Job Satisfaction: S-Tier Behavioral Designer’s Guide

Every workplace-engagement metric ever invented sits on top of a question Edwin Locke answered in 1976: how does a worker decide whether their job is good? Most companies measure job satisfaction as if it were a single number on a survey. Locke proved it was an equation, and the equation has terms most engagement programs never touch.

The equation is short. Satisfaction equals what you want minus what you have, multiplied by how much that dimension matters to you. The variables are personal, the importance weights are private, and the same job can produce a thriving worker and a quietly devastated one depending on which two people you put in the chair. Engagement programs that ship the same intervention to everyone are solving for a constant when the math has three variables.

This is the S-Tier Behavioral Designer’s Guide to Job Satisfaction. We will walk through Locke’s Range-of-Affect Theory, the three instruments that operationalized it (the Job Descriptive Index, the Minnesota Satisfaction Questionnaire, and the Job Satisfaction Survey), the construct-validity critiques that survived the field’s first 50 years, the neuroscience of how a brain actually computes “how do I feel about this job,” and the Octalysis crosswalk that names which Core Drive each facet runs on. By the end you will have a six-step audit that turns Locke’s equation into a redesign tool that holds up against the Great Resignation, Quiet Quitting, and the algorithmic-management decade we are still living inside.

Speed Run Notes

• Locke 1976: Job Satisfaction = (Want Value − Have Value) × Importance. The importance weight is the term most engagement programs ignore, and it is why the same intervention produces opposite results across workers.
• Three canonical instruments operationalize the construct: Job Descriptive Index (JDI, Smith 1969, five facets), Minnesota Satisfaction Questionnaire (MSQ, Weiss 1967, 20 facets), and Job Satisfaction Survey (JSS, Spector 1985, nine facets). Use facet, not global.
• Job satisfaction predicts job performance at about r = 0.30 (Judge 2001 meta), turnover at about r = −0.36 (Tett & Meyer 1993), and life satisfaction at about r = 0.44 (Judge & Watanabe 1993). The performance correlation is smaller than practitioners think.
• The construct survived 50 years because it is descriptive, not prescriptive. Locke described what a satisfied worker looks like; he did not specify which structural levers move which facet. The Octalysis crosswalk fills that gap.
• Quiet Quitting is what the Range-of-Affect equation predicts when Importance weights collapse to zero on Work Itself and Promotion while staying high on Pay. The worker stays, the work decays.
• Job satisfaction is the static-state engineering view; Self-Determination Theory and Employee Engagement are the dynamic-process views. Use Locke when you need a diagnostic snapshot, use SDT when you need a redesign theory.

Author Credibility: Yu-kai Chou

Yu-kai Chou created the Octalysis Framework after studying gamification since 2003 — years before the term entered mainstream vocabulary. As a Human-Systems Architect & Behavioral Designer, his framework has been applied by LEGO, Microsoft, Porsche, Coca-Cola, Salesforce, and MrBeast, impacting over 1.5 Billion Users.

Chou has taught the Octalysis methodology at Harvard, Stanford, Yale, Tesla, Google, BCG, and IDEO.

His work has been cited by Harvard, Stanford, MIT, Forbes, Wall Street Journal, Wired, US Department of Energy, NIST, NSF, NCBI, US Department of Education, ClinicalTrials.gov, and Google Scholar — with 3,700+ more academic publications. Explore his books here.

TEDxLausanne · Gamification to Improve Our World17 min

What Is Job Satisfaction?

Job satisfaction is the pleasurable or positive emotional state resulting from the appraisal of one’s job or job experiences. That definition is Locke’s, written in his 1976 chapter “The Nature and Causes of Job Satisfaction” in the Handbook of Industrial and Organizational Psychology. The wording sounds simple. Each word is doing work.

“Emotional state” matters because it places satisfaction in the affect family, not the cognition family. A worker can rationally agree their job is well-designed and still feel terrible about it. The reverse is also true: a worker can fail to articulate why they love a job and feel its rightness anyway. Programs that ask “are you satisfied?” on a 1-5 scale collect a cognitive proxy for an underlying affective signal, and the signal noise shows up in every meta-analysis that has tried to predict behavior from satisfaction scores.

“Appraisal” is the second load-bearing word. Satisfaction is not the job itself. It is the worker’s evaluation of the job, filtered through what the worker wanted, what alternatives they imagined, and how much each dimension matters to them personally. Two workers with identical roles can produce opposite satisfaction scores, and neither is wrong. They are appraising different equations.

“Of one’s job or job experiences” expands the construct from the role on paper to the lived experience of doing it. The role description is a poor predictor of satisfaction. The week-to-week reality of meetings, autonomy, recognition, coworker dynamics, and small daily frictions is the actual input. This is why job redesign that changes the role on paper without changing the experience usually fails to move satisfaction scores.

Locke’s definition was a corrective. The dominant theory before him was Herzberg’s Two-Factor Theory (1959, 1968), which claimed hygiene factors (pay, supervision, working conditions) could only cause dissatisfaction while motivators (achievement, recognition, work itself) could only cause satisfaction. The data did not support the asymmetry. Locke showed that any facet could produce satisfaction or dissatisfaction depending on the gap between Want and Have, weighted by Importance. The Two-Factor model collapsed back into a one-equation model with three terms.

The Range-of-Affect Theory

The Range-of-Affect Theory is Locke’s formal model of how each facet of a job contributes to overall satisfaction. The full equation:

Satisfaction_facet = (Want_facet − Have_facet) × Importance_facet

Overall job satisfaction is the sum (or sometimes the average) of facet satisfactions across the set of facets that matter to a worker. The multiplicative form on the right side is what gave the theory its name: the range of affect a facet can produce is bounded by how much that facet matters. Trivially-weighted facets can move the satisfaction needle only trivially, no matter how large the Want-Have gap. Highly-weighted facets dominate the score.

Want Value

Want value is what the worker expects, hopes for, or feels entitled to on a given facet. It is not the population average. It is not what HR thinks the worker should want. It is what this worker has set as their personal reference point, formed from their history, their peer comparisons, their public-discourse exposure, and their private aspirations.

