|
|
|
MULTIVARIATE SARMANOV COUNT DATA MODELS
(Click above to download a pdf copy of the paper)
- Abstract: I develop two flexible models of
multivariate, count data regression that overcome several difficulties to
extend Poisson regressions beyond the univariate case: i) account for
both over and underdispersion, ii) allow for correlations of any sign
among counts, iii) correlation and dispersion depend on different
parameters, and iv) estimation is computationally feasible. I address
whether the pricing strategies of competing duopolists in the early U.S.
cellular telephone industry are strategic complements or substitutes. I show
that a Sarmanov model with double Poisson marginals outperforms a count data
model based on a multivariate renewal process with gamma distributed arrival
times.
- Publication: Previously circulated as CEPR DP No. 7463.
- JEL: C16, C35, L11.
- First version: June 2008.
- Current version: September 2009.
- Funding: None.
- Seminars: None.
- Conferences: Texas Camp Econometrics XIV (2009).
- Presentation: PDF.
- Media Citations: None.
|
|
|
