Mental Health, Stress, Workplace Productivity, And Worker Welfare

Learning Objectives¶

By the end of Week 4, students should be able to:

1. explain why mental health is a labor-market force rather than an individual residual;

2. distinguish depression, anxiety, stress, pain, and substance use as conditions that can affect attendance, effort, productivity, job choice, safety, retention, and welfare;

3. separate four mechanisms: mental health affecting work, work affecting mental health, treatment and stigma, and measurement and welfare;

4. evaluate absenteeism, presenteeism, performance, quits, accommodations, prescriptions, claims, and symptom scales as imperfect labor-relevant measures;

5. identify why reverse causality, dynamic selection, diagnosis intensity, treatment selection, co-morbidity, socioeconomic confounding, and differential reporting make causal interpretation difficult;

6. design a research project that names the mental-health object, labor margin, identification strategy, and welfare object.

The central question is this: how do mental health and stress affect productivity, job quality, retention, and welfare inside labor markets?

Opening Orientation¶

Mental health enters labor markets through multiple margins at once. Depression, anxiety, stress, pain, and substance use can affect attendance, concentration, interpersonal functioning, safety, effort, job search, treatment seeking, disclosure, and the value of work itself. A worker may remain employed while losing productivity through presenteeism, missing work through episodic symptoms, avoiding promotion because the job requires more conflict or monitoring, or quitting a job that worsens distress. Firms may observe only fragments of this process: absences, mistakes, performance ratings, accommodation requests, turnover, or claims.

The relationship also runs the other way. Jobs, supervision, workload, autonomy, conflict, isolation, schedule volatility, job insecurity, and meaning at work can improve or harm mental health. That makes Week 4 a bridge between Labor I’s concern with worker welfare and job amenities and Labor II’s concern with productivity, supervision, personnel policy, and firm organization. The lecture is not a general clinical mental-health lecture. The organizing objects are labor supply, productivity, job quality, retention, safety, supervision, treatment access, stigma, and welfare.

Bridge¶

Week 1 framed health as work capacity, risk, and measurement. Week 2 studied disability and chronic conditions as dynamic constraints on labor supply, job choice, accommodations, and employer response. Week 3 moved to fertility, mortality, caregiving, aging, and lifecycle labor allocation. Week 4 moves inside workplaces and asks how less visible, more stigmatized, and harder-to-measure mental-health conditions affect productivity, retention, supervision, job quality, safety, and worker welfare.

The key continuity is that health is not a single coefficient. Mental health can reduce work capacity, raise the disutility of particular tasks, change job-search behavior, alter demand for treatment or accommodations, and reshape the worker’s outside options. Work can also generate mental-health outcomes through stress exposure, control, conflict, isolation, and meaning. A clean research design must say which direction and which margin it is identifying.

Field Core¶

Mental Health As A Labor-Market Object¶

A labor-economics starting point is to treat mental health as affecting both the capacity to work and the utility cost of work. Depression may lower energy, concentration, and persistence. Anxiety may make customer contact, monitoring, uncertainty, or conflict more costly. Stress can affect sleep, attention, error rates, and health investments. Pain and substance use can interact with mental health and labor-force attachment, creating cycles of nonemployment, reduced hours, lower job quality, and weaker recovery Krueger, 2017Case & Deaton, 2020.

The main labor margins are:

• Absenteeism: missed days, sick leave, disability leave, and episodic nonattendance.

• Presenteeism: being at work but operating with reduced concentration, speed, reliability, or social functioning.

• Productivity and safety: errors, task completion, customer outcomes, injury risk, and performance ratings.

• Retention and career paths: quits, job-to-job moves, promotions, occupational downgrading, and exits.

• Supervision and firm policy: monitoring, scheduling, accommodations, leave, employee assistance, and managerial discretion.

• Worker welfare: the psychic cost of work, job quality, stigma, treatment access, and the value of safer or more supportive jobs.

