Research

We employ a unique, transaction-level dataset of program licensing by Multichannel Video Programming Distributors (MVPDs) to study the relationship between firm size and bargaining leverage. We find per-subscriber-per-month licensing fees paid by MVPDs to program suppliers decrease significantly with MVPD size but increase significantly with program supplier size, consistent with the bargaining leverage hypothesis. These results are robust across a variety of specifications suggested by Nash bargaining theory. In particular, we control for contract vintage, distributor quality and competitive intensity.

We quantify how bargaining power derived from firm size affects the analysis of downstream mergers and the profitability of downstream entry. We estimate an empirical model of the television industry which features negotiations between upstream content and downstream distributors of varying size. We estimate that large distributors like Comcast are able negotiate about 25% lower content fees than smaller downstream firms. We evaluate the short-run welfare effects of several recently reviewed mergers involving size effects. In conjunction, we assess the degree to which size based bargaining power creates contracts which are a barrier to entry for new distribution firms.

The No Child Left Behind Act’s (NCLB) stated goal was to close the achievement gap. I use a regression discontinuity framework to analyze the NCLB Act’s attempt to reduce the achievement gap by requiring sociodemographic groups to meet group based targets. I study this question using a panel of North Carolina public school students’ test scores in grades three through eight from 2003-2012. I find reading and math scores improve when schools are held accountable for the achievement of each separate group.

This paper tests whether health maintenance organization (HMO) coverage leads to better preventive care for middle-aged and near-elderly adults. Seven common types of preventive care tests and three types of common chronic diseases whose diagnosis is strongly dependent on screening were analyzed. I control for selection into HMOs with a novel quasi-experiment: turnover in insurance contracts at the firm level creates exogenous variation in plan type. Rates of testing are slightly higher for some types of cancer screenings and about the same for other types of screening after controlling for selection. Furthermore, the paper explores self-awareness of several common health conditions: hypertension, hypercholesterolemia, and diabetes. These measures are constructed using a biological screening done by the Health and Retirement Survey. The probability of having one of these conditions and being unaware of it is about the same between members with HMO and non-HMO coverage.