Matching managerial talent to organizations shapes brand performance and raises a central question: what is the value of managers to organizations? Recovering this value is difficult because observed matches arise from a competitive two-sided market equilibrium. We develop a structural two-sided transferable-utility matching model, extending the literature by endogenizing organizations' decisions to dismiss and replace managers. With this new mechanism, we establish nonparametric identification of the value of managerial characteristics, complementing the match synergies identified in Fox (2010). Together, these two components reveal which characteristics each organization should prioritize in manager selection. For estimation, we extend the seminal matching maximum score estimator to incorporate dissolution decisions and realized performance outcomes with the performance verified by Monte Carlo simulations. We quantify managerial characteristics by applying a large language model to publicly available text. Applying the model to the English Premier League, we find priorities differ sharply: tactical expertise is most valuable to elite clubs, interpersonal skills matter most outside the top tier, and misconduct is especially costly for weaker clubs, where its harm even outweighs tactical gains. Counterfactuals show that one in four appointments of a manager from a higher-ranked club backfires. There is no universal ``good" manager: a manager's value depends jointly on managerial quality and organizational fit. The framework applies to other two-sided matching markets, including executive recruitment and influencer-brand partnerships.

Whether and with whom to form an alliance in Pharmaceutical R&D is a very high-stakes decision for a firm. We provide evidence on partner-specific attributes and synergies that lead to more valuable alliances for pharmaceutical innovation. We overcome the key challenge of endogeneity in alliance formation in a market equilibrium by estimating a two-sided matching model using a semi-parametric maximum score estimator framework. We extend the literature by (1) allowing firms to "go-it-alone'', thereby recovering the value of partner-specific characteristics necessary to compute an alliance's (total) value, and (2) estimating an alliance's value for an outcome, i.e., FDA approvals. Using data on 2,912 clinical trials between 2000 and 2011 by 301 firms, we find that more valuable alliances (ceteris paribus) comprise a primary firm with a partner that has: (i) greater experience working in alliances, with synergies from such experience, and (ii) a more specialized scope of experience with diseases and with drug classes (e.g., organic compounds), while the primary firm itself contributes more value with a broader scope. Thus, a potential partner's experience with alliances should be a critical consideration in forming an alliance. Consequently, alliance experience per se can be a strategic investment that confers a competitive advantage for firms seeking alliances. Our procedure has a microfoundation interpretation in the Heckman (1979) selection framework and wider applicability to other two-sided settings (e.g., digital platforms, buyer–supplier partnerships, co-branding, etc.).

In this paper, we investigate whether multimarket contact leads to collusive pricing in the retail lumber market and assess the effect of potential regulation on consumer welfare. In particular, we focus on the competition between Home Depot and Lowe's since they are the main players in retail lumber market. For our analysis, we assemble an original data set of prices and sales in the pressure treated lumber category. We first provide reduced form evidence for the effect of multimarket contact on collusion in pricing strategies by showing a positive and significant correlation between price and multimarket contact. Guided by this evidence we estimate a model, using the Berry et al. (1995) equilibrium framework, that includes both the demand and supply side with conduct parameters that capture the degree of collusion in setting prices. We specify the conduct parameters as a function of multimarket contact, following the framework developed by Sudhir (2001) and Ciliberto & Williams (2014). We find that the conduct parameter for capturing the effect of multimarket contact on collusion is significant and positive, implying that multimarket contact leads to higher prices than those from a competitive Bertrand-Nash equilibrium. Using the model estimates, we conduct a counterfactual analysis to measure the consumer welfare impact due to collusive pricing. We find that consumer surplus in January 2016 increases by $57.7k and this amounts to $692.4k in a year. To the best of our knowledge, this is the first paper that shows state-level multimarket contact facilitates collusive pricing with implications for consumer welfare. The finding has important policy implications that might suggest the need for monitoring prices in states where retail firms have a greater degree of multimarket contact.