This is the term workplace surveys are worst at measuring. Survey designers default to asking workers to rate “how satisfied are you with [facet]” on a Likert scale, which collapses Want, Have, and Importance into a single response with no way to separate them. The result is uninterpretable. A worker who rates Pay as 3/5 might be (a) earning what they want on a low-Importance facet, (b) earning less than they want on a high-Importance facet, or (c) earning more than they want and feeling guilty about it (Adams’s Equity Theory overpayment case). The same survey number maps to three different design problems.

The Want term also drifts. Cullen and Perez-Truglia 2022 documented that workers updated their Want value for Pay within weeks of learning peer salaries during a pay-transparency disclosure. The Want value is not stable across an information shock. Job satisfaction surveys taken before and after such a shock produce inconsistent results because the Want term moved underneath the measurement.

Have Value

Have value is what the job actually delivers on the facet, as perceived by the worker. This is closer to objective measurement (Pay can be looked up, Promotion timing can be tracked, Coworker count can be counted) but the relevant input is still the worker’s perception, not the ground truth.

The gap between objective Have and perceived Have is one of the dirtiest sources of satisfaction-survey noise. A manager can give a worker exactly what the policy promised and have the worker score Have as below what was delivered, because the worker’s mental accounting included a milestone the manager never agreed to. The reverse is equally common: a manager who fails to deliver a promised raise can still get an above-objective Have rating if the worker mentally discounts the unfulfilled promise.

Communication mechanics matter here. The same delivered Have produces different perceived Have depending on whether it was framed as a gain (above prior baseline) or a loss (below expectation). This is the Kahneman-Tversky 1979 reference-dependence effect, imported intact into workplace satisfaction. Job redesign that changes objective Have without changing the framing routinely fails to move satisfaction scores.

Importance Weighting

Importance is the term that distinguishes Locke’s theory from the additive-facet models that preceded it. Importance is the worker’s private weighting of how much each facet matters to their overall sense of “is this job good for me right now.”

Importance weights are heterogeneous across workers, and heterogeneous over time for the same worker. A junior worker with student debt typically weights Pay heavily. The same worker, ten years later with a paid-off mortgage and two young children, may have re-weighted toward Flexibility, Coworker Trust, and Work Itself, with Pay dropping to a midweight even if its objective amount has tripled. The Range-of-Affect equation predicts that the same Pay raise produces different satisfaction effects in those two life stages.

This is why the engagement-survey question “how satisfied are you with Pay?” is informationally weaker than the two-question pair “how important is Pay to you right now?” and “how does your Pay compare to what you want?” The second pair lets you compute Importance × (Want − Have) for each facet, identify the high-Importance and high-Gap facet (the bottleneck), and target the intervention. The first question gives you a number that confounds the three terms.

Why the Equation Matters for Design

The Range-of-Affect equation reframes the design question. Bad designers ask “how do we make workers happier?” Better designers ask “which facet has the largest (Want − Have) × Importance product for which segment of workers, and which structural surface activates the Core Drive that closes that gap?” The first question is unanswerable. The second is a decomposed engineering problem with five tractable variables.

It also explains why one-size-fits-all engagement programs underperform their pilot numbers when scaled. The pilot worked because the population happened to share an Importance distribution. The rollout fails because a different population has different weights, and the same intervention now targets the wrong facet for half of them.

The Facet Models: How We Measure It

Locke’s equation is a theory. To deploy it inside a company you need an instrument. Three canonical instruments have dominated 50 years of empirical work: the Job Descriptive Index, the Minnesota Satisfaction Questionnaire, and the Job Satisfaction Survey.

Job Descriptive Index (JDI, Smith Kendall Hulin 1969)

The JDI is the most-cited job satisfaction instrument in the literature, published in 1969 by Patricia Smith, Lorne Kendall, and Charles Hulin at Cornell. It measures five facets: Work Itself, Pay, Opportunities for Promotion, Supervision, and Coworkers. Each facet has 9-18 items. Workers respond yes / no / cannot decide to a list of adjectives describing the facet (for Pay: “income adequate for normal expenses,” “well paid,” “barely live on income,” etc.). The yes/no format reduces acquiescence bias relative to Likert scales.

The JDI is psychometrically the most defensible facet instrument. The five-facet structure replicates across populations, the internal consistency reliabilities are above .80 for all five facets in the 2009 revision, and the convergent validity with global single-item measures is around .60. The cost is that the JDI takes 8-15 minutes to complete and is licensed (Bowling Green State University holds the copyright). That license cost is why many companies use the cheaper alternatives.

Minnesota Satisfaction Questionnaire (MSQ, Weiss 1967)

The MSQ comes from the Minnesota Theory of Work Adjustment (Dawis and Lofquist) and was published in 1967 by Weiss, Dawis, England, and Lofquist. The long form has 100 items across 20 facets (Ability Utilization, Achievement, Activity, Advancement, Authority, Company Policies, Compensation, Coworkers, Creativity, Independence, Moral Values, Recognition, Responsibility, Security, Social Service, Social Status, Supervision Human Relations, Supervision Technical, Variety, Working Conditions). The short form has 20 items, one per facet.

The MSQ’s 20-facet decomposition is more granular than the JDI’s five and useful when the design team wants to localize the satisfaction gap to a specific behavioral dimension (e.g., Recognition separately from Supervision). The cost is that 20 facets multiply the survey load and the response variance inflates noise relative to the JDI’s wider facets.

Job Satisfaction Survey (JSS, Spector 1985)

The JSS is Paul Spector’s 1985 instrument, designed for human-service, public-sector, and nonprofit organizations where the JDI’s private-sector framing did not always fit. It measures nine facets (Pay, Promotion, Supervision, Fringe Benefits, Contingent Rewards, Operating Conditions, Coworkers, Nature of Work, Communication) across 36 items rated on a 6-point Likert scale. The JSS is free for non-commercial use, which is why academic studies often default to it, and it has been translated into 30+ languages.

