1. Competitive Search, Efficiency, and Multi-worker Firms. International Economic Review 54(1), Feb. 2013, 219-251.
    Published version. Older versions: March 2012 (accepted); September 2010 (incorporating dynamic model); first version: 2005.

    In competitive search where matching is between a single worker and a single job, a firm commits to the wage ex ante. Ex post, if it receives any applicants, it hires one and pays the promised wage. In this paper, I study wage posting when firms' production technology uses multiple workers but with decreasing returns to scale. A single promised wage must play two roles: it promises utility to workers who are hired but also affects how many applicants firms hire. In equilibrium, firms can post multiple wages to attract applicants (and more than two wages can be optimal).

  2. Search with multi-worker firms (with Daron Acemoglu). Theoretical Economics 9(3), Oct. 2014, 583-628.
    Published version. Older versions: May 2013 (accepted). First version: 2006. MATLAB code

    We generalize the standard random search model to the case where firms' production technology exhibits decreasing returns to labor. Wage determination is by bargaining over the marginal surplus, following Stole and Zwiebel. There are convex costs of posting more vacancies. Firms have a long-run target employment which is increasing in their idiosyncratic productivity, and grow to that target size over time. With productivity heterogeneity, the model is consistent with Zipf's and Gibrat's laws. It implies sluggish adjustment to aggregate shocks, as firm entry drives up wages.

  3. A Note on Wage Determination under Mismatch. Macroeconomic Dynamics, published online April 2014.
    Published version. Older versions: January 2014 (accepted); September 2011 version.

    Shimer's (2007, AER) mismatch model assumes competitive wages. I show how to allow for alternative wage determination mechanisms, such as the Shapley value. This gives an improvement  in microeconomic realism. The extra cyclical volatility of Shimer's model, relative to Mortensen-Pissarides, is almost all due to the wage determination assumption.

  4. Bargaining with Commitment between Workers and Large Firms. Review of Economic Dynamics, published online July 2014.
    Published version. Older versions: June 2014. First version: March 2010. Link to supplementary materials.

    Under decreasing returns to labor, if firms and workers bargain over wages, the equilibrium is inefficient: firms over-hire in order to drive down the wages they bargain with their workers. If instead firms and workers can sign long-term contracts, this inefficiency can be resolved. There can still be search inefficiencies if firms of different marginal values of hiring hire at the same time.

Working papers:

  1. Financial Frictions and Occupational Mobility (with Jose Mustre-del-Rio)

    When there are no complete markets against occupational risk, worker occupational mobility is affected. There are large welfare costs of incomplete markets, not just from worse consumption smoothing but also because workers leave occupations in which they have specific human capital (in case things get worse), rather than waiting to see if the occupation recovers.

  2. Identifying the Nature of Bargaining between Workers and Large Firms

    If bargaining causes firms to over-hire, then policy to reduce excessive vacancy posting could improve welfare. However, distinguishing over-hiring may be impossible using labor market data alone. It is also necessary to look at how much firms spend on recruiting.

  3. Do Large-Firm Bargaining Models Amplify and Propagate Aggregate Productivity Shocks?

    The ability of a search and matching model to match the cross-sectional distribution and dynamics of employment has no implications for its cyclical properties, and specifically, for the cyclical volatility of unemployment and vacancies.

  4. Worker Flows under Mismatch

  5. The Joint Dynamics of Capital and Employment at the Plant Level (with Ryan Michaels and Jiyoon Oh)