## A Simple Estimator for the Distribution of Random Coefficients

Stephen P. Ryan, Jeremy Fox, Kyoo-il Kim, and Patrick Bajari, Quantitative Economics, Vol.2, No. 3 (2011), pp. 381–418.

We propose a simple mixtures estimator for recovering the joint distribution of parameter heterogeneity in economic models, such as the random coefficients logit. The estimator is based on linear regression subject to linear inequality constraints, and is robust, easy to program, and computationally attractive compared to alternative estimators for random coefficient models. For complex structural models, one does not need to nest a solution to the economic model during optimization. We present a Monte Carlo study and an empirical application to dynamic programming discrete choice with a serially correlated unobserved state variable.

[PDF]

## Delivering Education to the Underserved through a Public-Private Partnership Program in Pakistan

Barrera-Osorio, Felipe, David S. Blakeslee, Matthew Hoover, Leigh Linden, Dhushyanth Raju, and Stephen P. Ryan. “Delivering education to the underserved through a public-private partnership program in Pakistan.” Review of Economics and Statistics 104, no. 3 (2022): 399-416.

Governments are increasingly partnering with the private sector to improve the delivery of education. We evaluate an innovative program that enlisted local entrepreneurs to open and operate new schools in 200 randomly selected, underserved villages in Sindh, Pakistan. School operators received a per-student subsidy from the local government to provide tuition-free primary education, and in half of the treated villages operators received a higher subsidy for female students. Management of these schools was highly decentralized, with school operators permitted to tailor inputs to local demand. The program increased enrollment in treatment villages by 31 percentage points, and test scores by 0.64 standard deviations, with no difference by gender or across the two subsidy schemes. Treatment effects are driven primarily by the establishment of schools in villages where they were previously absent, though program schools also improve educational outcomes even when nearby government schools are available. To gain greater insight into program school efficiency, we estimate a structural model of the demand and supply for school inputs. This exercise reveals that program schools selected inputs similar to those of a social planner who internalizes all the education benefits to society.

[PDF]

## Evaluating Wireless Consolidation Using Semiparametric Demand Estimation

Bajari, Patrick, Jeremy T. Fox, and Stephen P. Ryan. 2008. “Evaluating Wireless Consolidation Using Semiparametric Demand Estimation,” Quantitative Marketing and Economics, Vol. 6, No. 4, pp 299-338.

The US mobile phone service industry has dramatically consolidated over the last two decades. One justification for consolidation is that merged firms can provide consumers with larger coverage areas at lower costs. We estimate the willingness to pay for national coverage to evaluate this justification for past consolidation. As market level quantity data are not publicly available, we devise an econometric procedure that allows us to estimate the willingness to pay using market share ranks collected from the popular online retailer Amazon. Our semiparametric maximum score estimator controls for consumers’ heterogeneous preferences for carriers, handsets and minutes of calling time. We find that national coverage is strongly valued by consumers, providing an efficiency justification for across-market mergers. The methods we propose can estimate demand for other products using data from online retailers where product ranks, but not quantities, are observed.

[PDF]

## Heterogeneity and the Dynamics of Technology Adoption

Stephen P. Ryan and Catherine Tucker, “Heterogeneity and the Dynamics of Technology Adoption,” Quantitative Marketing and Economics, Volume 10, Issue 1 (2012), pp. 63–109.

We estimate the demand for a videocalling technology in the presence of both network effects and heterogeneity. Using a unique dataset from a large multinational firm, we pose and estimate a fully dynamic model of technology adoption. We propose a novel identification strategy based on post-adoption technology usage to disentangle equilibrium beliefs concerning the evolution of the network from observed and unobserved heterogeneity in technology adoption costs and use benefits. We find that employees have significant heterogeneity in both adoption costs and network benefits, and have preferences for diverse networks. Using our estimates, we evaluate a number of counterfactual adoption policies, and find that a policy of strategically targeting the right subtype for initial adoption can lead to a faster-growing and larger network than a policy of uncoordinated or diffuse adoption.

[PDF]

## Identification and Estimation of a Discrete Game of Complete Information

Bajari, Patrick, Han Hong, and Stephen P. Ryan. “Identification and estimation of a discrete game of complete information.” Econometrica 78, no. 5 (2010): 1529-1568.

We discuss the identification and estimation of discrete games of complete information. Following Bresnahan and Reiss (1990, 1991), a discrete game is a generalization of a standard discrete choice model where utility depends on the actions of other players. Using recent algorithms to compute all of the Nash equilibria to a game, we propose simulation-based estimators for static, discrete games. We demonstrate that the model is identified under weak functional form assumptions using exclusion restrictions and an identification at infinity approach. Monte Carlo evidence demonstrates that the estimator can perform well in moderately sized samples. As an application, we study entry decisions by construction contractors to bid on highway projects in California. We find that an equilibrium is more likely to be observed if it maximizes joint profits, has a higher Nash product, uses mixed strategies, and is not Pareto dominated by another equilibrium.

## Incentives Work: Getting Teachers to Go to School

Esther Duflo, Rema Hanna, and Stephen P. Ryan, American Economic Review, Volume 102, Issue 4, June 2012, pp. 1241–78

We use a randomized experiment and a structural model to test whether monitoring and financial incentives can reduce teacher absence and increase learning in India. In treatment schools, teachers’ attendance was monitored daily using cameras, and their salaries were made a nonlinear function of attendance. Teacher absenteeism in the treatment group fell by 21 percentage points relative to the control group, and the children’s test scores increased by 0.17 standard deviations. We estimate a structural dynamic labor supply model and find that teachers respond strongly to financial incentives. Our model is used to compute cost-minimizing compensation policies.

[PDF]

## Linear Regression Estimation of Discrete Choice Models with Nonparametric Distributions of Random Coefficients

Bajari, Patrick, Jeremy T. Fox, and Stephen P. Ryan. 2007. “Linear Regression Estimation of Discrete Choice Models with Nonparametric Distributions of Random Coefficients.” American Economic Review P&P, 97 (2): 459-463.

This paper describes a method for estimating random coefficient discrete choice models that is both flexible and simple to compute. We demonstrate that, with a finite number of types,

choice probabilities are a linear function of the model parameters. Because of this linearity,

our model can be estimated using linear regression subject to inequality constraints. We can approximate an arbitrary distribution of random coefficients by allowing the number of types to be sufficiently large. Therefore, we say our estimator is nonparametric for the distribution of heterogeneity.

[PDF]

## Machine Learning Methods for Demand Estimation

Patrick Bajari, Denis Nekipelov, Stephen P. Ryan, and Miayou Yang, American Economic Review: Papers and Proceedings, 2016, Vol. 105, No. 5, pp. 481-85.

We survey and apply several techniques from the statistical and computer science literature to the problem of demand estimation. To improve out-of-sample prediction accuracy, we propose a method of combining the underlying models via linear regression. Our method is robust to a large number of regressors; scales easily to very large data sets; combines model selection and estimation; and can flexibly approximate arbitrary non-linear functions. We illustrate our method using a standard scanner panel data set and find that our estimates are considerably more accurate in out-of-sample predictions of demand than some commonly used alternatives.

[PDF]

## Market-based Emissions Regulation and Industry Dynamics

Meredith Fowlie, Mar Regaunt, and Stephen P. Ryan, Journal of Political Economy, Vol. 124, No. 1, 2016, pp. 249–302

We assess the static and dynamic implications of alternative market-based policies limiting greenhouse gas emissions in the US cement industry. Our results highlight two countervailing market distortions. First, emissions regulation exacerbates distortions associated with the exercise of market power in the domestic cement market. Second, emissions “leakage” in trade-exposed markets offsets domestic emissions reductions. Taken together, these forces can result in social welfare losses under policy regimes that fully internalize the emissions externality. Market-based policies that incorporate design features to mitigate the exercise of market power and emissions leakage deliver welfare gains when damages from carbon emissions are high.

[PDF]

## Selection on Moral Hazard in Health Insurance

Stephen P. Ryan, Liran Einav, Amy Finkelstein, Paul Schrimpf, and Mark Cullen, American Economic Review, Volume 103, Issue 1, February 2013, pp. 178–219.

We use employee-level panel data from a single firm to explore the possibility that individuals may select insurance coverage in part based on their anticipated behavioral (“moral hazard”) response to insurance, a phenomenon we label “selection on moral hazard.” Using a model of plan choice and medical utilization, we present evidence of heterogenous moral hazard as well as selection on it, and explore some of its implications. For example, we show that, at least in our context, abstracting from selection on moral hazard could lead to overestimates of the spending reduction associated with introducing a high-deductible health insurance option.