The JSS handles the Fringe Benefits and Contingent Rewards facets separately, which is useful in sectors where total compensation is heavily non-cash (healthcare, government, education). The Communication facet is also distinct from Supervision, which catches a satisfaction driver the JDI bundles into Supervision.

Global vs Facet: Which to Use

Global single-item satisfaction measures (“All things considered, how satisfied are you with your job?”) are not as bad as facet purists once claimed. Wanous, Reichers, and Hudy 1997 meta-analyzed single-item satisfaction measures and found a minimum reliability of about .67, sufficient for many practical purposes. Single-item global measures are also predictively comparable to facet-summed measures for outcomes like turnover intention.

The case for facet measurement is not about predictive power. It is about diagnostic power. A global score tells you the worker is dissatisfied; it does not tell you which facet is the bottleneck. The Range-of-Affect equation’s whole design value comes from facet decomposition. If you are running a quarterly health check, a global score is acceptable. If you are designing an intervention, you need facets, and you need Importance weights separately from the Want-Have gap.

What Locke Got Right

Three of Locke’s core claims have held up across 50 years of follow-on research, including waves of construct-validity critique. They are worth naming directly because each one is still misunderstood by practitioners reaching for satisfaction surveys.

The first is the affective primacy claim. Locke insisted satisfaction was an emotional state, not a cognitive judgment. The neuroscience that arrived in the 2000s confirmed this directly. Functional imaging shows that satisfaction-evaluation tasks activate affect-processing regions (ventromedial prefrontal cortex, amygdala) before they activate the higher-cognitive regions (dorsolateral prefrontal cortex) that workers reach for when asked to justify their satisfaction rating. The number on a survey is a post-hoc rationalization of an already-formed affective signal.

The second is the Importance weighting claim. Locke’s multiplicative term has been validated repeatedly. Mikes and Hulin 1968, McFarlin and Rice 1992, and the more recent Diener and Tay 2015 cross-cultural work all show that facet satisfactions weighted by Importance predict overall satisfaction better than unweighted sums. The effect is modest in absolute terms (the R-squared bump from adding Importance weighting is typically .03-.05), but the diagnostic value of knowing which facets matter most to which workers is much higher than the predictive bump suggests.

The third is the discrepancy claim itself: satisfaction is driven by the gap between Want and Have, not by the absolute level of Have. This is why high-Pay companies do not always have high-Pay satisfaction (the Want term scales with the perceived peer reference), and why low-Pay companies sometimes have high-Pay satisfaction (the Want term is calibrated to a peer group with lower Pay). Reference-dependence is not a satisfaction edge case; it is the satisfaction mechanism.

These three claims compound. Satisfaction is affective, weighted by Importance, and driven by the Want-Have gap. The implication is that workplace satisfaction is a private equation that companies can observe imperfectly from outside, and any program that treats it as a public, additive, level-of-Have problem will miss the mechanism the worker is actually computing.

Where Job Satisfaction Theory Falls Apart

Locke’s theory is descriptive. Its failures show up where the field tried to use it prescriptively without filling in the missing structural layer.

The Performance Correlation Is Smaller Than Practitioners Think

The popular claim that “satisfied workers are productive workers” predates Locke and predates the data. Judge, Thoresen, Bono, and Patton 2001 published the canonical meta-analysis on the satisfaction-performance relationship and found r = 0.30 across 312 samples. That is a real correlation, but it is not the half-correlation many engagement decks imply. The shared variance is about 9 percent, which means satisfaction and performance share roughly one-tenth of their variance and have ninety percent that is not shared.

Practical implication: investing in satisfaction primarily to drive performance is a weak bet. Investing in satisfaction because retention, life satisfaction, customer treatment, and reduced absenteeism are all moderately correlated with it is a stronger bet. The case for caring about satisfaction is moral and operational, not productivity-narrow.

The Theory Does Not Specify Mechanism

Locke described what a satisfied worker looks like; he did not specify which structural levers move which facet. A company that measures satisfaction with the JDI, identifies Pay as the lowest-scoring facet, and then increases Pay is making an inferential leap the theory does not authorize. The Want term could have moved (peer disclosure shifted the reference point), the Importance weight could have spiked (a recent life event made Pay temporarily heavier), or the Have term could be misperceived (the worker is not seeing the equity grant correctly). Each of those produces the same satisfaction score and requires a different intervention.

This is why the Octalysis crosswalk that follows in the Apply section is the load-bearing addition to Locke’s framework. Locke gives you the equation; Octalysis gives you the structural levers that move each facet’s Want, Have, and Importance terms.

Common-Method Variance Inflates Most Job-Satisfaction Findings

Most published satisfaction research uses self-report for both the predictor (satisfaction) and the outcome (performance, turnover intention, citizenship behavior). Common-method variance inflates correlations between two self-reports collected from the same person at the same time. Podsakoff, MacKenzie, Lee, and Podsakoff 2003 estimated common-method variance can account for as much as 35% of the observed correlation between two self-report measures collected together. The implication is that even the Judge 2001 r = 0.30 is probably a slight overstatement of the true satisfaction-performance link.

For practitioner decisions, this means satisfaction surveys cross-checked against objective behavior (attendance records, output metrics, churn data) carry more decision weight than satisfaction surveys cross-checked against self-reported engagement. The instrumentation choice is not neutral.

Cross-Cultural Generalizability Is Weaker Than the Theory Implies

The Range-of-Affect equation assumes Want, Have, and Importance are commensurable across workers within a cultural frame. Hofstede’s dimensions and the GLOBE study findings show that the Importance weighting structure shifts across cultures: collectivist-culture workers weight Coworker and Supervision facets more heavily than individualist-culture workers, who weight Work Itself and Promotion more heavily (Robert, Probst, Martocchio, Drasgow, and Lawler 2000 JAP). The satisfaction equation still applies inside each frame, but a multinational using one survey instrument across all subsidiaries is computing different things in each location, even if the items are translated correctly.

