Inference on Market Power in Markets with Multiple Equilibria
Project Information
- Discipline
- Economics (903)
- Orientation
- Research
We provide a method to estimate demand functions and infer market power when firms strategically self-select themselves into markets that better match their characteristics. The econometric problem is a multivariate version of the classic Gronau/Heckman self-selection model where multiple agents play an entry game. Our methodology takes into account both the selection problem and the endogeneity of price in the demand equation and so we allow researchers to assess whether correcting for selection (in addition to endogeneity) is empirically important. We use cross-sectional data from the US airline industry to estimate the demand for airline travel controlling for self-selection. We use the demand estimates and a common assumption on firm conduct to infer the market power of US airline firms. We find that the price-cost percentage margins are twice as large when market structure is assumed to be endogenous to the pricing decision than when we assume that the entry decision is exogenous.
Intellectual MeritTheir methodology takes into account both the selection problem and the endogeneity of price in the demand equation and so we allow researchers to assess whether correcting for selection (in addition to endogeneity) is empirically important. Used cross-sectional data from the US airline industry to estimate the demand for airline travel controlling for self-selection. Used the demand estimates and a common assumption on firm conduct to infer the market power of US airline firms.
Broader ImpactsFound that the price-cost percentage margins are twice as large when market structure is assumed to be endogenous to the pricing decision than when we assume that the entry decision is exogenous.
Project Contact
- Project Lead
- Katherine Holcomb (kholcomb)
- Project Manager
- Katherine Holcomb (kholcomb)
Resource Requirements
- Hardware Systems
-
- alamo (Dell optiplex at TACC)
- hotel (IBM iDataPlex at U Chicago)
- india (IBM iDataPlex at IU)
- sierra (IBM iDataPlex at SDSC)
Running Metlab compiled executables for high throughput computing.
Scale of UseThousands of cores for periods of days.
Project Timeline
- Submitted
- 07/02/2013 - 16:24