1: \begin{abstract}

2: Many real-life settings of consumer-choice involve social interactions,

3: causing targeted policies to have spillover-effects. This paper develops novel

4: empirical tools for analyzing demand and welfare-effects of

5: policy-interventions in binary choice settings with social interactions.

6: Examples include subsidies for health-product adoption and vouchers for

7: attending a high-achieving school. We establish the connection between

8: econometrics of large games and Brock-Durlauf-type interaction models, under

9: both I.I.D. and spatially correlated unobservables. We develop new convergence

10: results for associated beliefs and estimates of preference-parameters under

11: increasing-domain spatial asymptotics. Next, we show that even with fully

12: parametric specifications and unique equilibrium, choice data, that are

13: sufficient for counterfactual \emph{demand}-prediction under interactions, are

14: \emph{insufficient} for welfare-calculations. This is because distinct

15: underlying mechanisms producing the same interaction coefficient can imply

16: different welfare-effects and deadweight-loss from a policy-intervention.

17: Standard index-restrictions imply distribution-free bounds on welfare. We

18: illustrate our results using experimental data on mosquito-net adoption in

19: rural Kenya.

20: 

21: \end{abstract}