What’s Really Happening Inside the Brain

The neuroscience of job satisfaction was not available to Locke in 1976. What we now know vindicates the affective-primacy claim and explains why satisfaction surveys are noisier than most managers assume.

Satisfaction appraisals start in the ventromedial prefrontal cortex (vmPFC), the brain region that integrates value signals across domains. Hare, Camerer, and Rangel 2009 showed that vmPFC activity tracks the overall valuation of a decision option across different attribute types, and this is what a worker is doing when asked “how satisfied are you with your job?” The brain is computing a weighted sum of facet-level affective signals and producing an integrated value estimate. This is precisely the operation Locke’s equation models, performed by neural hardware that predates the equation by 200,000 years of human evolution.

The Importance weighting also has a neural correlate. Lim, O’Doherty, and Rangel 2011 showed that attention modulates the vmPFC value signal: features the worker attends to more get higher weighting in the integrated valuation. Importance, in Locke’s terms, is attention in neuroeconomic terms. A worker who is currently attending to Coworker dynamics (because the team just changed) will weight Coworker satisfaction more heavily in their overall job satisfaction appraisal than a worker whose team has been stable for two years.

The reference-dependence of the Have value maps to dopamine reward-prediction-error signals in the ventral striatum. Schultz 2016 reviewed the canonical literature: midbrain dopamine neurons fire when reality exceeds expectation, fall silent when reality matches expectation, and pause-then-resume in a depression-like pattern when reality undershoots expectation. The Have term in Locke’s equation is not the absolute Have. It is the Have minus the expected Have, which is exactly the reward-prediction error the dopamine system encodes. This is why a worker who gets exactly the bonus they expected feels nothing, while a worker who gets the same bonus unexpectedly feels great, and a worker who gets the same bonus after being told to expect more feels disappointed despite the identical Have.

Finally, the chronic-dissatisfaction state has a neural cost. Workers in sustained high (Want − Have) × Importance dissatisfaction show elevated cortisol on diurnal sampling (McEwen 1998 allostatic load), reduced functional connectivity between the prefrontal cortex and the amygdala (Liston, McEwen, Casey 2009), and shorter telomere length over multi-year follow-up (Epel et al 2004). Job satisfaction is not a soft variable in the long run. It is a biological exposure.

Job Satisfaction vs Other Theories

Job Satisfaction is one of several constructs the engagement-research field uses to describe how workers feel about their work. Each of the neighboring constructs captures something Locke’s theory does not.

vs Job Characteristics Model (Hackman & Oldham 1976)

The Job Characteristics Model (JCM) ships the same year as Locke’s chapter and answers a related question from the opposite direction. JCM specifies which five structural features of a job (Skill Variety, Task Identity, Task Significance, Autonomy, Feedback) produce three psychological states (felt meaning, felt ownership, knowledge of results) that drive outcomes including job satisfaction. JCM is prescriptive; it tells you which structural levers to pull. Locke is descriptive; it tells you what the worker’s mental computation looks like once any structural state has been delivered.

The two theories compose cleanly. Use JCM to identify which structural variables to redesign, then use Locke to predict which facet’s (Want − Have) × Importance product will move when the JCM redesign lands. A job redesign that increases Skill Variety should move the JDI Work Itself facet upward; if the survey shows it did not, either the perceived Have did not change (a framing failure) or the worker’s Importance weight on Work Itself was lower than the survey assumed.

vs Self-Determination Theory (Ryan & Deci 2000)

Self-Determination Theory (SDT) is the dynamic-process counterpart to Locke’s static-state model. Satisfaction is a snapshot of how a worker feels about the job right now. SDT specifies the three psychological needs (autonomy, competence, relatedness) whose satisfaction drives the higher-quality intrinsic motivation that makes satisfaction sustainable, not just instantaneous.

In practice this means a job that scores high on Locke’s facets but low on SDT need-satisfaction will deliver short-term satisfaction that decays. The worker likes the Pay and the Coworkers, but the work itself does not satisfy the autonomy-competence-relatedness triad and the affective tone of the job drifts downward over months. Locke’s instrument catches the drift after it has happened; SDT predicts it before.

vs Employee Engagement (Kahn 1990; Gallup Q12)

Employee Engagement, as Kahn 1990 defined it, is the personal employment of physical, cognitive, and emotional energy into the work role. Engagement is something the worker does. Satisfaction is something the worker feels. They correlate around r = 0.40, which is high enough to confuse practitioners and low enough that they are not interchangeable.

The clearest case where they diverge is the Quiet Quitter: high job satisfaction (the Want-Have gap is closed, Pay is fine, Coworkers are good), low engagement (the worker is no longer investing discretionary effort). Locke’s instrument shows green; Kahn’s shows red. A company that measures only one of these will miss half of the worker population that needs intervention.

vs Workplace Burnout (Maslach & Leiter 1997)

Workplace Burnout sits on the opposite end of the affective spectrum from satisfaction. Maslach and Leiter’s three dimensions (Emotional Exhaustion, Cynicism, Reduced Personal Accomplishment) describe what happens when sustained dissatisfaction across the Six Areas of Worklife (workload, control, reward, community, fairness, values) damages the worker’s capacity to feel anything about the job at all. Burnout is what Range-of-Affect dissatisfaction looks like after the affect-processing machinery itself is depleted.

The diagnostic implication: a worker scoring low on satisfaction AND low on Maslach’s Reduced Personal Accomplishment dimension is in the acute dissatisfaction phase, and the intervention is to close a Want-Have gap. A worker scoring low on satisfaction AND high on Reduced Personal Accomplishment is in the chronic burnout phase, and closing the gap will not be enough; recovery requires structural rest and meaning repair. Same survey input, two different interventions.

Job Satisfaction in the Real World

The Range-of-Affect equation works on paper. The four real-world settings below are where it earns or fails its practitioner credibility.

Knowledge Work

Knowledge workers (software engineers, designers, lawyers, analysts, academics) typically weight Work Itself and Autonomy as their top Importance facets. The labor-market mobility of this group is high, which means dissatisfaction translates to attrition faster than in lower-mobility populations. Companies that compete for knowledge-worker satisfaction by raising Pay without addressing Work Itself often see Pay-satisfaction rise and overall-satisfaction stay flat, because Pay is not the bottleneck facet for that worker segment.