The productivity side is concrete in Bubonya, Cobb-Clark, and Wooden, who study mental health and workplace productivity margins including absenteeism and presenteeism Bubonya et al., 2017. The broader labor point is larger: a worker can have unchanged wages but worse welfare, unchanged employment but lower productivity, or improved retention after treatment without an immediate earnings gain.

Stigma, Disclosure, And Treatment As Labor-Market Mechanisms¶

Stigma is not just background culture. It is a mechanism that changes information, investment, matching, and career dynamics. Workers may hide symptoms because disclosure could be interpreted as unreliability, poor fit, weak leadership potential, safety risk, or low commitment. They may avoid treatment because treatment records, schedule demands, medication use, or counseling appointments feel career-relevant. Firms may fail to offer accommodations because mental-health constraints are less visible than physical limitations, or because managers lack information about how to respond.

Stigma can move several labor outcomes at once:

• Disclosure: workers may not reveal symptoms until performance, attendance, or conflict has already deteriorated.

• Help-seeking and treatment access: available care may not translate into use if workers fear reputation, promotion, licensing, or job-security penalties.

• Accommodations: flexible schedules, workload changes, quiet space, leave, or supervisory changes may be underused when workers expect negative signals.

• Attendance and retention: untreated symptoms can raise absences and quits, while treatment may require time away from work in the short run.

• Career choices: workers may sort away from high-conflict, public-facing, monitored, or high-autonomy roles depending on symptoms, stigma, and anticipated support.

This is why treatment access and treatment use are endogenous. Even when services exist, the treated group is not a random subset of workers with need. Experimental destigmatization work and intervention studies are useful because they show how beliefs, information, and access can change help-seeking and labor-relevant outcomes, but the labor evidence is still thin relative to the size of the problem Lacey et al., 2022Lund et al., 2024.

Work Causing Mental Health Outcomes: Job Quality, Supervision, And Meaning¶

Mental health is also an outcome of labor-market conditions. Jobs expose workers to supervision, schedules, workload, effort-reward tradeoffs, risk, conflict, isolation, autonomy, and meaning. The same observed hours can have very different mental-health implications depending on whether the work is predictable, controlled, socially supported, meaningful, and safe.

Important workplace channels include:

• Workload and time pressure: high demands, overtime, and unrealistic targets can raise stress and reduce recovery time.

• Autonomy and control: control over pace, methods, scheduling, and task order can buffer stress.

• Supervision and monitoring: supervision can provide support and clarity, but surveillance or arbitrary management can raise anxiety and quits.

• Conflict and harassment: interpersonal conflict, customer abuse, discrimination, and harassment can make the workplace a source of mental-health harm.

• Isolation and remote work: social isolation may lower support and belonging even when flexibility improves some constraints.

• Meaning and identity: workers may experience better mental health when effort connects to purpose, values, or accomplishment Kranton & Thomas, 2025.

Job quality is therefore a mental-health object, not merely a wage shifter. Evidence on job quality and mental health highlights that the quality of work can matter as much as or more than the quantity of work Wang et al., 2022. For labor economists, this connects amenities and compensating differentials to welfare: a higher wage may compensate for some stress, but observed wages may not reveal the mental-health cost borne by workers with few outside options.

Measurement And Data: What Is The Mental-Health Object?¶

Mental-health labor research is unusually measurement-intensive. Common objects include:

• Survey measures: PHQ-9, GHQ, CES-D, K6, self-reported distress, self-reported stress, anxiety, burnout, and life satisfaction.

• Administrative proxies: prescriptions, insurance claims, diagnosis records, sick leave, disability claims, treatment episodes, hospitalizations, and provider visits.

• Workplace measures: absenteeism, presenteeism, performance ratings, output, errors, safety incidents, quits, promotions, supervisor reports, and accommodations.

• Treatment and access measures: provider supply, waiting time, coverage, employee assistance programs, appointment availability, copayments, and leave flexibility.

These measures are not substitutes. Symptom scales can capture distress without diagnosis, but reporting may differ by gender, occupation, culture, stigma, and survey mode. Prescriptions and claims capture treatment contact, not untreated need. Sick leave and disability claims reflect symptoms, job rules, access, diagnosis intensity, workplace norms, and program incentives. Presenteeism is attractive because it captures on-the-job productivity loss, but it is hard to measure because self-reports, supervisor ratings, and output data each miss different dimensions.

