Nathan Yang

Unobserved Heterogeneity in Dynamic Games: Cannibalization and Preemptive Entry of Hamburger Chains in Canada

We study cannibalization and preemption in the evolution of market structure. Because a market can accommodate only a finite number of products/outlets, forwardlooking firms face the tradeoff between cannibalization and preemption. We develop a dynamic entry model of multi-product oligopoly, and estimate it using data on hamburger shops in Canada (1970–2005). The results suggest cannibalization is the main determinant of profit and entry. Moreover, counterfactual simulations imply preemptive motives generate highly nonlinear entry strategies, and caution against the use of static approaches when market structure is evolving.We develop a dynamic entry model of multi-store oligopoly with heterogeneous markets, and estimate it using data on hamburger chains in Canada (1970-2005). Because more lucrative markets attract more entry, firms appear to favor the presence of more rivals. Thus unobserved heterogeneity across geographical markets creates an endogeneity problem and poses a methodological challenge in the estimation of dynamic games, which we address by combining the procedures proposed by Kasahara and Shimotsu (2009), Arcidiacono and Miller (2011), and Bajari, Benkard, and Levin (2007), respectively. The results suggest the omission of unobserved market heterogeneity attenuates the estimates of competition, and the tradeoff between cannibalization and preemption is an important factor behind the evolution of market structure.

Burger King and McDonald’s: Where’s the Spillover?

Abstract This paper studies how spillover effects from competitors’ choices affect a firm’s decision to open a store. Using panel data from the UK’s fast food industry, I propose and estimate a game of entry under incomplete information that incorporates spillover effects between firms’ entry decisions. A positive spillover is identified for Burger King – increasing the stock of existing McDonald’s by one outlet increases Burger King’s estimated equilibrium probability of opening a new store by approximately 18 percentage points. Furthermore, the estimated model suggests that this spillover affects Burger King’s variable profit, as opposed to its fixed cost of entry. It is less clear whether this externality matters for McDonald’s.

Firm Expansion, Size Spillovers, and Market Dominance in Retail Chain Dynamics

We develop and estimate a dynamic game of strategic firm expansion and contraction decisions to study the role of firm size in future profitability and market dominance. Modeling firm size is important because retail chain dynamics are more richly driven by expansion and contraction than de novo entry or permanent exit. Additionally, anticipated size spillovers may influence the strategies of forward-looking firms, making it difficult to analyze the effects of size without explicitly accounting for these in the expectations and, hence, decisions of firms. Expansion may also be profitable for some firms while detrimental for others. Thus, we explicitly model and allow for heterogeneity in the dynamic link between firm size and profits as well as potential for persistent brand effects through firm-specific unobservable factors. As a methodological contribution, we surmount the hurdle of estimating the model by extending a two-step procedure that circumvents solving the game. The first stage combines semipar...

Learning in Retail Entry

This paper advances our collective knowledge about the role of learning in retail agglomeration. Uncertainty about new markets provides an opportunity for sequential learning, where one firms past entry decisions signal to others the potential profitability of risky markets. The setting is Canadas hamburger fast food industry from its early days in 1970 to 2005, for which simple analysis of my unique data reveals empirical patterns pointing towards retail agglomeration. The notion that uninformed potential entrants have an incentive to learn, but not informed incumbents, motivates an intuitive double-difference approach that separately identifies learning by exploiting differences in the way potential entrants and incumbents react to spillovers. This identification strategy confirms that information externalities are key drivers of agglomeration. Estimates from a dynamic oligopoly model of entry with information externalities provide further evidence of learning, as I show that common uncertainty matters. Counterfactual analysis reveals that an industry with uncertainty is initially less competitive than an industry with certainty, but catches up over time. Furthermore, there are many instances in which chains enter markets they would have avoided had they not faced uncertainty. Finally, consistent with the interpretation of uncertainty as an entry barrier, I find that chains place significant premiums on certainty at proportions beyond 2% of their total value from being monopolists.Retailers may face uncertainty about the profitability of local markets, which provide opportunities for learning when making entry decisions. To study these informational aspects, I develop an empirical framework for studying industry dynamics with uncertainty and learning (from others). Using data about fast food chains, I estimate the model with a two-step estimation approach augmented with particle filtering as a way to flexibly account for unobserved posterior beliefs about market profitability, as well as persistent unobserved states that evolve exogenously. The estimates confirm the presence of uncertainty and learning. Furthermore, simulations using the estimated model demonstrate that learning from others partially offsets the negative business-stealing effects of rival presence.

Harnessing the Small Victories: Empirical Evidence from a Calorie and Weight Loss Tracking Application

We investigate the role of short-term goal achievement on long-term goal achievement under the context of weight loss. Using unique and large-scale data from a freemium mobile weight management application (Lose It!), we track the daily dynamics of weight loss across a large number of users. The application sets a salient daily budget for calories, and by comparing cases in which the user is slightly under or over-budget, we provide a causal link between short-term goal achievement and long-term outcomes such as future weight loss, achievement of goal weight, and setting of more ambitious weight loss targets. Short-term goal achievement also have implications on future customer development as staying within the daily budget leads to an increase in premium account upgrades. Furthermore, we show that the impact of short-term goal achievement varies across user segments. We later demonstrate using a dynamic regression discontinuity design that the short-term goal achievement effects persist over time, and in fact, induce users to accomplish even more ambitious short-term goals in the future. Finally, estimates from a dynamic structural model of calories management reveal that users receive positive utility from past short-term goal accomplishments, and counterfactual analysis with the estimated model quantify the long-run user benefits of various hypothetical policies that adjust the daily budget of calories.

Aggressive Growth in Retail: A Trade-Off between Deterrence and Survival?

We study the relationship between aggressiveness and survival under the context of retail entry strategy. Using a dynamic game of entry, calibrated with novel monthly data about fast casual taco chains (Chipotle and Taco Cabana) in Texas (1993-2015), we conduct a series of simulations to explore the extent to which a trade-off exists between aggression and survival. The results show that this trade-off can exist in certain conditions. Most importantly, we establish a set of qualifiers (i.e., opportunity zone) for aggressive entry to improve survival: (1) A rivals presence has to hurt the incumbents profit enough to make preemption attractive, (2) the damage of the rivals presence on the firms profit should not be too much (in the event that rival still enters), and (3) the incumbent has to be competitive enough so as to hurt its rivals profit sufficiently.