The 2026 software engineering hiring market is the cleanest natural experiment. Pay packages have inflated 40-60% above 2019 levels, and developer engagement scores (Forsgren 2021 SPACE metrics, Forsgren-Storm-Smith-Murphy-Hall 2022 DevEx) have moved less than expected. The Want term scaled with Pay, the Have term scaled with Pay, the gap closed only modestly. Meanwhile Autonomy and Meaningful Work scores moved less because the underlying work design did not change. Locke’s equation predicts this outcome; the practitioners who built compensation-only retention programs did not.

Healthcare

Healthcare workers (nurses, physicians, technicians) typically weight Coworker quality, Supervision, and Meaningful Work as top facets. The healthcare attrition crisis post-2020 was not primarily a Pay crisis. Aiken et al 2014 in The Lancet showed that nurse engagement and retention tracked Supervision quality, Operating Conditions (staffing ratios), and Coworker fabric far more closely than they tracked Pay across European hospitals. Hospitals that responded with sign-on bonuses and retention pay saw modest short-term improvement; hospitals that simultaneously fixed staffing ratios, supervisor training, and shift-handoff protocols saw sustained improvement.

This is the cleanest validation of facet-weighted Locke that the post-pandemic decade produced. The dominant Importance facets in healthcare are not Pay, and any intervention that targets Pay alone will not move the satisfaction equation for the workforce that needs the intervention.

Gig Economy and Platform Work

Platform workers (rideshare, delivery, marketplace sellers, freelance creators) face a satisfaction equation in which the Have term is dynamic and the Want term has limited reference data. A rideshare driver does not have a stable peer-reference group whose Pay and Working Conditions they can observe; the platform deliberately atomizes the workforce. The result is that Want-value formation is driven by media reports, social-media testimonials, and personal projection, all of which produce wide Want-Have gaps and chronic facet-level dissatisfaction.

This pattern matches the empirical record: Berger, Frey, Levin, and Danda 2019 found that rideshare drivers report lower job satisfaction than comparable hourly employees with similar Pay, primarily on the Supervision facet (algorithmic management substitutes for human supervisor) and the Coworker facet (no Coworkers exist). The Importance weights on those facets are not zero, and a job that delivers zero on a positively-weighted facet cannot reach satisfaction parity with a job that delivers some on it.

Public Sector and Education

Public-sector and education workers typically weight Mission, Meaningful Work, and Coworker facets heavily, with Pay weighted lower than private-sector comparison. The empirical record (Bright 2008 public service motivation studies, Perry and Wise 1990) shows public-sector workers accepting 10-25% below-market Pay in exchange for higher delivery on Mission, Stability, and Coworker stability. When public-sector reform pushes toward Pay-for-Performance and reduces Stability, the satisfaction equation flips: the Pay improvement is small (PfP rarely closes the private-sector gap), and the Stability and Mission losses are large. Net effect is satisfaction decline, attrition rise, and exactly the workforce hollowing the reforms were supposed to prevent.

The Elephant in the Room

Quiet Quitting was Locke’s equation rendered visible in 2022. The viral TikTok of August 2022 (Khan 2022) named what tens of millions of workers were already doing: staying in their jobs at the contractual minimum effort while withdrawing the discretionary investment that engagement programs had been measuring as a proxy for satisfaction.

Inside Locke’s framework, Quiet Quitting is what the Range-of-Affect equation looks like when Importance weights collapse to near-zero on Work Itself, Promotion, and Recognition while staying high on Pay and Stability. The (Want − Have) × Importance product on the deweighted facets becomes small in absolute terms even if the Want-Have gap is large; the worker stops caring about the gap. Simultaneously, the (Want − Have) × Importance product on Pay and Stability stays modest because the gaps are small and the workers stay. Net satisfaction score: not low enough to trigger HR intervention. Net engagement score: collapsing.

The Great Resignation phase (2021-2022) was the opposite signal from the same equation. Workers who had collected pandemic-era information shocks (remote work was possible, peer salaries were visible, life-goal Importance weights had been reshuffled by 2020) discovered their (Want − Have) × Importance products had grown larger than they thought, and they acted. The aggregate quit rate in the US reached 3.0% per month in November 2021 (Bureau of Labor Statistics JOLTS), the highest in two decades. The same workers did not become structurally different. Their Want terms updated, and Locke’s equation re-balanced.

The honest read on both phenomena is that they were predictable, given Locke’s theory and the 2020-2022 information shocks. The “surprise” was practitioner surprise; the equation did not surprise. Companies that ran facet-with-Importance-weighting surveys in early 2021 saw the Want-term drift coming and adjusted retention strategy. Companies that ran Likert-only satisfaction surveys did not, and discovered the drift after their workforce had already used it.

Imports the Libertarian Paternalism pillar publicity-test ethics frame: would your job-satisfaction intervention survive being described accurately to the workers it targets, including which facets you are deliberately moving and which you are not? Most engagement programs would fail this test. They optimize the easy facets (Pay, Recognition, Perks) and avoid the hard ones (Work Itself, Autonomy, Mission). The publicity test forces the choice into the open and makes the worker the decider, not the program designer.

How to Apply Job Satisfaction with the Octalysis Framework

Locke’s equation tells you what is happening. The Octalysis Framework tells you which behavioral lever moves each facet. The crosswalk below assigns each of the JDI’s five facets to the Core Drives that produce its Have value, and names the structural surfaces that move the Want term and the Importance weight.

Work Itself: Core Drive 2 + Core Drive 3

Work Itself satisfaction is what the JCM five characteristics combine to produce, viewed from inside the worker’s appraisal. It runs primarily on Core Drive 2 (CD2): Development & Accomplishment (the worker grows from doing the work) and Core Drive 3 (CD3): Empowerment of Creativity & Feedback (the worker shapes the work and sees feedback close the loop). When Work Itself satisfaction is low, the structural surface to design is one of: (a) decision-rights expansion (more CD3 latitude), (b) skill-growth runway (more CD2 development loop closure), (c) feedback channel quality (faster, more honest, more useful feedback per unit of work).

