Real Work
Technology and Labor Markets: Past, Present, and Future; Evidence from Two Centuries of Innovation
Huben Liu et al.
NBER Working Paper, October 2025
Abstract:
We use recent advances in natural language processing and large language models to construct novel measures of technology exposure for workers that span almost two centuries. Combining our measures with Census data on occupation employment, we show that technological progress over the 20th century has led to economically meaningful shifts in labor demand across occupations: it has consistently increased demand for occupations with higher education requirements, occupations that pay higher wages, and occupations with a greater fraction of female workers. Using these insights and a calibrated model, we then explore different scenarios for how advances in artificial intelligence (AI) are likely to impact employment trends in the medium run. The model predicts a reversal of past trends, with AI favoring occupations that are lower-educated, lower-paid, and more male-dominated.
Future Shock or Future Shrug? Public Responses to Varied Artificial Intelligence Development Timelines
Anil Menon & Baobao Zhang
Journal of Politics, forthcoming
Abstract:
The increased pace of generative artificial intelligence (AI) development has prompted many experts to predict widespread workplace automation in the near future. We examine how messaging about different AI development timelines affects the public’s views about work-place automation and policies to mitigate technological unemployment. Our survey experiment (N=2,440 U.S. adults) randomly assigned participants to a control group or one of three treatments, each presenting a different timeline for job loss due to “transformative AI” (2026, 2030, or 2060). Generally, shorter timelines did not increase concerns about job loss or support for universal basic income, limiting automation, or funding job retraining. The informational treatments, when pooled, increased concerns regarding technological unemployment but did not impact policy preferences. Our null results speak to a larger literature, which finds that making individuals feel more proximate to risks does not necessarily increase support for policies aimed at mitigating these risks.
Careers of Minimum Wage Workers
Sari Pekkala Kerr, William Kerr & Louis Maiden
NBER Working Paper, October 2025
Abstract:
We characterize the careers of minimum wage workers by merging SIPP panels covering 1992-2016 into the LEHD. A long-run analysis shows strong earnings growth for these workers in subsequent decades, becoming indistinguishable from peers earning modestly more initially. Most of this growth is due to the steep earnings trajectories of young workers. Older workers earning minimum wages show a modest dip in earnings at that moment compared to earlier and later periods. Increases in state minimum wages do not significantly alter the future careers of workers who are on the minimum wage when the increases occur.
New employer payroll taxes and entrepreneurship
Audrey Guo & Melanie Wallskog
Journal of Public Economics, October 2025
Abstract:
How costly are taxes for young firms? In this paper, we demonstrate that even small payroll taxes significantly distort hiring decisions and employment growth. First, we leverage cross-sectional variation in the taxes faced by new employers to study how these taxes affect entrepreneurs’ decisions to become employers. We find that higher taxes discourage new firms from hiring their first workers; we estimate an elasticity of the number of new employers to taxes of −0.1. Second, we study tax changes a new employer faces after it enters. We find that higher taxes lead more firms to exit, while also reducing employment for those who survive and leading some firms to avoid taxes by using non-taxable contract labor.
Better Labor Market Options Reduce Workplace Injuries
Jisung Park, Paul Stainier & Anna Stansbury
University of Pennsylvania Working Paper, October 2025
Abstract:
When workers' labor market outside options improve, their wages go up. What about other non-wage amenities? Using a Bartik shift-share IV that isolates shocks to workers' outside options in the labor market, we find that better labor market options meaningfully reduce the workplace injury rate: a one-standard-deviation improvement in outside options lowers injuries by about 0.18 standard deviations. These results suggest that external labor market conditions influence incentives to maintain safe workplaces, which has important implications for designing public policy, including those that regulate workplace health and safety and alter labor market competition.
The Economic Incidence of Schedule Unpredictability in Hourly Work
Hannah Farkas
Columbia University Working Paper, November 2025
Abstract:
Workers in hourly service jobs frequently experience shift cancellations, schedule adjustments, and unpredictable hours, often referred to as ‘just-in-time’ scheduling. Despite imposing significant costs on workers, data limitations have made studying such patterns challenging. I use a large administrative dataset featuring information on nearly 1 million employees’ scheduled and worked hours to illustrate patterns of schedule unpredictability at thousands of small food and drink and retail businesses across the U.S. I demonstrate that baseline schedule unpredictability is widespread, the most severe for the lowest-wage and least-tenured workers, and that a machine learning model trained on past schedules and deviations cannot capture all unpredictability. Using exogenous weather shocks, which are known to diminish consumer demand, I illustrate how these customer-facing establishments pass risk of slow business days onto workers through the channel of unpredictable scheduling. I then leverage the fact that most hourly service workers earn at or near the minimum wage to demonstrate the tradeoff between schedule predictability and wages. Following large exogenous minimum wage increases, schedule unpredictability increases by 20% per week and schedules become even more responsive to weather shocks. This highlights how some of the welfare gains workers realize from a minimum wage may be offset by increased schedule unpredictability.
