Disease Exposure, Environmental Health, Demographic Change, And Labor-Market Adjustment

Learning Objectives¶

By the end of Week 5, students should be able to:

1. explain why disease, environmental exposure, and demographic change are labor-market shocks rather than background population facts;

2. distinguish worker-capacity, firm-production, migration/reallocation, distributional, and policy-institution channels;

3. compare plague, tuberculosis and early public health, HIV/AIDS, and COVID as labor-market incidence cases with different timing, policy response, and persistence;

4. analyze health inequality as unequal exposure plus unequal capacity to adjust through remote work, job switching, migration, insurance, or retirement;

5. connect aging, fertility decline, dependency structure, and demographic composition to wages, labor shortages, technology adoption, migration, and late-career work;

6. identify the main data sources and empirical designs used to study disease, environmental health, demography, and labor-market adjustment;

7. design a research project that names the shock, labor margin, unit of analysis, adjustment horizon, identifying variation, and welfare object.

The central question is this: how do disease, environmental exposure, and demographic structure reshape workers, firms, local labor markets, and aggregate adjustment?

Opening Orientation¶

Weeks 1 to 4 studied health as work capacity, disability, lifecycle allocation, and workplace mental-health constraints. Week 5 scales up. Infectious disease, pollution, heat, mortality shocks, fertility decline, and population aging can move entire workplaces, neighborhoods, local labor markets, sectors, and migration systems. The labor-economics question is not only whether exposure reduces work. It is who is exposed, who can adjust, whether the shock appears in wages or in hidden productivity and welfare losses, and whether the adjustment is temporary disruption, medium-run reallocation, or long-run institutional and distributional change.

This week is therefore not a generic epidemiology, climate, or demography lecture. The objects are labor supply, labor demand, productivity, occupational sorting, migration, local adjustment, technology adoption, retirement, worker welfare, and inequality. Disease and environment matter because they change the feasible set for work and the real value of jobs. Demographic change matters because it changes the composition of labor supply, the demand for care, the incentives facing firms, and the margins through which places adjust.

Bridge¶

Week 4 treated mental health and stress as workplace forces affecting attendance, productivity, retention, disclosure, treatment, supervision, and welfare. Week 5 keeps the same labor discipline but changes the scale. Instead of asking how one worker or workplace adjusts to a health constraint, we ask how exposure and population structure reshape groups of workers, firms, places, and cohorts.

The bridge is incidence. In Week 4, stigma and job quality determined who bore invisible mental-health costs inside work. In Week 5, exposure, mobility, remote-work feasibility, housing, occupation, age, and migration status determine who bears disease and environmental costs across labor markets. Wages and employment remain important, but they are not enough. A shock can lower output per hour, raise fatigue, reduce job quality, move workers across places, induce automation, delay retirement, or change the welfare value of the same nominal wage.

Field Core¶

A Labor-Market Framework For Scaled Health Shocks¶

A useful starting point is to write labor-market adjustment as a system. Let local labor outcomes depend on worker health capacity, environmental and disease exposure, demographic composition, firm labor demand, migration opportunities, and policy response:

where $Y_{rgt}$ is a labor outcome for group $g$ in region $r$ at time $t$; $H_{rgt}$ is health capacity or mortality risk; $E_{rt}$ is disease or environmental exposure; $D_{rt}$ is demographic structure; $F_{rt}$ is firm labor demand and production technology; $M_{rt}$ is migration or commuting adjustment; and $P_{rt}$ is public-health, labor-market, or social-insurance policy.

The same observed employment decline can therefore mean very different things. It may reflect illness, fear of exposure, school closure, demand collapse, shutdown policy, firm exit, sectoral reallocation, migration, early retirement, or data composition. A strong paper names the channel it can identify and the channels it cannot separate.

The main channels are:

• Worker capacity: exposure changes health, fatigue, mortality risk, cognition, physical performance, attendance, or retirement.

• Firm production: employers change scheduling, safety protocols, task assignment, hiring, technology, location, or exit.

• Migration and sorting: workers move, commute differently, switch occupations, work remotely, retire, or become immobile.

• Distributional incidence: exposure and adjustment differ by job, housing, wealth, race, gender, age, migrant status, and local public capacity.

• Policy and institutions: public health, shutdowns, unemployment insurance, paid leave, disability systems, labor regulation, pensions, and infrastructure mediate outcomes.