The Importance weight on Work Itself moves with worker tenure and life stage. Early-career workers often under-weight it (Pay is more salient), mid-career workers often peak-weight it, late-career workers re-weight toward Coworkers and Mission. The same structural intervention will produce different satisfaction effects across the worker-tenure distribution, which is why facet-level intervention design must be segmented by tenure.

Pay: Core Drive 4 + Core Drive 8

Pay satisfaction is the facet most workers say matters most and the facet that moves overall satisfaction least once a floor is met (Easterlin 1974 and follow-up work; Kahneman and Deaton 2010 set the satiation point around $75K-$95K in US 2010 dollars, though more recent work suggests the satiation point is higher and the relationship more continuous). Pay runs on Core Drive 4 (CD4): Ownership & Possession (the worker’s fair share of the value they create) and Core Drive 8 (CD8): Loss & Avoidance (the dread of underpayment relative to peers).

The Want term on Pay is the most volatile term in the entire Range-of-Affect equation because peer-reference shocks (a salary leak, a market repricing, a pay-transparency mandate) can move Want without changing Have. Cullen and Perez-Truglia 2022 showed median-or-below earners reduce effort by 5.7% after seeing they are below median in their pay-transparency-mandated peer group, even though their Pay did not change. The Have stayed; the Want moved; the gap opened; the satisfaction and effort dropped.

Opportunities for Promotion: Core Drive 2 + Core Drive 6

Promotion satisfaction tracks the worker’s perception of advancement runway. It runs on Core Drive 2 (CD2): Development & Accomplishment (the long-arc growth path) and Core Drive 6 (CD6): Scarcity & Impatience (the constraint that only a few promotions exist). Companies with flat structures and limited promotion slots can produce strong Work Itself satisfaction but weak Promotion satisfaction; the Importance weight on Promotion is the swing variable that determines whether this pattern is fine or fatal for a given worker.

The clearest structural fix for low Promotion satisfaction without inventing slots that do not exist is to redesign the development-loop visibility. If workers can see their skill growth, scope expansion, and reputational accumulation independent of title change, the Have on Promotion increases without the title structure changing. The Want term may also move downward if the public discourse around “promotion” stops being the load-bearing measure of career success in the company.

Supervision: Core Drive 5 + Core Drive 8

Supervision satisfaction runs on Core Drive 5 (CD5): Social Influence & Relatedness (the worker’s relational bond with their manager) and Core Drive 8 (CD8): Loss & Avoidance (the manager’s perceived fairness as risk insurance). Bad supervisors produce dissatisfaction far out of proportion to the time the worker spends with them, because every interaction is high-stakes affective signal. Good supervisors produce satisfaction that buffers other facet dissatisfactions: workers with weak Pay but strong Supervision often stay; workers with strong Pay but weak Supervision often leave (the “people quit managers, not companies” cliche has empirical support; Branham 2005 and Gallup engagement data).

Supervisor training is the cheapest high-impact intervention available to most companies. The classic Hersey and Blanchard Situational Leadership training raises the floor on Supervision satisfaction more reliably than nearly any other intervention because it equips managers to adjust style to follower readiness, which is the Have signal the worker is reading.

Coworkers: Core Drive 5 + Core Drive 1

Coworker satisfaction runs on Core Drive 5 (CD5): Social Influence & Relatedness (the peer-relational fabric) and Core Drive 1 (CD1): Epic Meaning & Calling (the sense of working with people toward something larger than the paycheck). The Coworker facet is the only facet where the worker’s satisfaction is jointly determined by people the company hires; bad coworkers contaminate satisfaction across other facets through spillover, and good coworkers buffer dissatisfaction across other facets through the same mechanism.

The structural surfaces that move Coworker satisfaction are hiring quality (the most important), team composition stability (Edmondson 2018 on teaming and psychological safety), and shared-project meaning (CD1 ties). Remote-first companies that lose the casual-interaction surface area routinely see Coworker satisfaction decline; the structural fix is intentional design of low-stakes interaction surfaces (cohort programs, project clusters, optional in-person gatherings) rather than mandatory return-to-office that breaks Autonomy.

Moderating Core Drives: Core Drive 7 and CD1 Spillover

Core Drive 7 (CD7): Unpredictability & Curiosity sits underneath every facet as the engagement variable. Predictable jobs satisfy the Want-Have equation but feel flat; surprise-graded jobs (new projects, rotational programs, learning expeditions) keep CD7 active and prevent satisfaction from going stale even when the Range-of-Affect equation is balanced. CD1 spillover means a worker who experiences strong Epic Meaning & Calling on their work absorbs facet shortfalls elsewhere more tolerantly than a worker with weak CD1. Mission alone cannot substitute for facet shortfalls (the public-sector finding above) but it modulates how much each shortfall weighs.

The 6-Step Job Satisfaction × Octalysis Audit

This is the structural test that turns Locke’s diagnostic into a redesign tool.

1. Measure facet satisfaction with separated Importance weights. Use the JDI five facets (or MSQ 20 if you need granularity) AND a parallel Importance survey on the same facets. Do not collapse them into one Likert question. Compute (Want − Have) × Importance per facet per worker.
2. Segment by Importance pattern, not by demographic. Workers with high Importance on Work Itself + Autonomy are a different design segment than workers with high Importance on Pay + Stability, even if they have identical roles and demographics. Pool by Importance pattern to find the bottleneck for each segment.
3. Map the highest (Want − Have) × Importance product to its Core Drive. Pay = CD4 + CD8. Work Itself = CD2 + CD3. Promotion = CD2 + CD6. Supervision = CD5 + CD8. Coworkers = CD5 + CD1. The Core Drive identifies the structural surface to design.
4. Design the structural intervention with the publicity test applied. Would you describe this intervention accurately to the workers it targets, including which facet you are moving and why? If not, redesign it. (Imports Libertarian Paternalism publicity-test convention.)
5. Anchor the intervention against Coworker spillover and CD1 spillover. The intervention will land inside a Coworker fabric and inside a Mission ecology. A redesign that ignores either will produce ripple effects that distort the satisfaction signal on adjacent facets.
6. Re-measure at six months on facet satisfaction AND facet Importance. If the targeted facet moved upward and Importance stayed roughly stable, the intervention worked. If Importance shifted downward (the worker stopped caring about the facet you fixed), the intervention shifted the equation, not closed the gap.

