New and Old Jobs
Firms' GitHub Copilot adoption and labor market outcomes for software engineers
Matthew Baird et al.
Contemporary Economic Policy, forthcoming
Abstract:
Using LinkedIn and GitHub data, this paper examines how firms' adoption of GitHub Copilot (GHC), a generative AI coding assistant, relates to software engineer (SWE) skills and labor outcomes. GHC adoption is associated with around a 3%–5% higher monthly probability of hiring SWEs, driven by entry-level hires. New hires exhibit around 5% more non-programming skills, with no decrease in coding skills. These findings are consistent with, for SWEs, GAI's productivity impacts and creation of new tasks outweighing potential displacement effects from automation of some SWE tasks.
The Jevons Paradox and Insatiable Humans: Why AI Won't Empty the Finance Suite
Eldar Maksymov
Arizona State University Working Paper, May 2026
Abstract:
The conversation around AI and white-collar work has fixated on the wrong question. Anthropic's March 2026 finding that AI can theoretically perform 94.3 percent of business and finance tasks has executives debating which jobs will survive. They should be asking which jobs are about to come into existence. The Jevons Paradox -- William Stanley Jevons's 1865 observation that efficiency gains expand rather than contract resource use -- provides the framework. Its cleanest modern test: U.S. accountants quadrupled between 1980 and 2022, growing at nearly seven times the rate of population growth, after spreadsheets automated their core work. AI is to today's accountant what VisiCalc was to 1980's -- except more powerful. The near-term displacement is real and painful. But firms that treat AI only as a headcount-reduction tool will miss the expansion. This article maps that expansion and offers concrete prescriptions for students, executives, and educators.
Does Employment Slow Cognitive Decline? Evidence from Labor Market Shocks
Noah Arman Kouchekinia, David Neumark & Tim Bruckner
NBER Working Paper, April 2026
Abstract:
With large gains in life expectancy, the population share of disability due to cognitive decline and dementia has substantially increased. Many older adults in the United States leave the workforce well before age 65. Correlational evidence suggests that leaving the workforce before retirement age could accelerate the pace of cognitive decline. We offer causal evidence, using HRS data for the United States, exploiting plausibly exogenous shifts in labor demand in local labor markets as a Bartik instrument for employment variation across these markets. We find substantial declines over time in cognitive scores stemming from negative labor demand shocks. These findings are concentrated among men aged 51 to 64, whose employment decisions and outcomes may be more sensitive to local labor market conditions than are these decisions or outcomes for women or for older men. Our evidence extends past work focusing narrowly on the retirement age window and provides further support to the notion that working to older ages may delay age-related cognitive decline.
State Right to Work Laws and Economic Dynamism in U.S. Counties, 1946 to 2019
Alec Rhodes & Tom VanHeuvelen
American Sociological Review, forthcoming
Abstract:
Do state Right to Work (RTW) laws unleash economic dynamism, or the ability for local economies to respond, thrive, and grow in changing conditions? Although the specific goals of RTW laws to limit union security agreements appear to narrowly target unionized firms, proponents and opponents alike argue that RTW laws have broad labor market consequences. Union monopoly theories suggest that unions increase labor costs and exert greater worker control over the labor process in ways that distort and dampen firm investment and growth, and hence that RTW promotes economic dynamism. Institutionalist theories counter that unions increase labor productivity and build a stronger local consumer base, and hence that RTW inhibits economic dynamism. We provide a novel test of these divergent hypotheses using 75 years of County Business Patterns data and county-border-pair fixed-effects regression models to address unobserved heterogeneity. We fail to find consistent evidence that RTW passage is associated with meaningful changes in employment or workplace establishment concentration relative to geographically proximate counties in non-RTW states that share a common border. We develop an alternative competitive labor policy mitigation perspective that highlights how policymakers respond to policies in neighboring states to help explain this null result. Consistent with our arguments, we find that non-RTW states made tax and incentive policy more attractive for employers during this period, and that tax and incentive policies have a meaningful association with local economic dynamism. This highlights tax incentives as an alternative policy lever that non-RTW states used to mitigate the competitive advantages of RTW states.
How do holistic wrap-around anti-poverty programs affect employment and individualized outcomes?
Javier Espinosa et al.
Journal of Public Economics, May 2026
Abstract:
A new wave of social service programs aims to build a pathway out of poverty by helping clients define their own goals and then supporting them flexibly and intensively over multiple years to meet those goals. We conduct a randomized controlled trial of one such program, Bridges to Success. Two cohorts of participants were randomly assigned to intensive, holistic, wrap-around services that typically last two years versus a control offered help with an immediate need. Since the intervention has a clear goal of exiting poverty but is also holistic, we pre-specified both employment and non-employment outcomes. The measured treatment effect on employment three years after random assignment is 9 +/- 9 percentage points (pp). The proportion of people reporting high housing quality after one year has a treatment effect of −2 +/- 12 pp.