Robot adoption and inflation dynamics
Henrique Basso & Omar Rachedi
Journal of Monetary Economics, forthcoming
Abstract:
Leveraging variation in robot adoption across U.S. metropolitan areas, we document that automation reduces the sensitivity of inflation to unemployment. To rationalize this finding, we build a New Keynesian model with search frictions in the labor market where robot adoption flattens the Phillips curve. The key channel is the option value of automation: the threat of automating labor tasks alters effective workers’ bargaining power, muting the wage sensitivity to unemployment. We validate the relevance of this channel in the data by showing that robot adoption reduces the sensitivity of inflation to unemployment relatively more in highly unionized metropolitan areas.
Making Talk Cheap: Generative AI and Labor Market Signaling
Anaïs Galdin & Jesse Silbert
Princeton Working Paper, November 2025
Abstract:
Large language models (LLMs) like ChatGPT have significantly lowered the cost of producing written content. This paper studies how LLMs, through lowering writing costs, disrupt markets that traditionally relied on writing as a costly signal of quality (e.g., job applications, college essays). Using data from Freelancer.com, a major digital labor platform, we explore the effects of LLMs’ disruption of labor market signaling on equilibrium market outcomes. We develop a novel LLM-based measure to quantify the extent to which an application is tailored to a given job posting. Taking the measure to the data, we find that employers had a high willingness to pay for workers with more customized applications in the period before LLMs were introduced, but not after. To isolate and quantify the effect of LLMs’ disruption of signaling on equilibrium outcomes, we develop and estimate a structural model of labor market signaling, in which workers invest costly effort to produce noisy signals that predict their ability in equilibrium. We use the estimated model to simulate a counterfactual equilibrium in which LLMs render written applications useless in signaling workers’ ability. Without costly signaling, employers are less able to identify high-ability workers, causing the market to become significantly less meritocratic: compared to the pre-LLM equilibrium, workers in the top quintile of the ability distribution are hired 19% less often, workers in the bottom quintile are hired 14% more often.
Contrasting the Local and National Demographic Incidence of Local Labour Demand Shocks
Richard Mansfield
Economic Journal, forthcoming
Abstract:
This paper examines how spatial frictions that differ among heterogeneous workers and establishments shape the geographic and demographic incidence of alternative local labour demand shocks, with implications for the appropriate level of government at which to fund local economic initiatives. LEHD data featuring millions of job transitions facilitate estimation of a rich two-sided labour market assignment model. The model generates simulated forecasts of many alternative local demand shocks featuring different establishment compositions and local areas. Workers within 10 miles receive only 11.2% (6.6%) of nationwide welfare (employment) short-run gains, with at least 35.9% (62.0%) accruing to out-of-state workers, despite much larger per-worker impacts for the closest workers. Local incidence by demographic category is very sensitive to shock composition, but different shocks produce similar demographic incidence further from the shock. Furthermore, the remaining heterogeneity in incidence at the state or national level can reverse patterns of heterogeneous demographic impacts at the local level. Overall, the results suggest that reduced-form approaches using distant locations as controls can produce accurate estimates of local shock impacts on local workers, but that the distribution of local impacts badly approximates shocks’ statewide or national incidence.
How local is local development? Evidence from casinos
Ari Anisfeld & Jordan Rosenthal-Kay
Regional Science and Urban Economics, November 2025
Abstract:
One rationale for place-based policy is that local development produces positive productivity spillovers. We examine the employment spillovers from a large local development project: opening a casino. Comparing employment in neighborhoods that won a casino license to runner-up neighborhoods that narrowly lost, we find that casinos create jobs in their immediate vicinity. However, we estimate net job losses overall when considering the broader neighborhood. Employment gains concentrate in the leisure and hospitality industry, suggesting spillovers are industry-specific or are driven by demand-side forces like trip-chaining. We develop theory to show that our estimates imply a rapid spatial decay of productivity spillovers.