Practical Steps for Improving Job Satisfaction

The audit above is structural. The six practical steps below are the operational version a manager or engagement lead can apply this quarter.

1. Separate Importance from satisfaction in your survey. Add a parallel Importance question for every facet you measure. Without the Importance term, you have Locke’s instrument without Locke’s equation.
2. Cross-check facet survey results against objective behavior. Pay satisfaction against actual peer-pay disclosure, Promotion satisfaction against tenure-to-promotion data, Coworker satisfaction against team-stability metrics. Common-method variance is your default failure mode.
3. Train managers on Supervision behaviors that move the Have on CD5. Active listening, calibrated feedback, calendar-protected 1:1s, and Hersey-Blanchard situational style flex move Supervision satisfaction faster than any policy change. Cheapest high-impact intervention available.
4. Audit Work Itself satisfaction at the role-design layer, not the perk layer. Free lunch does not move Work Itself satisfaction. Decision-rights expansion, scope ownership, and feedback-channel quality do.
5. Treat Pay satisfaction as a Want-management problem in transparent markets. Pay-transparency mandates (Colorado, NYC, California, Washington, EU 2026 directive) made the Want term observable to every worker. Pay raises that do not address the median-peer reference will not close the gap. Either re-anchor compensation against the visible market or accept that Pay satisfaction will trend down while Pay levels rise.
6. Run a Quiet-Quitting check separately from satisfaction. Add a discretionary-effort question (“compared to six months ago, how much extra effort are you putting into your work?”) that catches the Importance-collapse signal Locke’s instrument alone misses.

Job Satisfaction Was the Beginning, Not the End

Locke’s 1976 chapter gave the field a clean equation and a generation of instruments. What it did not give the field was a theory of how to move the equation in a specific direction at a specific company on a specific timeline. The 50 years of follow-on work, including JCM, SDT, Engagement, Burnout, and the Octalysis Framework that this post builds on, are the field’s collective attempt to fill in the missing prescriptive layer.

For any practitioner building a behavioral-design system around worker satisfaction in 2026, the right move is to use Locke as the diagnostic, JCM and SDT as the structural prescription, Engagement and Burnout as the dynamic monitoring, and the Octalysis crosswalk above as the Core-Drive-level lever assignment. None of these tools by itself is sufficient. Together they give a behavioral designer enough structure to redesign work intentionally instead of reflexively reaching for the next perk or compensation tweak that the Range-of-Affect equation already predicts will fail.

If your company is running an engagement program and you have not asked which facet has the largest (Want − Have) × Importance product for which worker segment, you do not have a satisfaction strategy. You have a survey. Locke’s equation is what turns the survey into the strategy.

Frequently Asked Questions

What is the Range-of-Affect Theory in one sentence?

Job satisfaction on any facet equals the gap between what the worker wants and what the job delivers, multiplied by how much that facet matters to the worker; overall job satisfaction is the sum of facet satisfactions across the facets that matter to that worker.

How is Locke’s theory different from Herzberg’s Two-Factor Theory?

Herzberg 1959 claimed hygiene factors could only cause dissatisfaction and motivators could only cause satisfaction, an asymmetry the data did not support. Locke 1976 showed any facet can produce satisfaction or dissatisfaction depending on the Want-Have gap weighted by Importance, collapsing Herzberg’s asymmetry into a single equation with three terms.

Which job satisfaction instrument should I use: JDI, MSQ, or JSS?

Use the JDI if you want the most psychometrically defensible five-facet measure and can afford the license. Use the MSQ if you need 20-facet granularity and a free academic license. Use the JSS if you work in a public-sector, nonprofit, or human-service setting where the JDI’s private-sector framing does not fit. All three operationalize Locke’s theory; the difference is the facet decomposition and the cost.

Does job satisfaction predict job performance?

Yes, but the correlation is smaller than practitioners assume. Judge et al 2001 meta-analyzed 312 samples and found r = 0.30, which means satisfaction and performance share about 9 percent of variance. Investing in satisfaction primarily to drive performance is a weak bet; investing in satisfaction for retention, life satisfaction, customer treatment, and reduced absenteeism is a stronger one.

Is job satisfaction the same as employee engagement?

No. Satisfaction is something the worker feels; engagement is something the worker does. They correlate around r = 0.40, which is high enough to confuse practitioners and low enough that they are not interchangeable. A Quiet Quitter often scores high on satisfaction and low on engagement; a passionate worker in a difficult role often scores low on satisfaction and high on engagement.

Why does Pay satisfaction not rise with Pay?

Locke’s Want term scales with peer reference. When Pay rises across the market, the Want term rises along with it (the worker compares against the new peer median, not the old one), and the gap stays roughly constant. This is why a company can give significant raises and see Pay satisfaction stay flat or even decline, especially when pay-transparency mandates make the peer reference more visible.

How does Job Satisfaction theory explain Quiet Quitting?

Quiet Quitting is what the Range-of-Affect equation looks like when Importance weights collapse to near-zero on Work Itself, Promotion, and Recognition while staying high on Pay and Stability. The (Want − Have) × Importance product on the deweighted facets becomes small even if the gap is large; the worker stops caring about the gap and stays. Satisfaction scores look fine; engagement scores collapse.

How does it apply across cultures?

The Range-of-Affect equation generalizes, but the Importance weights shift across cultures. Robert et al 2000 and the GLOBE study findings show collectivist cultures weight Coworker and Supervision more heavily, individualist cultures weight Work Itself and Promotion more heavily. A multinational using one survey instrument across subsidiaries should expect different (Want − Have) × Importance equilibria in each location, even with the items translated correctly.

