

A 35-year retrospective by Edwin A. Locke and Gary P. Latham on their goal-setting theory, published in American Psychologist (September 2002, Vol. 57, No. 9, pp. 705–717).

Core findings

Specific, difficult goals consistently produce higher performance than vague “do your best” instructions. The effect is robust: validated on over 100 tasks, with more than 40,000 participants in at least eight countries, across laboratory, simulation, and field settings. Effect sizes range from d = 0.42 on complex tasks to d = 0.80 on simple ones.

Four mechanisms explain how goals regulate action: they direct attention toward goal-relevant activities, energize effort, increase persistence, and arouse discovery of task-relevant knowledge and strategies.

Moderators

Goal effects depend on three key moderators:

Goal commitment — enhanced by importance (public commitment, monetary incentives) and self-efficacy. Participation in goal-setting has cognitive rather than motivational benefits: it stimulates information exchange and strategy formulation, but has negligible direct effect on performance when goal difficulty is controlled.
Feedback — goals plus feedback outperform goals alone. Feedback without goals is also insufficient: people need both a reference standard and information about current standing. This echoes the observability requirement from control theory.
Task complexity — on complex tasks, learning goals outperform performance goals. Effect sizes shrink as complexity rises (d = 0.48 for most complex vs. 0.67 for least complex). Proximal subgoals help decompose complex tasks and improve error management.

Critique of control theory

The paper explicitly rejects pure feedback control as the basis for motivation. Locke & Latham argue that Carver and Scheier’s (1981) control theory — a negative feedback loop modeled on the thermostat — treats discrepancy reduction as the source of motivation, implying that the natural state of the organism is rest. Instead, they adopt Bandura’s (1989) position: goal-setting is primarily a discrepancy-creating process. People who attain a goal set a higher one, producing new gaps to master. The dual cycle is “disequilibratory discrepancy production followed by equilibratory reduction.”

This creates an important nuance for the knowledge base’s framing of management as feedback control (see xettel, xettel). The user’s xettel cards characterize Locke & Latham as borrowing feedback control from cybernetics. The paper tells a more complex story: they use the language of goals and discrepancy, but argue that pure negative feedback is insufficient — feed-forward (setting new, higher goals) is the primary driver.

High-performance cycle

The integrated model: specific difficult goals → high performance → rewards (recognition, promotion) → high satisfaction → higher self-efficacy → setting even higher goals. This cycle explains why satisfaction does not directly cause performance (a long-standing puzzle in organizational psychology): satisfaction is the result of high performance when rewards match effort. It affects future performance only indirectly, through organizational commitment and subsequent goal-setting.

Practical applications

The paper reports dramatic productivity gains: truck drivers increased loads from 60% to 90% of legal weight, saving $250,000 in 9 months. Self-regulation training (setting attendance goals, monitoring obstacles, self-administering rewards/punishments) significantly increased job attendance and self-efficacy, with effects sustained over 9 months.