Browsing by Author "Fox, Jeremy T."

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    A note on identification of discrete choice models for bundles and binary games
    (Wiley, 2017) Fox, Jeremy T.; Lazzati, Natalia
    We study nonparametric identification of single-agent discrete choice models for bundles (without requiring bundle-specific prices) and of binary games of complete information. We show that these two models are quite similar from an identification standpoint. Moreover, they are mathematically equivalent when we restrict attention to the class of potential games and impose a specific equilibrium selection mechanism in the data generating process. We provide new identification results for the two related models.
  • Thumbnail Image
    ItemEmbargo
    Essays on Two-Sided Matching Games
    (2023-04-12) Kazempouresmati, Amir; Fox, Jeremy T.
    Empirical methods for transferable-utility matching games have previously been developed using the key outcome of the matches formed in equilibrium. In the first chapter, we explore the identification and estimation of match production functions and agent valuation functions using data on two additional outcomes of such matching games: monetary transfers (prices) and profits. We provide identification results for nonparametric models for the case of data on profits and more parametric models for the case of data on prices. We provide estimators paralleling the identification results for both profit data and price data. Importantly, our identification results allow for agents to have valuations defined over the unmeasured characteristics of potential partners. Further, using data on CEO compensation of large public firms in the U.S., we estimate the effects of the CEO’s experience, the scope of the firm’s operations, and the firm’s size on the match valuations of CEOs and firms. In the second chapter, we investigate how data on firms’ profits allow the identification of the distribution of unobserved heterogeneity under a specific equilibrium selection rule in a two-sided matching model with a finite number of firms. We prove an extension of Theorem 9 in Heckman and Honoré (1990) where we show data on market-wide profits identify the joint distribution of total unobserved match productions. Finally, under an equilibrium selection rule that relies on the core structure in assignment games, we show that data on equilibrium matching, individual firm profits, and match-level measured characteristics of all agents identify the joint distribution of match-level unmeasured characteristics.
  • Thumbnail Image
    Item
    Estimating matching games with transfers
    (Wiley, 2018) Fox, Jeremy T.
    I explore the estimation of transferable utility matching games, encompassing many‐to‐many matching, marriage, and matching with trading networks (trades). Computational issues are paramount. I introduce a matching maximum score estimator that does not suffer from a computational curse of dimensionality in the number of agents in a matching market. I apply the estimator to data on the car parts supplied by automotive suppliers to estimate the valuations from different portfolios of parts to suppliers and automotive assemblers.
  • Thumbnail Image
    Item
    Reallocating Bonus Payments Through Competition to Improve Medicare Advantage Plan Quality: A Dynamic Game Approach
    (2022-04-21) Bhowal, Rajarshi; Fox, Jeremy T.
    This thesis contains two chapters that study competition among firms in regulated markets and auctions. The first chapter of the thesis explores how competition among firms can be used to improve the quality of plan offerings in a managed care setting like Medicare Advantage through changes in the reimbursement policy of the government. In a managed market, private firms provide government sponsored services at regulated prices and compete for subsidies. I study how firms offering Medicare Advantage plans compete in terms of quality and evaluate how the markets would evolve under a competitive bonus payment system rewarding them based on their relative quality performance in a local market. I introduce a dynamic discrete game model of firm quality investment choice and use it to estimate the cost of quality improvements. The estimated model is used to predict the market outcomes in terms of average plan quality under the alternative payment system by calculating the counterfactual equilibria and simulating the markets forward. The results show that 65\% of the counties improve under the new payment rule compared to their observed outcomes in the data, with underperforming counties improving more. \\ The second chapter of the thesis studies collusive behavior of bidders in an ascending price auction. Any exchange of private information or an agreement among participants regarding their behavior in an auction is prohibited by anti-trust laws. I introduce a method of empirically detecting the identities of such colluders under selective entry using the participation and bidding behavior of the potential bidders. I discard the assumption that only the ring leader is observed to participate in every auction, which has been previously used to create the suspect group. I then devise a two-step estimation process where the identity of the colluding members is identified from the identity of the participants, winners, and the value of the winning bid in an ascending price auction. We use both the participation probabilities and distribution of final price across auctions to create the suspect group and show that the identities of the colluders can be detected using a pairwise comparison index.
  • Thumbnail Image
    Item
    Unobserved Heterogeneity in Matching Games
    (The University of Chicago Press, 2018) Fox, Jeremy T.; Yang, Chenyu; Hsu, David H.
    Agents in two-sided matching games vary in characteristics that are unobservable in typical data on matching markets. We investigate the identification of the distribution of unobserved characteristics using data on who matches with whom. In full generality, we consider many-to-many matching and matching with trades. The distribution of match-specific unobservables cannot be fully recovered without information on unmatched agents, but the distribution of a combination of unobservables, which we call unobserved complementarities, can be identified. Using data on unmatched agents restores identification.