The key identification challenges are reverse causality, dynamic selection, diagnosis intensity, treatment selection, co-morbidity with physical health or pain, socioeconomic confounding, and differential reporting. Credible causal evidence is still limited in many areas. Stronger designs often use treatment-access shocks, waiting-time shocks, policy changes in coverage or leave, pandemic and workplace shocks, job-loss shocks, firm reorganization, or structural and job-search models that jointly model mental health and labor-market choices Jolivet et al., 2025Gallo et al., 2000Gallo et al., 2006Picciotto & others, 2024.

Productivity, Retention, Safety, And Worker Welfare¶

Mental health is often introduced through productivity loss, but the labor-market object is broader. A firm may care about output, absences, retention, safety, customer interaction, and team coordination. Workers also care about whether the job worsens distress, whether disclosure is safe, whether treatment can be scheduled, whether supervisors interpret symptoms fairly, and whether wages compensate for stress. A policy that improves treatment take-up, attendance, or retention may improve welfare even if wages are unchanged.

This distinction matters for interpretation:

• A decline in absenteeism can be productivity gain, improved health, reduced leave access, or stronger presenteeism pressure.

• A decline in quits can indicate better job quality, weaker outside options, lower search capacity, or treatment-supported attachment.

• A performance improvement after treatment may reflect symptom relief, accommodation, changed tasks, or selection into treatment.

• A wage premium in a stressful job may be compensation for disamenities, rent sharing, selection of resilient workers, or limited worker options.

The welfare question cannot be read off wages alone. It requires attention to job quality, mental-health trajectories, treatment access, stigma, and worker choice sets.

Frontier Research Architecture¶

The field is open because many central questions remain underidentified. Students should treat this as an opportunity rather than as a reason to avoid the topic. Useful research designs make four distinctions explicit:

1. Mental health affecting work: symptoms or treatment change attendance, productivity, search, quits, and wages.

2. Work affecting mental health: job loss, supervision, workload, remote work, conflict, meaning, or policy changes alter distress and treatment.

3. Treatment and stigma: access, beliefs, disclosure, and accommodations change both mental health and labor outcomes.

4. Measurement and welfare: observed productivity or employment does not fully capture worker well-being.

Promising designs include treatment-access or provider-supply shocks, waiting-time variation, policy changes in coverage or paid leave, job-loss event studies, pandemic-era workplace shocks, matched employer-employee designs around management practices, randomized destigmatization or information interventions, and structural or job-search approaches that model mental health, job choice, and treatment together. Each design should state what it identifies and what remains bundled.

Research Lab¶

The Week 4 research lab follows Reproduce -> Diagnose -> Transfer. The primary anchor is Bubonya, Cobb-Clark, and Wooden on mental health and productivity at work because it provides a feasible empirical object: mental-health measures linked to absenteeism, presenteeism, job characteristics, and productivity-related outcomes Bubonya et al., 2017. The challenge anchor is Kranton and Thomas on meaning, identity, and employee mental health, which pushes the lab toward job design, values alignment, and worker welfare Kranton & Thomas, 2025.

The lab is not an official replication package for either paper. It uses deterministic synthetic data to teach the empirical logic conservatively: symptom measures are correlated with productivity and job quality; stigma and treatment affect reporting and use; and causal interpretation remains limited without stronger variation.

Reproduce. Students recover a bounded Bubonya-style descriptive profile. Outcomes include absenteeism days, presenteeism, performance, quit risk, accommodation use, and a welfare index. Students compare workers with low and high distress and then stratify by job strain, autonomy, stigma, and treatment access.

Diagnose. Students classify each result as a possible productivity effect, reporting artifact, job-quality effect, treatment-selection margin, stigma margin, or dynamic-selection problem. They explicitly ask whether mental health is affecting work, work is affecting mental health, or both.