Persistence channels and adjustment margins¶

Interpretation: the same health or demographic shock can change labor markets through multiple margins at once. A good paper makes clear whether the relevant response is worker capacity, firm adaptation, migration/reallocation, or institutional/policy change.

Historical Health Episodes And Long-Run Labor Effects¶

Historical disease episodes are useful because they stretch the time horizon. They show how a health shock can begin as mortality or morbidity and become a labor-market event through wages, bargaining power, human capital, public-health investment, care burdens, migration, and institutions.

Plague and the Black Death. The Black Death is the canonical case for labor scarcity, wage pressure, and institutional change, but the lesson is not mechanical population decline. Jedwab, Johnson, and Koyama emphasize that the economic effects of plague depended on market structure, land-labor ratios, urban networks, and institutions Jedwab et al., 2022. A labor reading asks whether mortality raised the marginal product of surviving workers, whether bargaining power changed, whether institutions constrained wage adjustment, and whether long-run inequality shifted. Short-run disruption can dominate in some places, while medium-run scarcity and institutional adaptation matter in others.

Tuberculosis and early public health. Tuberculosis illustrates how disease surveillance and public-health capacity enter labor markets through survival, schooling, human capital, productivity, and local inequality. Anderson, Charles, Rees, and Wang study tuberculosis reporting laws and early urban mortality, showing that public-health institutions can alter disease environments rather than merely record them Anderson et al., 2019. The labor question is whether healthier cohorts acquire more human capital, work more productively, migrate differently, or face a different local disease amenity.

HIV/AIDS. HIV/AIDS connects mortality, morbidity, household labor allocation, skill premia, and development. Chicoine’s South African evidence shows how AIDS mortality affected wages and employment, making the shock a local labor-supply and distributional event rather than only a health crisis Chicoine, 2012. HIV/AIDS also forces students to keep caregiving, orphanhood, household income, skill composition, and migration in the labor framework.

COVID. COVID belongs in the same family, but with unusually fast data, strong policy response, sectoral concentration, remote-work feasibility, and school/care disruptions. Albanesi and Kim show highly unequal employment effects across occupations, gender, and family status during the COVID recession Albanesi & Kim, 2021. Chetty and coauthors show how real-time private-sector data can identify heterogeneous spending and employment responses at high frequency Chetty et al., 2020. COVID is not a historical exception to the labor-incidence framework. It is a modern case that makes policy, care, occupation, remote work, and distributional incidence visible.

The common timing distinction is essential:

• Short-run disruption: illness, fear, closures, absences, sector shutdowns, demand collapse, and care constraints.

• Medium-run reallocation: job switching, remote work, migration, firm reorganization, safety investment, and sectoral recovery.

• Long-run effects: human-capital scarring, public-health institutions, occupational restructuring, inequality, technology adoption, and changes in worker bargaining power.

COVID In Historical Perspective¶

COVID makes several older disease-shock questions newly measurable. Like plague, it raised questions about labor scarcity, bargaining power, and sectoral bottlenecks. Like tuberculosis, it made public-health capacity and reporting central. Like HIV/AIDS, it generated unequal morbidity, mortality, caregiving, and labor-market risks across groups. Unlike many older episodes, it arrived with high-frequency administrative, payroll, mobility, transaction, and remote-work data.

The interpretation problem is that COVID moved many margins at once. A decline in employment among service workers could reflect infection risk, customer demand, public restrictions, school closures, employer closure, transportation risk, or household care constraints. A rise in wages in some lower-wage sectors could reflect quits, search frictions, hazard compensation, sectoral demand, or temporary labor scarcity. A migration response could reflect health risk, remote-work opportunities, housing costs, or amenity demand. The empirical task is to state which margin the design can isolate and which margins remain bundled.

COVID also changed the methods frontier. It normalized real-time labor-market measurement, linked private-sector data, occupation-by-exposure designs, and rapid heterogeneity analysis. But high-frequency data do not solve incidence by themselves. They often measure transactions, payroll, mobility, or posted jobs rather than worker welfare, health, household constraints, or long-run scarring.

