Human Work
Population Growth and Firm Dynamics
Michael Peters & Conor Walsh
Journal of Political Economy Macroeconomics, forthcoming
Abstract:
Firm-based theories of creative destruction in the spirit of Klette and Kortum (2004) are the workhorse models linking firm dynamics and growth. Because they typically assume a constant population, they are ill-equipped to study the implications of falling population growth. We propose a framework that overcomes this shortcoming. Our theory implies that neither the growth rate nor the size distribution are subject to scale effects, and it generates a thick-tailed employment distribution. Lower population growth reduces entry, increases concentration, raises market power, and lowers growth. These predictions are consistent with the experience of the US and many developed economies.
Task-Specific Technical Change and Comparative Advantage
Lukas Althoff & Hugo Reichardt
Stanford Working Paper, January 2026
Abstract:
Artificial intelligence is changing which tasks workers do and how they do them. Predicting its labor market consequences requires understanding how technical change affects workers’ productivity across tasks, how workers adapt by changing occupations and acquiring new skills, and how wages adjust in general equilibrium. We introduce a dynamic task-based model in which workers accumulate multidimensional skills that shape their comparative advantage and, in turn, their occupational choices. We then develop an estimation strategy that recovers (i) the mapping from skills to task-specific productivity, (ii) the law of motion for skill accumulation, and (iii) the determinants of occupational choice. We use the quantified model to study generative AI’s impact via augmentation, automation, and a third and new channel -- simplification -- which captures how technologies change the skills needed to perform tasks. Our key finding is that AI substantially reduces wage inequality while raising average wages by 21 percent. AI’s equalizing effect is fully driven by simplification, enabling workers across skill levels to compete for the same jobs. We show that the model’s predictions line up with recent labor market data.
O-Ring Automation
Joshua Gans & Avi Goldfarb
NBER Working Paper, January 2025
Abstract:
We study automation when tasks are quality complements rather than separable. Production requires numerous tasks whose qualities multiply as in an O-ring technology. A worker allocates a fixed endowment of time across the tasks performed; machines can replace tasks with given quality, and time is allocated across the remaining manual tasks. This “focus” mechanism generates three results. First, task-by-task substitution logic is incomplete because automating one task changes the return to automating others. Second, automation decisions are discrete and can require bundled adoption even when automation quality improves smoothly. Third, labour income can rise under partial automation because automation scales the value of remaining bottleneck tasks. These results imply that widely-used exposure indices, which aggregate task-level automation risk using linear formulas, will overstate displacement when tasks are complements. The relevant object is not average task exposure but the structure of bottlenecks and how automation reshapes worker time around them.
Artificial intelligence tools expand scientists’ impact but contract science’s focus
Qianyue Hao et al.
Nature, forthcoming
Abstract:
Developments in artificial intelligence (AI) have accelerated scientific discovery. Alongside recent AI-oriented Nobel prizes, these trends establish the role of AI tools in science. This advancement raises questions about the influence of AI tools on scientists and science as a whole, and highlights a potential conflict between individual and collective benefits. To evaluate these questions, we used a pretrained language model to identify AI-augmented research, with an F1-score of 0.875 in validation against expert-labelled data. Using a dataset of 41.3 million research papers across the natural sciences and covering distinct eras of AI, here we show an accelerated adoption of AI tools among scientists and consistent professional advantages associated with AI usage, but a collective narrowing of scientific focus. Scientists who engage in AI-augmented research publish 3.02 times more papers, receive 4.84 times more citations and become research project leaders 1.37 years earlier than those who do not. By contrast, AI adoption shrinks the collective volume of scientific topics studied by 4.63% and decreases scientists’ engagement with one another by 22%. By consequence, adoption of AI in science presents what seems to be a paradox: an expansion of individual scientists’ impact but a contraction in collective science’s reach, as AI-augmented work moves collectively towards areas richest in data. With reduced follow-on engagement, AI tools seem to automate established fields rather than explore new ones, highlighting a tension between personal advancement and collective scientific progress.
Capital composition and the decline of the labor share: Why buildings matter
Jacob Kerspien, Jakob Madsen & Holger Strulik
European Economic Review, April 2026
Abstract:
This paper argues that the decline in the labor share is not driven by the overall quantity of capital, but by its changing composition. Constructing annual macro data for 16 advanced countries over two centuries, we show that, since 1980, the relative decline in buildings capital and the associated increase in real prices of buildings have reduced the labor share because buildings and labor are complements. The decline in the labor share has been reinforced by the increase in machinery capital and the associated decline of real prices of machinery capital because machinery capital and labor are substitutes. Together, these shifts in capital composition account for a substantial portion of the observed decline in the labor share of income.
How Adaptable Are American Workers to AI-Induced Job Displacement?