Transfer. Students redesign the logic for one frontier setting: a treatment-access shock, a waiting-time shock, a job-loss event study, a pandemic or workplace reorganization shock, a disclosure or destigmatization intervention, or a structural job-search model. The transfer task names the latent mental-health object, labor outcome, identifying variation, main threat, and welfare margin.

Methods Box¶

Reading Ladder And References¶

Core framing. Start with Currie and Madrian for the classic health-labor map, Bound for measurement discipline, and Pintor for a recent review of health and labor outcomes Currie & Madrian, 1999Bound, 1991Pintor, 2024.

Mental health and workplace productivity. Use Ettner, Frank, and Kessler for early labor-market evidence on psychiatric disorders, then Bubonya, Cobb-Clark, and Wooden for absenteeism and presenteeism at work Ettner et al., 1997Bubonya et al., 2017.

Pain, substance use, and labor-force attachment. Read Krueger and Case and Deaton as labor-relevant background on pain, opioids, nonemployment, and despair, while keeping the lecture focused on work, job quality, and welfare rather than broad clinical epidemiology Krueger, 2017Case & Deaton, 2020.

Work causing mental health. Use Wang and coauthors on job quality versus job quantity and Kranton and Thomas on meaning, identity, and employee mental health Wang et al., 2022Kranton & Thomas, 2025.

Treatment, stigma, and interventions. Use Lacey and coauthors for destigmatization and help-seeking, and Lund and coauthors for mental-health interventions and labor-market outcomes Lacey et al., 2022Lund et al., 2024.

Frontier methods. Use Jolivet, Postel-Vinay, and Robin for structural mental-health and labor-market links, Garcia-Gomez, Jones, and Rice for administrative health-shock thinking, and job-loss evidence as a cautionary example of work affecting mental health Jolivet et al., 2025García-Gómez et al., 2022Gallo et al., 2000Gallo et al., 2006Picciotto & others, 2024.

Exercises And Discussion Prompts¶

1. Why is presenteeism conceptually attractive but empirically difficult?

2. Suppose a workplace mental-health program improves retention but not wages. Is that a productivity success, a welfare success, both, or neither?

3. How would you separate “job quality harms mental health” from “workers with poor mental health sort into worse jobs”?

4. In what sense is stigma a labor-market institution rather than only a social attitude?

5. Design a treatment-access study. What labor outcome would you measure, and why might treatment selection still bias the estimate?

6. Choose one workplace measure: absences, performance, quits, supervisor ratings, or accommodations. Which mental-health mechanisms does it miss?

7. What would a welfare analysis need beyond wages, employment, and productivity?

Reproducibility And Code Lab Note¶

The Week 4 code lab lives at labs/04-mental-health-stress-workplace-productivity-and-worker-welfare/. It is a bounded synthetic teaching path, not an official replication of Bubonya, Cobb-Clark, and Wooden or Kranton and Thomas. The smoke path creates deterministic worker-level data; estimates a compact mental-health and productivity profile; diagnoses measurement, stigma, job-quality, treatment-selection, and reverse-causality threats; and transfers the design to treatment access, waiting time, job loss, pandemic and workplace shocks, disclosure interventions, and structural job-search settings.

The lab is conservative by design. It does not claim access to official household panel data, employer performance records, claims data, treatment records, or employee-assistance data.

Slide Companion Note¶

The Week 4 slide deck lives at slides/week4/04-mental-health-stress-workplace-productivity-and-worker-welfare.tex. The deck mirrors the chapter logic without duplicating the prose: it opens with mental health as a labor-market force, distinguishes productivity and welfare, covers stigma and treatment, shows how work conditions shape mental health, summarizes measurement and causal challenges, and closes with frontier research opportunities and the Research Lab design.

Bridge Forward¶

Week 5 broadens from worker-level and workplace-level mental-health mechanisms to disease exposure, environmental health, demographic change, and labor-market adjustment. The bridge is scale. Week 4 asks how mental health and stress change productivity, job quality, disclosure, treatment, retention, and welfare inside jobs. Week 5 asks how population-health and environmental shocks change workers, firms, local labor markets, and aggregate adjustment.