Working Papers
Identification and Estimation of Binary Choices with Hidden Information Diffusion [pdf]
Abstract: This paper studies the hidden information diffusion and consumer preferences for a new product, where only informed individuals who have acquired product information can make adoption decisions. Due to unobserved information acquisition/diffusion, we cannot distinguish informed but unwilling-to-adopt individuals from uninformed ones. In addition to a random utility model, we introduce a binary threshold crossing model to characterize information diffusion: an individual becomes informed if the exposure to the information, which depends on the fraction of already informed neighbors and personal attributes, exceeds some idiosyncratic threshold. We show the identification of diffusion and utility parameters through two subsets of observations: (i) adoption behaviors of seeds (first-informed individuals) over time; (ii) adoption behaviors of seeds' neighbors at time two. We propose a two-step estimation procedure: in the first stage, we estimate diffusion and utility parameters using the subset of observations that yields identification; in the second stage, we re-estimate utility parameters using all the observations and plug-in first-step estimates of diffusion parameters. With regularity conditions on decaying dependence with distance and network structure, we establish consistency of the two-step estimator. Monte Carlo simulations demonstrate the efficiency gain from the two-step estimator over the first-step estimator. As a counterfactual analysis, we present a greedy seeding strategy using network structure, the pattern of information diffusion, and consumer preferences to select seeds. Simulation studies indicate that the proposed strategy yields higher mean adoption rates than many existing strategies uniformly across seed set sizes.
Presentation: American Economic Association Annual Meeting 2024 (Poster), Southern Economic Association 93rd Annual Meeting 2023, California Econometrics Conference 2023 (Poster), Asian Meeting of the Econometric Society in China 2023 (Virtual), North American Summer Meeting of the Econometric Society 2023, Seminar at Peking University - Guanghua School of Management 2022 (Virtual)
Nonparametric Identification in Generalized Separable Models
Abstract: This paper provides new sufficient conditions for nonparametrically identifying structural models with generalized additive or multiplicative separability. The conditions generalize common normalizations used in the literature, such as differentiability of link and component functions, functional form normalization, and large support of an explanatory variable. We show that when a generalized separable model has two component functions, three location normalizations and weak support conditions can lead to the identification. The new results can be applied to a wide range of influential econometric models after transformation.
Inference for Social Interactions in Large Endogenous Networks, with Shuo Qi
Abstract: We study the identification and estimation of social interactions in large endogenous networks. Our analysis focuses on binary-action games of incomplete information in which an agent's expected payoff relies on her characteristics, peers' average characteristics, the average of her beliefs about peers' actions, and some preference shocks. Endogeneity in networks results from unobserved characteristics that affect both link formation and individual decision-making. To identify the utility functions, we express the unobserved characteristics as some unknown function of observed variables and address the endogeneity issue through a control function approach. The identification strategy holds even if multiple equilibria exist. We employ the strategy to develop a semiparametric estimator and assess finite-sample performance through Monte Carlo simulations.