This paper seeks to understand how market structure for freight transportation affects domestic trade outcomes. I modify a standard Ricardian trade model to incorporate imperfectly-collusive transporters spanning multiple modes. Estimating the model's fundamentals reveals the expected per-mile iceberg trade cost for specific modes, the correlation of trade costs across disparate markets, and the ferocity of competition for freight transportation along individual segments of the domestic transit network. Calibrating the model to domestic trade flows, I find that: i) current losses due to non-competitive pricing are substantial, amounting to roughly 5% of baseline welfare; ii) the exercise of freight market power has an outsize impact on exports; iii) these impacts are concentrated in rural areas throughout the Southeast and Mountain West, as well as small urban areas in the Midwest and Gulf states; iv) exercise of freight market power exacerbates the impact of mode-specific shocks; and v) non-competitive freight pricing does little to attenuate international trade shocks, due to the concentration of freight market power away from international gates.
with Jeronimo Carballo and Richard Mansfield
Multi-establishment firms account for around 60% of U.S. workers' primary employers, providing ample opportunity for workers to change their work location without changing their employer. Using U.S. matched employer-employee data, this paper analyzes workers' access to and use of such between-establishment job transitions, and estimates the effect on workers' earnings growth of greater access, as measured by proximity of employment at other within-firm establishments. While establishment transitions are not perfectly observed, we estimate that within-firm establishment transitions account for 7.8% percent of all job transitions and 18.2% of transitions originating from the largest firms. Using variation in worker's establishment locations within their firms' establishment network, we show that having a greater share of the firm's jobs in nearby establishments generates meaningful increases in workers' earnings: a worker at the 90th percentile of earnings gains from more proximate within-firm job opportunities can expect to enjoy 2% higher average earnings over the following five years than a worker at the 10th percentile with the same baseline earnings.
This paper seeks to understand how freight market concentration changes the impact of domestic market access improvements throughout the late 19th and early 20th Century. I modify a multi-sector Ricardian trade model to allow for endogenous transport costs set by a profit-maximizing freight monopolist. This framework yields a modified market access term that accounts not only for exogenous bilateral frictions, but endogenous bilateral markups. Reduced-form analysis reveals that omitting endogenous freight prices from the calculation of market access reduces the estimated impact by as much as half. I further document marked convexity in this treatment effect by initial output: smaller, remote counties stand the most to gain from market access improvements, which also means these counties are the most at-risk from the exercise of freight market power. I show that, over time, the most-affected counties move west along the American Frontier.
with Jeronimo Carballo and Richard Mansfield