Hospital visits. Geil et al. (1997) fit a negative binomial random effects model to explain the number of hospital visits using panel data on 5180 individuals drawn from 8 waves of the GSOEP from 1984-1994. The 1990, 1991, and 1993 waves were excluded because they did not provide information on hospitalization. The individuals were between the ages of 25 to 64 and excluded children, students, and retired people. This is an unbalanced panel of 30,590 observations. The explanatory variables are described in Table I of Geil et al. (1997). These include, age, age-squared, age-cubed, dummy for male, dummy for private insurance, dummy for private insurance with copayment obligation, dummy for public insurance, dummy for voluntary public insurance, dummy for family public insurance, dummy for public insurance company legally obliged to accept all risks, dummy for public insurance with voluntary additional coverage through a private scheme, dummy for chronic conditions, dummy for handicapped, monthly net income, dummy for living outside city center, dummy for married, dummy for at least secondary education, dummy for university or technical college, dummy for passing vocational training, dummy for working in a health-related field, dummy for being in the labor force, dummy for blue collar, dummy for white collar, dummy for civil servant, dummy for self-employed, dummy for part-time, dummy for a non-German from Western countries, dummy for other non-German nationals, dummy for children below the age of 16 in the household. The data set can be downloaded from the Journal of Applied Econometrics archive web site: (http://qed.econ.queensu.ca/jae/).

(a) Replicate the descriptive statistics given in Table II of Geil et al. (1997)?

(b) Run the Negative Binomial random effects regressions for males and females given in Table III of Geil et al. (1997). Compare the estimates and their significance for males versus females with regard to their hospital visits in Germany?