We provide a new method to empirically analyze discrete choice models with state dependence and individual-by-product fixed effects, and use it to analyze consumer choices in a policy-relevant environment (a subsidized health insurance exchange). The method infers state dependence from consumers' switching choices in response to changes in product attributes. Moment inequalities are constructed that do not depend on the fixed effects and do not rely on assumptions about initial conditions.  The method infers much smaller switching costs on the health insurance exchange than would be inferred from standard logit and/or random effects methods.  A counterfactual policy evaluation illustrates that the policy implications of this difference are substantive.

Policymakers can encourage public program enrollment through either financial incentives or streamlined enrollment. We study a defaults policy that streamlines health insurance take-up by auto-enrolling qualifying individuals who fail to respond when asked to select a health plan. This “targeted” auto-enrollment policy has a major impact, increasing enrollment by 30-50% and differentially enrolling young, healthy, low-cost people, with little evidence of improper enrollment of the already-insured. We evaluate the policy's tradeoffs using a model of auto-enrollment as removing a non-price ordeal to take-up. Consistent with the classic rationale for ordeals, marginal enrollees have attributes suggesting low (private) value of insurance. But the same attributes correlate with low public costs, making the targeting efficiency case less clear. Relative to financial subsidies, auto-enrollment has similar targeting properties but is 36-125% more cost-effective by avoiding new spending on inframarginal enrollees.

As the ACA Marketplaces face challenges with high premiums and limited insurer competition, there is significant interest in how policymakers can stabilize markets and control costs. We describe a unique set of active purchasing policies used by Massachusetts’ health insurance exchange to shape the rules of competition and reward lower-price insurers with additional customers. In contrast to the typical focus on recruiting new insurers to an exchange, Massachusetts focused on shaping insurer incentives through carrots and sticks for setting prices below certain thresholds or below other insurers’ prices. We provide evidence that insurer pricing was significantly influenced by active purchasing policies. Between 2010 and 2013, over 80% of insurer prices were set exactly at or within 1% of pricing thresholds created by active purchasing policies. One key “limited choice” policy—which restricted the choice set of fully-subsidized consumers to the two cheapest plans—was associated with a 16-20% reduction in average insurance prices relative to comparative insurance markets in 2012-2014. The state’s slower price growth continued during the ACA’s first years, with the Connector having among the lowest benchmark premiums of any state starting in 2017.