Environmental Health, Productivity, And Hidden Incidence¶

Pollution, heat, wildfire smoke, toxic exposure, and unsafe worksites can reduce labor supply and productivity even when they do not appear as large employment effects. Hanna and Oliva use a natural experiment in Mexico City to show that pollution can affect labor supply Hanna & Oliva, 2015. Graff Zivin and Neidell study pollution and worker productivity, while their temperature work highlights time allocation under heat exposure Graff Zivin & Neidell, 2012Graff Zivin & Neidell, 2014. Chang, Graff Zivin, Gross, and Neidell show that pollution reduces call-center productivity, which matters because environmental health is not only an outdoor manual-labor issue Chang et al., 2019. Somanathan and coauthors extend the heat-productivity logic to Indian manufacturing, where heat affects both productivity and labor supply Somanathan et al., 2021.

The labor point is that environmental exposure changes the real value of work. Workers may stay employed but experience fatigue, errors, absences, lower output, greater injury risk, or lower welfare. Firms may respond by changing shifts, investing in ventilation or cooling, redesigning tasks, relocating, or substituting capital. Places may adjust through migration, commuting, housing prices, or industrial composition. If only employment and wages are observed, much of the incidence is hidden.

Health Inequality, Migration, And Labor-Market Choice¶

Health inequality enters this week in two parts: unequal exposure and unequal adjustment capacity. Workers differ in whether they can avoid exposure through remote work, safer housing, private transportation, flexible scheduling, job switching, insurance, savings, migration, or retirement. Deaton’s health-inequality framework is useful because it pushes us to specify mechanisms rather than treat health gradients as descriptive background Deaton, 2003.

Migration and immobility are central. A household may leave a polluted area, but only if it has resources, information, legal status, portable work, family support, and destination access. A worker may switch out of risky work, but only if credentials, contracts, language, job networks, and local demand permit it. Bossavie and coauthors show how migrant workers faced elevated economic and health risks during COVID, connecting occupational sorting, legal and institutional vulnerability, and exposure in one labor-market object Bossavie et al., 2021. Development settings make these margins especially visible because informality, crowding, weak insurance, and limited public-health capacity can amplify both exposure and immobility.

Health inequality, migration, and labor-market incidence¶

Research implication: health inequality is not just a health outcome. It is a labor-market incidence and mobility problem.

The central labor questions are:

1. Who is more exposed because of occupation, housing, transport, race, wealth, migrant status, age, or local institutions?

2. Who can adjust through remote work, migration, commuting, job switching, reduced hours, retirement, or preventive investment?

3. Who bears the welfare loss when wages do not compensate for risk, fatigue, stress, or unsafe amenities?

4. Which firm and local responses amplify or dampen inequality?

Demographic Change, Aging, And Aggregate Adjustment¶

Population aging, fertility decline, mortality change, and dependency structure are labor-market forces because they change labor supply, care demand, retirement, skill composition, migration pressure, and firms’ incentives to reorganize production. Bloom, Canning, and Fink frame population aging as a macroeconomic and labor-market challenge, but this course keeps the labor margins visible: participation, hours, retirement, productivity, wages, care work, and job design Bloom et al., 2010. Lee and Mason connect fertility decline and human capital to demographic transition, which matters for the composition and quality of future labor supply Lee & Mason, 2010.

Aging can also induce firm response. Acemoglu and Restrepo show that demographic change can encourage automation, so aging affects labor markets through endogenous technology adoption as well as labor supply Acemoglu & Restrepo, 2022. Maestas, Mullen, and Powell show that population aging affects economic growth, labor force, and productivity, making local demographic composition a labor-market adjustment object Maestas et al., 2023. Regional evidence on demographic change highlights that adjustment can occur through wages, migration, job redesign, retirement behavior, and sectoral composition rather than a single aggregate employment effect Little & Triest, 2002Böhm, 2021.

Demographic change should be connected to both ends of the working life. Fertility decline alters the future age structure and the supply of young workers. Aging increases late-career work, retirement timing, elder-care demand, and the value of flexible jobs. Dependency structure changes fiscal pressure, pension incentives, and care-sector labor demand. These are labor-market margins, not demographic background.

Methods, Data, And Research Architecture¶

Week 5 has a heavier methods layer because the empirical objects are multi-scale and often spatial. The strongest papers name the shock, the labor margin, the timing, the unit of analysis, and the equilibrium response.

Historical methods and data sources¶

Good design principle: define whether the shock is health, environment, demography, or policy; then define whether the main response margin is work capacity, firm behavior, migration, or welfare.