What is the fastest way to raise job satisfaction in a team?

Train the manager. Supervision satisfaction is the most under-invested facet at most companies, and supervisor training has the strongest expected effect per dollar of any common intervention. Active listening, calibrated feedback, calendar-protected 1:1s, and Hersey-Blanchard situational style flex move Supervision satisfaction in weeks. Pay, Promotion, and Work Itself interventions take quarters to years to land.

Does the Octalysis crosswalk replace Locke’s theory?

No. It extends it. Locke described what a satisfied worker looks like; he did not specify which structural levers move which facet. The Octalysis crosswalk assigns each facet to its Core Drives (Work Itself = CD2 + CD3, Pay = CD4 + CD8, Promotion = CD2 + CD6, Supervision = CD5 + CD8, Coworkers = CD5 + CD1) so that a designer can pick the lever, not just diagnose the gap.

References

1. Locke, E. A. (1976). The nature and causes of job satisfaction. In M. D. Dunnette (Ed.), Handbook of Industrial and Organizational Psychology (pp. 1297-1349). Chicago: Rand McNally. Field-naming chapter.
2. Locke, E. A. (1969). What is job satisfaction? Organizational Behavior and Human Performance, 4(4), 309-336. Foundational definition.
3. Smith, P. C., Kendall, L. M., & Hulin, C. L. (1969). The Measurement of Satisfaction in Work and Retirement. Chicago: Rand McNally. Job Descriptive Index (JDI) canonical.
4. Weiss, D. J., Dawis, R. V., England, G. W., & Lofquist, L. H. (1967). Manual for the Minnesota Satisfaction Questionnaire. Minnesota Studies in Vocational Rehabilitation, 22. MSQ canonical.
5. Spector, P. E. (1985). Measurement of human service staff satisfaction: Development of the Job Satisfaction Survey. American Journal of Community Psychology, 13(6), 693-713. JSS canonical.
6. Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K. (2001). The job satisfaction-job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127(3), 376-407. Canonical meta-analysis.
7. Tett, R. P., & Meyer, J. P. (1993). Job satisfaction, organizational commitment, turnover intention, and turnover: Path analyses based on meta-analytic findings. Personnel Psychology, 46(2), 259-293. Turnover meta.
8. Judge, T. A., & Watanabe, S. (1993). Another look at the job satisfaction-life satisfaction relationship. Journal of Applied Psychology, 78(6), 939-948. Life-satisfaction correlation.
9. Herzberg, F. (1968). One more time: How do you motivate employees? Harvard Business Review, 46(1), 53-62. Two-Factor Theory canonical.
10. Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology, 82(2), 247-252. Single-item validation.
11. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review. Journal of Applied Psychology, 88(5), 879-903. Common-method variance canonical.
12. Robert, C., Probst, T. M., Martocchio, J. J., Drasgow, F., & Lawler, J. J. (2000). Empowerment and continuous improvement in the United States, Mexico, Poland, and India. Journal of Applied Psychology, 85(5), 643-658. Cross-cultural canonical.
13. Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work. Organizational Behavior and Human Performance, 16(2), 250-279. Job Characteristics Model canonical.
14. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation. American Psychologist, 55(1), 68-78. SDT canonical.
15. Kahn, W. A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33(4), 692-724. Employee Engagement canonical.
16. Maslach, C., & Leiter, M. P. (1997). The Truth About Burnout. San Francisco: Jossey-Bass. Six Areas of Worklife canonical.
17. Adams, J. S. (1963). Toward an understanding of inequity. Journal of Abnormal and Social Psychology, 67(5), 422-436. Equity Theory canonical.
18. Cullen, Z., & Perez-Truglia, R. (2022). How much does your boss make? The effects of salary comparisons. Journal of Political Economy, 130(3), 766-822. Pay-transparency canonical.
19. Hare, T. A., Camerer, C. F., & Rangel, A. (2009). Self-control in decision-making involves modulation of the vmPFC valuation system. Science, 324(5927), 646-648. vmPFC valuation canonical.
20. Schultz, W. (2016). Dopamine reward prediction-error signalling: A two-component response. Nature Reviews Neuroscience, 17(3), 183-195. Reward-prediction-error canonical.
21. McEwen, B. S. (1998). Stress, adaptation, and disease: Allostasis and allostatic load. Annals of the New York Academy of Sciences, 840, 33-44. Allostatic load canonical.
22. Aiken, L. H., Sloane, D. M., Bruyneel, L., et al. (2014). Nurse staffing and education and hospital mortality in nine European countries. The Lancet, 383(9931), 1824-1830. Healthcare workforce canonical.
23. Berger, T., Frey, C. B., Levin, G., & Danda, S. R. (2019). Uber happy? Work and well-being in the “Gig Economy.” Economic Policy, 34(99), 429-477. Platform-worker satisfaction canonical.
24. Chou, Y. (2015). Actionable Gamification: Beyond Points, Badges, and Leaderboards. Octalysis Media. Octalysis canonical.

Related Reading

• Job Characteristics Model: S-Tier Behavioral Designer’s Guide — the structural-design counterpart to Locke’s diagnostic.
• Self-Determination Theory: S-Tier Behavioral Designer’s Guide — the dynamic-process counterpart to Locke’s static-state model.
• Employee Engagement: S-Tier Behavioral Designer’s Guide — Locke’s closest neighbor in the engagement-research field.
• Workplace Burnout: S-Tier Behavioral Designer’s Guide — the failure-mode neighbor when Range-of-Affect dissatisfaction depletes the affect-processing machinery itself.
• Equity Theory: S-Tier Behavioral Designer’s Guide — the comparative-fairness mechanism underneath Pay satisfaction and the Want-term update during transparency shocks.
• The Octalysis Framework — the 8-Core-Drive system that names the structural levers moving each Locke facet.

WOULD YOU LIKE YU-KAI CHOU TO WORK WITH YOUR ORGANIZATION?