Hamid Firooz

12/ This is because a robot subsidy can alleviate markup distortions and, by lowering the user cost of robots, a subsidy reallocates production toward large, automating firms that have high productivity, bringing aggregate output closer to the efficient level.

11/ Under our calibration, a modest robot subsidy (about 1.41% of the value of robots) maximizes the steady-state welfare, yielding a welfare gain equivalent to about 4.23% of steady-state consumption compared to the laissez-faire benchmark.

10/ Under our calibration, the model predicts that a 40% fall in the relative price of robots—observed during the past two decades—can explain ~50% of the rise in sales concentration in the U.S. manufacturing and ~25% of the divergence between sales and employment concentration.

9/ the Annual Business Survey: The adoption rate of robots in 2016-2018 by firms in the top percentile of the employment distribution within 6-digit industries was about 3 times of the adoption rate among firms in the 50th to 75th percentile (5.1% vs. 1.7%).

8/ We calibrate the model to match several moments observed in the U.S. manufacturing sector. Our calibrated model predicts that the usage of automation technology is highly skewed toward a small fraction of large firms, in line with the cross-sectional evidence from ...

7/ Since robots substitute for workers, the expansion of those large firms relies more on robots than on workers. Thus, a rise in automation raises sales concentration more than employment concentration, as we observe in the data.

6/ Automation improves a firm’s labor productivity, allowing large, robot-using firms to expand their sales share further. This economy-of-scale effect leads to a positive connection between automation and sales concentration.

5/ Operating the automation technology incurs a fixed cost, but it reduces the marginal cost of production. Larger firms are more likely to automate because they have higher productivity, higher market power, and thus higher profits.

4/ To understand the empirical link between automation and industry concentration, we construct a dynamic general equilibrium model featuring heterogeneous firms, endogenous automation decisions, and variable markups.

3/ employment concentration. These empirical relations are robust when we estimate the correlations using an instrumental variable (IV) approach.

2/ manufacturing and ~25% of the divergence between sales and employment concentration.
We first document that, at the industry level, robot density has a significantly positive correlation with sales concentration but a small and insignificant correlation with ...

1/ A thread on “Automation and the Rise of Superstar Firms” recently published @ Journal of Monetary Economics. Link to the paper:
drive.google.com/file/d/1KvVA...
Summary: The rise in automation in the past two decades explains ~50% of the rise in sales concentration in the US ...

New at JIE: "Reshoring, automation, and labor markets under trade uncertainty", by Hamid Firooz (@hamid-firooz.bsky.social), Sylvain Leduc and Zheng Liu

doi.org/10.1016/j.ji...

11/11 Instead, it raises unemployment, thereby turning the pro-competitive gains from trade negative. In the language of Arkolakis et al. (2019), I show that the pro-competitive gains from trade liberalization can be even more “elusive” in the presence of labor market frictions.

10/ resulting in a more efficient allocation of resources and therefore leading to positive pro-competitive gains from trade. In frictional labor markets, however, the increase in firms’ responsiveness does not costlessly shift employment toward more efficient firms.

9/ taking labor market frictions into account makes the pro-competitive gains from trade liberalization negative. In the absence of search frictions, trade liberalization makes firms more responsive to their idiosyncratic shocks via variable markups, ...

8/ The surge in unemployment caused by variable markups tends to reduce the pro-competitive gains from trade in the steady state. Interestingly, I show that while abstracting from labor market frictions in my framework implies positive pro-competitive gains from trade, ...

7/ This in turn increases the labor elasticities of revenue, which makes firms’ employment decisions more responsive to their idiosyncratic shocks. More responsiveness to shocks increases unemployment and residual wage inequality in a frictional labor market.

6/ The counterfactual outcomes under variable markups are more consistent with Colombia's trade liberalization.
The link between trade, markups, and labor market outcomes is due to trade liberalization increasing product market competition, thereby raising demand elasticities.

5/ In particular, I show that once markups are allowed to respond to trade liberalization, unemployment and residual wage inequality rise almost three times more than in a model with constant markups (in the steady state).

4/ I develop and estimate a dynamic general equilibrium model of international trade featuring
endogenously variable markups to show that trade-induced changes in markups have important
impacts on labor market outcomes in a frictional labor market.

3/ while Colombia’s trade liberalization in the 1990s raises its real income by 1.7% (in the steady state) under constant markups, accounting for variable markups reduces these gains to only 0.3%.

2/ Moreover, the pro-competitive “gains” from trade in a frictional labor market are quite negative—that is, the steady-state welfare gains are smaller under variable markups than those under constant markups:

1/ A thread on “The Pro-Competitive Consequences of Trade in Frictional Labor Markets” published @jintlecon.bsky.social. Link to the paper:
drive.google.com/file/d/1gNpO...
Summary: Taking variable markups into account in frictional labor markets reduces the gains from trade liberalization by 80%.