By Jordyn Gracey
This paper takes the assertion, made by Gentzkow et al., that newspaper slant is primarily determined by slant as given. Both that paper and this one use Hotelling as a foundation. However, this paper considers what happens when the distribution of ideological preferences differs at national and county levels. This paper controls for the size of the market in which the newspapers are operating as well as for the make-‐up of the county-‐level population. Findings show that demand is a robust determinant of slant across market sizes and that supply-‐side factors rarely have significant impact on slant. In the two cases where ownership does have an effect on slant, it is in regressions where the largest-‐circulating newspapers have been dropped. We determine that if ownership is important it is when control is more centralized,if a newspaper is operating in a small market or if the owner chooses the slant before deciding which market to enter.
Advisor: Michelle Connolly | JEL Codes: L2, L21, L4, L22, L25, L44, Y8 | Tagged:
By Peichun Wang
Ever since the Deregulation Act in 1978 in the U.S. airline industry, there have been series of major airline mergers and acquisitions, notably three major waves in the 1980’s, 1990’s, and late 2000’s. These mergers, especially the more recent multi-billion mergers (e.g. Delta- Northwest, United-Continental) have shown a trend of substantial market consolidation that inevitably worries consumers as well as the U.S. Department of Justice (DoJ). Most academic literature to date have tried to study mergers in a static setting where these mergers are assumed to be exogenous. However, the clear pattern of merger waves in the airline industry, as well as many other industries, suggests strong correlation between mergers. A few studies that attempted at a dynamic merger model remain theoretical due to computational barriers. In this paper, I found empirical evidence of merger waves by investigating the change of airline carriers’ incentive to merge after another merger between two other carriers. These results are based on a structural model of the U.S. airline industry, in which I estimate demand with a standard (for dierentiated product markets) discrete-choice nested logit model, but allow for selection on entrants’ costs and qualities, i.e. rms with lower costs and higher qualities would have been selected into the market before the merger, suggesting that post-merger entry is less likely than what non-selective entry models have predicted.
Advisor: James Roberts | JEL Codes: L13, L25, L93 | Tagged: