Competition and Costs in Medicare Advantage (JOB MARKET PAPER)
As of 2010, government expenditure per enrollee in Medicare Advantage (MA) was 13% higher than in traditional Medicare. I develop and estimate a model of competition among MA insurers to disentangle the sources of this cost discrepancy. Building on the existing literature, in my model, insurers endogenously choose premiums and the generosity of the coverage, taking into account the complex structure of reimbursements and subsidies employed in MA. My estimates reveal considerable heterogeneity in insurer costs. First, HMO plans, compared to PFFS plans, provide more generous coverage with lower costs. Second, the relationship between plan cost and generosity of the coverage varies across type of plans. Further, my estimates imply that MA generates a surplus larger than its cost to the government; however, this surplus is mainly captured by insurers. To identify possible areas for efficiency gains, I run simulations to study the impact of alternative reimbursements policies, such as premium support, on government expenditure, consumer surplus, and insurer profits.
Insurers Response to Selection Risk: Evidence from Medicare Enrollment Reforms, SIEPR Discussion Paper #15-030, Submitted, with Francesco Decarolis (Boston University)
Evidence on insurers behavior in environments with both risk selection and market power is largely missing. We fill this gap within the context of privatized Medicare providing one of the first empirical accounts of how insurers adjust plan features when faced with a potential change in selection. Our empirical strategy exploits the combined effects of a Medicare reform that altered the potential selection risk of the highest quality (5-star) Part C and D plans and the geographical dispersion of such plans over the US territory. Starting in 2012, exclusively for 5-star plans the open enrollment window was widened to allow enrollments at anytime during the year. We estimate that, due to the reform, the within-year enrollment of 5-star plans increases, but their risk pool does not worsen and actually slightly improves. Correspondingly, when estimating impacts on the market-level distribution of various plan features, we find lower premiums and decreased coverage generosity for 5-star plans relative to competing plans, leading us to argue that 5-star plans became more appealing for most beneficiary, but less so for those in worse health conditions.
The Effect of Discretion on Procurement Performance, R&R Management Science, with Decio Coviello (HEC Montreal) and Giancarlo Spagnolo (SITE – Stockholm School of Economics)
We run a regression discontinuity design analysis to document the causal effect of increasing buyers’ discretion on procurement outcomes in a large database for public works in Italy. Works with a value above a given threshold have to be awarded through an open auction. Works below this threshold can be more easily awarded through a restricted auction, where the buyer has some discretion in terms of who (not) to invite to bid. Our main result is that discretion increases the probability that the same firm wins repeatedly, and it does not deteriorate (and may improve) the procurement outcomes we observe. The effects of discretion persist when we repeat the analysis controlling for the geographical location, corruption, social capital and judicial efficiency in the region of the public buyers running the auctions.