Superstar Teams
[Latest version (v2.0) | Dec 2024] -- New version!

A theory where firm performance is rooted in the quality of its workforce; and heterogeneity across firms naturally arises from talent sorting motivated by complementarities in team production that endogenously arise due to skill specialization. This theory can help explain salient features and trends in the macro economy, such as the growing importance of between-firm wage inequality. It also demonstrates that for productivity gains from specialization to materialize, workers need to be matched with colleagues possessing complementary expertise.

• Econ JM Best Paper Award, European Economic Association/Unicredit Foundation 

• Companion note "Human Capital at Work: Five Facts about the Role of Skills for Firm Productivity, Growth, and Wage Inequality" (OECD/with C. Criscuolo & P. Gal)
  [OECD WP] [Manuscript]

• [SSRN WP (v1.3) | Nov 2023] 

For Whom the Bot Tolls: Specialization and the Earnings Effects of AI (with L. Mann) -- Draft coming soon!
Who will gain and who will lose from AI-induced task automation? This paper argues that skill specialization is an important part of the answer. In a canonical task-based model, in which occupations bundle tasks and workers possess heterogeneous portfolios of task-specific skills, workers whose skills are concentrated in automated tasks face losses, while automation frees others from performing tasks they were relatively less skilled in. Quantifying these effects requires measuring (i) the exposure of tasks to automation, for which we exploit a direct mapping between our model and prominent exposure measures; (ii) the task content of occupations, which we recover by leveraging advances in natural language processing; and (iii) the multidimensional distribution of skills. We use the model's structure to estimate this skill distribution from data on wages and occupational choices. We show that workers' skill specialization explains empirically salient features of mobility. In a proof-of-concept study guided by empirical exposure measures, we then quantify the dynamic earnings effects of tools that automate data analysis. Our results show that occupation-level outcomes hide substantial individual-level heterogeneity. While selection into occupation based on skill specialization implies that incumbents lose on average, wage declines are concentrated among leavers, whereas stayers win by shifting towards other productive tasks. Meanwhile, workers entering highly exposed occupations benefit from the transformed task composition. Our framework provides a flexible approach for analyzing the distributional impacts across multiple types of automation shocks.

• Presentations: Columbia U, Minneapolis Fed, BSE Summer Forum, SED

Volatile Hiring: Uncertainty in Search and Matching Models, with W. Den Haan and P. Rendahl
Journal of Monetary Economics, Vol. 123, pp. 1-18, 2021 [Published version - open access],  [Online appendix ], [Replication files], [Short slides | April 2021], [VoxEU column | Sep 2021] 

We study the hypothesis that heightened uncertainty leads to higher unemployment because firms prefer to adopt a "wait-and-see" approach to posting vacancies. Contrary to common belief, option-value considerations  play no role in the standard search-and matching model with free entry. Constructively, we show that when the mass of entrepreneurs is finite and there is heterogeneity in firm-specific productivity, a rise in perceived uncertainty robustly increases the option value of waiting and reduces job creation.