In exploratory analysis, additional evidence suggests a stronger case for effects on employment than non-employment outcomes. Employment after one year shows a treatment effect of 10 +/- 8 pp. Pooling our data with the most similar existing study increases precision relative to either study alone and indicates that such programs likely generate moderately positive employment effects. On the other hand, we find little evidence that intensive, holistic services affect any of a wide variety of other non-employment outcomes beyond housing, even when other areas of life are participants’ primary goals.
Innovation and Employment Cyclicality: Evidence from U.S. States
Nune Hovhannisyan & Jeremy Schwartz
Eastern Economic Journal, April 2026, Pages 425-443
Abstract:
Although there is agreement on the positive role of innovation in the economy in the long run, there is little evidence to show whether its impact on employment varies during the business cycle. In this paper, we use patent and R&D data as a measure of innovation and exploit US state variation to fill this gap in the literature. We find that, measured by patents or R&D expenditures, greater innovation amplifies the variation of employment over the business cycle. In other words, highly innovative states see larger employment gains during expansions, but also higher employment losses during downturns.
Intermediate Input Prices and the Labor Share
Juanma Castro-Vincenzi & Benny Kleinman
NBER Working Paper, April 2026
Abstract:
We argue that the relative price of materials is an important determinant of the labor share of income. When materials and primary inputs are complements and the profit share is positive, a higher price of materials lowers the labor share and raises the profit share of value added, without requiring markups or returns to scale to change. We show that materials-price fluctuations align with U.S. labor-share trends, provide causal evidence on this mechanism across industries and commuting zones, and quantify its importance in a dynamic quantitative model. Finally, we use our mechanism to rationalize differential labor-share trends across countries.
Do re-employment bonuses increase employment? Evidence from the Idaho Return to Work Bonus programme
Duncan Hobbs & Michael Strain
Economica, forthcoming
Abstract:
In June 2020, Idaho announced the Return to Work Bonus (RWB) programme, which provided residents who returned to work with bonuses of up to $1500. We present difference-in-differences, triple differences, event studies and synthetic control estimates suggesting that the programme may have supported employment and workforce participation following its enactment. For example, from difference-in-differences estimates, we find that the employment–population ratio rose by 3.6 percentage points, the unemployment rate fell by 0.7 percentage points, and the non-participation rate fell by 2.9 percentage points in Idaho relative to other states following the introduction of the RWB programme. If these results from Idaho were to generalize to the economy as a whole, then they would not be enough to arrest a moderate recession, but they would meaningfully accelerate labour market recovery. To the best of our knowledge, this is the first paper to study the effects of re-employment bonuses on the US labour market outside an experimental setting.
Marijuana Legalization and Firms’ Cost of Equity
Scott Guernsey, Matthew Serfling & Cheng Yan
Journal of Financial and Quantitative Analysis, May 2026, Pages 1112-1147
Abstract:
After medical marijuana legalization (MML) by U.S. states, firms’ cost of equity (COE) decreases, especially for those with more growth opportunities, higher productivity, or a more skilled workforce. This policy change also reduces firm risk and leads to an increase in labor supply through increased labor force participation, employment, hours worked, and net migration. Further, home prices rise after MML, reflecting increased local housing demand due to a growing supply of workers. These findings align with theoretical models that link asset prices to labor markets and suggest that MML can lower firms’ COE by mitigating labor search frictions.
Employment Under Marijuana
Wei-Fong Pan
Journal of Financial and Quantitative Analysis, forthcoming
Abstract:
This study examines the impact of recreational marijuana laws (RMLs) on firm-level employment using an imputation-based difference-in-differences approach across U.S. states. RMLs significantly reduce employment, particularly among firms with high-skilled labour, strong union presence, permissive corporate cultures, and in states with greater dispensary density. Alternative explanations -- including economic crises, COVID-19, fiscal changes, labour regulations, and related policies such as smoking bans and right-to-work laws -- are systematically ruled out through a series of placebo and robustness tests. RMLs also reduce investment, sales growth, and innovation, suggesting that legalisation introduces labour-related frictions with broad implications for firm performance and long-term dynamism.
Outplaying elite table tennis players with an autonomous robot
Peter Dürr et al.
Nature, 23 April 2026, Pages 886-891
Abstract:
Artificial intelligence (AI) systems now challenge or surpass human experts in many computer games. Physical and real-time sports such as table tennis, however, remain a major open challenge because of their requirements for fast, precise and adversarial interactions near obstacles and at the edge of human reaction time. Here we present Ace, to our knowledge the first real-world autonomous system competitive with elite human table tennis players. Ace addresses the challenges of physical real-time interaction through a new, high-speed perception system using event-based vision sensors, and a new control system based on model-free reinforcement learning, as well as state-of-the-art high-speed robot hardware. Evaluated in matches against elite and professional players under official competition rules, Ace achieved several victories and demonstrated consistent returns of high-speed, high-spin shots. These results highlight the potential of physical AI agents to perform complex, real-time interactive tasks, suggesting broader applications in domains requiring fast, precise human–robot interaction.