Common design families include:

• Cohort-by-place exposure designs for historical epidemics, fetal health, early public health, and long-run human-capital or earnings outcomes Almond, 2006Bleakley, 2007.

• Event studies around onset or intensity shocks for COVID waves, local outbreaks, shutdowns, heat waves, pollution spikes, plant closures, and region-level mortality.

• Difference-in-differences and exposure-gradient designs using policy timing, reporting laws, environmental regulation, pollution monitors, weather, or disease intensity.

• Worker or firm panel designs linking payroll, attendance, output, claims, task assignment, and workplace safety to exposure.

• Spatial sorting and migration designs using IRS migration, ACS/Census, commuting flows, cell-phone mobility, housing markets, and local labor demand.

• Demographic composition designs exploiting variation in aging, fertility decline, cohort size, dependency ratios, and retirement or pension rules.

Common data sources include mortality and vital statistics, disease surveillance, hospital or claims records, administrative earnings, UI records, tax data, payroll records, establishment data, weather and pollution monitors, satellite exposure data, Census/ACS, historical census, public-health archives, migration and commuting data, mobility records, private-sector transactions, and occupational exposure measures.

The main threats are also predictable: survival selection, endogenous residence, avoidance behavior, local pre-trends, concurrent policies, demand shocks, migration selection, measurement breaks, changing diagnosis or reporting intensity, equilibrium spillovers, and bundled public-health responses. A credible paper does not need to solve every threat, but it must say which object is identified.

Research Lab¶

The Week 5 research lab follows Reproduce -> Diagnose -> Transfer. The primary anchor is Albanesi and Kim on the COVID recession and unequal labor-market incidence because it gives students a bounded modern health-shock design with worker-type and occupation heterogeneity Albanesi & Kim, 2021. The challenge anchor is Chicoine on HIV/AIDS and labor markets in South Africa because it pushes the design toward mortality, development, skill premia, and longer-run distributional adjustment Chicoine, 2012.

The lab is not an official replication package for either paper. It uses deterministic synthetic data to teach a conservative labor-incidence workflow: disease risk, policy intensity, occupation, remote-work feasibility, care burden, migrant status, environmental exposure, and demographic structure jointly shape employment, hours, wages, migration, automation, and welfare.

Reproduce. Students recover a compact COVID-style incidence profile by occupation and worker type. The synthetic profile compares employment, hours, wage, remote-work feasibility, exposure risk, care burden, migration response, automation adoption, and welfare loss before, during, and after the shock.

Diagnose. Students classify each pattern as direct disease risk, policy/demand collapse, caregiving constraint, occupational sorting, migration selection, firm reorganization, environmental exposure, demographic pressure, or hidden welfare incidence. They explicitly separate short-run disruption from medium-run reallocation.

Transfer. Students redesign the same workflow for one non-COVID setting: Black Death labor scarcity, tuberculosis reporting laws, HIV/AIDS mortality, pollution and productivity, heat and manufacturing, migrant exposure, or aging and automation. The transfer task names the shock, data, comparison group, labor outcome, main threat, and welfare margin.

Methods Box¶

Reading Ladder And References¶

Core framing. Start with Currie and Madrian for health and labor markets, Deaton for health inequality, and Adda for how economic activity and viral disease interact Currie & Madrian, 1999Deaton, 2003Adda, 2016.

Historical and long-run episodes. Use Jedwab, Johnson, and Koyama for the Black Death, Anderson and coauthors for tuberculosis reporting laws, Almond for the 1918 influenza cohort design, Bleakley for hookworm eradication, and Chicoine for HIV/AIDS and South African labor markets Jedwab et al., 2022Anderson et al., 2019Almond, 2006Bleakley, 2007Chicoine, 2012.

COVID and modern epidemic labor markets. Pair Albanesi and Kim with Chetty and coauthors to connect worker-type heterogeneity, sectoral incidence, real-time data, and stabilization policy Albanesi & Kim, 2021Chetty et al., 2020.

Environmental health and work. Read Hanna and Oliva on pollution and labor supply, Graff Zivin and Neidell on pollution and productivity, Chang and coauthors on call-center productivity, Graff Zivin and Neidell on heat and time allocation, and Somanathan and coauthors on temperature, productivity, and labor supply in Indian manufacturing Hanna & Oliva, 2015Graff Zivin & Neidell, 2012Chang et al., 2019Graff Zivin & Neidell, 2014Somanathan et al., 2021.