Sam Manning & Tomás Aguirre
NBER Working Paper, January 2026
Abstract:
We construct an occupation-level adaptive capacity index that measures a set of worker characteristics relevant for navigating job transitions if displaced, covering 356 occupations that represent 95.9% of the U.S. workforce. We find that AI exposure and adaptive capacity are positively correlated: many occupations highly exposed to AI contain workers with relatively strong means to manage a job transition. Of the 37.1 million workers in the top quartile of AI exposure, 26.5 million are in occupations that also have above-median adaptive capacity, leaving them comparatively well-equipped to handle job transitions if displacement occurs. At the same time, 6.1 million workers (4.2% of the workforce in our sample) work in occupations that are both highly exposed and where workers have low expected adaptive capacity. These workers are concentrated in clerical and administrative roles. Importantly, AI exposure reflects potential changes to work tasks, not inevitable displacement; only some of the changes brought on by AI will result in job loss. By distinguishing between highly exposed workers with relatively strong means to adjust and those with limited adaptive capacity, our analysis shows that exposure measures alone can obscure both areas of resilience to technological change and concentrated pockets of elevated vulnerability if displacement were to occur.
The Long-Lived Cyclicality of the Labor Force Participation Rate
Tomaz Cajner, John Coglianese & Joshua Montes
Review of Economics and Statistics, forthcoming
Abstract:
How cyclical is the U.S. labor force participation rate (LFPR)? We examine exogenous state-level business cycle shocks, finding that the LFPR is highly cyclical, but with significantly longer-lived responses than the unemployment rate. After a negative shock, the LFPR declines for about four years -- substantially lagging unemployment -- and only fully recovers after about eight years. Our main specifications use age-sex-adjusted LFPR, and we show that using unadjusted LFPR is problematic because local shocks spur changes in the population of high-LFPR age groups. Cyclicality varies across groups, with larger and longer-lived responses among men, younger workers, less-educated workers, and Black workers.
(Artificial) Intelligence Saturation and the Future of Work
Konrad Kording & Ioana Marinescu
University of Pennsylvania Working Paper, November 2025
Abstract:
Macroeconomic models typically treat AI as just another form of capital, and predict a slowly evolving world, while computer science scaling laws applied to the whole economy predict explosive growth and the potential for a singularity-like event. Both views gloss over the asymmetric reality that intelligence capital or AI scales at computer-science speeds, whereas physical capital and labor do not. What’s missing is a unified, parameter-driven framework that can nest assumptions from both economics and computer science to generate meaningful predictions of AI’s wage and output impacts. Here we use a constant elasticity of substitution (CES) production function framework that separates physical and intelligence sectors. Whereas physical capabilities let us affect the world, intelligence capabilities let us do this well: the two are complementary. Given complementarity between the two sectors, the marginal returns to intelligence saturate, no matter how fast AI scales. Because the price of AI capital is falling much faster than that of physical capital, intelligence tasks are automated first, pushing human labor toward the physical sector. The impact of automation on wages is theoretically ambiguous and can be non-monotonic in the degree of automation. A necessary condition for automation to decrease wages is that the share of employment in the intelligence sector decreases; this condition is not sufficient because automation can raise output enough to offset negative reallocation effects. In our baseline simulation, wages increase and then decrease with automation. Our interactive tool www.intelligencesaturation.org shows how parameter changes shift that trajectory. Wage decreases are steeper at high levels of automation when the outputs of the physical and intelligence sectors are more substitutable. After full automation, more AI and more physical capital increase wages, a classic prediction from standard production functions in capital and labor. Yet, when intelligence and physical are complementary, the marginal wage impact of AI capital saturates as AI grows large. More broadly, the model offers a structured way to map contrasting intuitions from economics and computer science into a shared parameter space, enabling clearer policy discussions, and guiding empirical work to identify which growth and wage trajectories are plausible.
Scaling Laws for Economic Productivity: Experimental Evidence in LLM-Assisted Consulting, Data Analyst, and Management Tasks
Ali Merali
Yale Working Paper, December 2025
Abstract:
This paper derives "Scaling Laws for Economic Impacts" -- empirical relationships between the training compute of Large Language Models (LLMs) and professional productivity. In a preregistered experiment, over 500 consultants, data analysts, and managers completed professional tasks using one of 13 LLMs. We find that each year of AI model progress reduced task time by 8%, with 56% of gains driven by increased compute and 44% by algorithmic progress. However, productivity gains were significantly larger for non-agentic analytical tasks compared to agentic workflows requiring tool use. These findings suggest continued model scaling could boost U.S. productivity by approximately 20% over the next decade.
Understanding the Employment Effects of Opportunity Zones
Matthew Freedman, Noah Arman Kouchekinia & David Neumark
NBER Working Paper, December 2025
Abstract:
The Opportunity Zone program was designed to encourage investment in distressed communities across the United States. Early research found no evidence of impacts of the program on employment, earnings, or poverty of zone residents, but some evidence of positive effects on employment among businesses in zones. Using the latest survey-based as well as administrative data, we adopt a longer-run and more comprehensive perspective on the labor market impacts of OZs. We find that OZ designation increases job creation among businesses within zones. However, a large share of the newly created jobs in zones is offset by declines in nearby low-income communities. While we detect gains in OZ resident employment over the longer run, the increase comes from jobs with workplaces outside of OZs that, in light of the changing demographic composition of zones, are likely held by new as opposed to existing residents. Overall, our results suggest that OZs have limited benefits for existing residents of targeted areas and are associated mainly with a spatial reallocation of jobs and households.