Demography and aggregate adjustment. Use Bloom, Canning, and Fink on population aging, Lee and Mason on fertility and human capital, Acemoglu and Restrepo on demographics and automation, Maestas, Mullen, and Powell on aging and productivity, and regional demographic-change work as comparative background Bloom et al., 2010Lee & Mason, 2010Acemoglu & Restrepo, 2022Maestas et al., 2023Little & Triest, 2002Böhm, 2021.

Migration, inequality, and vulnerability. Use Bossavie and coauthors on migrant workers during COVID as a bridge between disease exposure, occupational sorting, institutional vulnerability, and labor-market risk Bossavie et al., 2021.

Frontier and reading map¶

Core framing¶

• Currie and Madrian (1999) Currie & Madrian, 1999

• Deaton (2003) Deaton, 2003

Historical episodes¶

• Jedwab, Johnson, and Koyama (2022) Jedwab et al., 2022

• Anderson et al. (2019) Anderson et al., 2019

• Chicoine (2012) Chicoine, 2012

COVID and modern epidemic labor markets¶

• Albanesi and Kim (2021) Albanesi & Kim, 2021

• Chetty et al. (2020) Chetty et al., 2020

Environmental health and work¶

• Hanna and Oliva (2015) Hanna & Oliva, 2015

• Graff Zivin and Neidell (2014) Graff Zivin & Neidell, 2014

• Chang et al. (2019) Chang et al., 2019

Demography and aggregate adjustment¶

• Acemoglu and Restrepo (2022) Acemoglu & Restrepo, 2022

• Böhm (2021) Böhm, 2021

Inequality, migration, and vulnerability¶

• Bossavie et al. (2021) Bossavie et al., 2021

Frontier project ideas¶

• Health shocks and migration traps

• Pandemic labor-market persistence

• Aging, automation, and worker welfare

• Pollution, productivity, and local reallocation

• Health inequality as a labor-market incidence problem

Exercises And Discussion Prompts¶

1. Why can the same health shock appear as an employment shock in one setting, a wage shock in another, and an automation shock in a third?

2. Choose plague, tuberculosis, HIV/AIDS, or COVID. What is the short-run disruption margin, the medium-run reallocation margin, and the possible long-run institutional or distributional margin?

3. How would you distinguish worker-capacity effects from firm-reorganization effects in pandemic, pollution, or heat data?

4. Which workers are least able to adjust to disease or environmental exposure, and what labor-market frictions make them immobile?

5. When should aging be modeled as a labor-supply shock, and when should it be modeled as a technology-adoption shock?

6. Design a migration study around environmental health. What comparison group would you use, and how would you handle selection into moving?

7. Pick one outcome that wages and employment would miss. How would you measure it?

Reproducibility And Code Lab Note¶

The Week 5 code lab lives at labs/05-disease-exposure-environmental-health-demographic-change-and-labor-market-adjustment/. It is a bounded synthetic teaching path, not an official replication of Albanesi and Kim or Chicoine. The smoke path creates deterministic local labor-market data; estimates a compact disease/environment/demography incidence profile; diagnoses health risk, policy/demand collapse, care constraints, occupational sorting, migration, firm response, demographic pressure, and hidden welfare incidence; and transfers the design to historical epidemics, HIV/AIDS, pollution, heat, migration vulnerability, and aging/automation.

The lab is conservative by design. It does not claim access to official Current Population Survey extracts, private-sector real-time data, South African census or mortality records, public-health archives, payroll data, pollution monitor microdata, migration records, or proprietary firm output data.

Slide Companion Note¶

The Week 5 slide deck lives at slides/week5/05-disease-exposure-environmental-health-demographic-change-and-labor-market-adjustment.tex. The deck mirrors the chapter logic without duplicating the prose: it opens with disease, environment, and demography as labor-market shocks; compares historical health episodes; places COVID in historical perspective; centers health inequality, migration, and labor adjustment; covers aging and demographic change; summarizes methods and data; and closes with the Research Lab design.

Bridge Forward¶

Week 6 turns the whole course into research design. Week 5 supplies the scaled-up objects: disease, environmental exposure, demographic structure, migration, firm response, local adjustment, and welfare. The final week asks students to choose a labor outcome, define the health or population mechanism, state the counterfactual, and build a credible design without losing sight of inequality and worker welfare.