Abstract
Zero inflation and over-dispersion issues can significantly affect the predicted probabilities as well as lead to unreliable estimations in count data models. This paper investigates whether considering this issue for German Socioeconomic Panel (1984–1995), used by Riphahn et al. (2003), provides any evidence of misspecification in their estimated models for adverse selection and moral hazard effects in health demand market. The paper has the following contributions: first, it shows that estimated parameters for adverse selection and moral hazard effects are sensitive to the model choice; second, the random effects panel data as well as standard pooled data models do not provide reliable estimates for health care demand (doctor visits); third, it shows that by appropriately accounting for zero inflation and over-dispersion there is no evidence of adverse selection behaviour and that moral hazard plays a positive and significant role for visiting more doctors. These results are robust for both males and females’ subsamples as well as for the full data sample.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
As will be discussed in following sections, recently various extensions of zero-inflated models have been emphasised by Harris et al. (2014) and are incorporated in STATA. More details about some of these models are discussed in Hilbe (2011). While, STATA is not able to provide panel data estimates for zero-inflated models, LIMDEP is able to estimate fixed effect and random effect models in this context.
- 2.
- 3.
Vuong test is a likelihood ratio based test for selecting a specific model among non-nested models.
- 4.
- 5.
- 6.
Since the predicted values might not be integers, we convert them to the closest integer.
- 7.
Codes for Bivariate NB2 model are found in Cameron, C., and Trivedi, P. (2013) (see page 336).
- 8.
For the estimating Bivariate NB2 by ML, initial values we find by estimating non-linear seemingly unrelated regression (NLSUR) and assuming initial value for α equal to 2. For first stage, correlation between doctor visits and hospital visits for males and females are estimated as 0.125 and 0.078, respectively.
- 9.
Stata gives this probability using the command: predict f0, pr(0).
- 10.
Stata gives this probability using the command: predict fk, pr(k).
References
Ainsworth. (2007). Zero-inflated spatial models: Web supplement. Lecture note. http://people.math.sfu.ca/~lmainswo/
Amponsah, S. (2013). Adverse selection, moral hazard, and income effect in health insurance: The case of Ghana. Bulletin of Political Economy, Tokyo International University, 14, 35.
Asante, F., & Aikins, M. (2008). Does the NHIS cover the poor. Accra: Danida Health Sector Support Office.
Bago d’Uva, T., & Jones, A. (2009). Health care utilization in Europe: New evidence from the ECHP. Journal of Health Economics, 28, 265–279.
Bajari, P., Hong, H., & Khwaja, A. (2011). A semiparametric analysis of adverse selection and moral hazard in health insurance contracts. Working Paper.
Bajari, P., Dalton, C., Hong, H., & Khwaja, A. (2014). Moral hazard, adverse selection, and health expenditures: A semiparametric analysis. The Rand Journal of Economics, 45(4), 747–763.
Bundorf, M. K., Herring, B., & Pauly, M. (2005). Health risk, income, and employment-based health insurance. NBER Working Paper No. 11677.
Cameron, C., Trivedi, T., Milne, F., & Piggott, J. (1988). A microeconometric model of the demand for health care and health insurance in Australia. Review of Economic Studies, Oxford University Press, 55(1), 85–106.
Cameron, C., & Trivedi, T. (2013). Regression analysis of count data. NewYork: Cambridge University Press.
Cardon, J. H., & Handel, I. (2001). Asymmetric information in health insurance: Evidence from the national medical expenditure survey. The Rand Journal of Economics, 32(3), 408–427.
Chiappori, P., & Salanie, B. (2000). Testing for asymmetric information in insurance markets. Journal of Political Economy, 108(1), 56–78.
Cohen, A. C. (1960). Estimation in a truncated Poisson distribution when zeros and ones are missing. Journal of the American Statistical Association, 55, 342–348.
Famoye, F. (1995). Generalized binomial regression model. Biometrical Journal, 37, 581–594.
Famoye, F., & Singh, K. P. (2006). Zero-inflated generalized poisson regression model with an application to domestic violence data. Journal of Data Science, 4, 117–130.
Geil, P., Million, A., Rotte, R., & Zimmermann, K. F. (1997). Economic incentives and hospitalization in Germany. Journal of Applied Econometrics, 12(3), 295–311.
Greene, W. H. (1994). Some accounting for excess zeros and sample selection in poisson and negative binomial regression models. (Working Paper EC-94-10): Department of Economics, New York University.
Greene, W. (2008). Functional forms for the negative binomial model for count data. Economic Letters, 99(3), 585–590.
Gupta, P. L., Gupta, R. C., & Tripathi, R. C. (1996). Analysis of zero-adjusted count data. Computational Statistics and Data Analysis, 23, 207–218.
Harris, T., Hilb, J., & Hardin, J. (2014). Modeling count data with generalized distributions. The Stata Journal, 14(3), 562–579.
Hilbe, J. (2011). Negative binomial regression (2nd ed.). NewYork: Cambridge University Press.
Hilbe, J. (2014). Modeling count data. Cambridge University Press.
Irwin, J. O. (1968). The Generalized Waring Distribution Applied to Accident Theory. Journal of the Royal Statistical Society. Series A, 131(2), 205–225
Johnson, N. L., Kemp, A. W., & Kotz, S. (2005). Univariate discrete distributions (3rd ed.). New York: Wiley.
Keane, M., & Stavrunova, O. (2016). Adverse selection, moral hazard and the demand for Medigap insurance. Journal of Econometrics, 190(1), 62–78.
Kirigia, J. M., Sambo, L. G., Nganda, B., Mwabu, G. M., Chatora, R., & Mwase, T. (2005). Determinants of health insurance ownership among South African women. BMC Health Services Research, 5(1), 1.
Kuhnert, P. M., Martin, T. G., Mengersen, K., & Possingham, H. P. (2005). Assessing the impacts of grazing levels on bird density in woodland habitat: A Bayesian approach using expert opinion. Environmetrics, 16, 717–747.
Li, T., Trivedi, P. K., & Guo, J. (2003). Modeling response bias in count: A structural approach with an application to the national crime victimization survey data. Sociological Methods and Research, 31, 415–545.
Martin, T. G., Wintle, B. A., Rhodes, J. R., Kuhnert, P. M., Field, S. A., Low-Choy, S. J., Tyre, A. J., & Possingham, H. P. (2005). Zero tolerance ecology: Improving ecological inference by modelling the source of zero observations. Ecology Letters, 8, 1235–1246.
Marvasti, A. (2014). An estimation of the demand and supply for physician services using a panel data. Economic Modelling, 43, 279–286.
Melkersson, M., & Rooth, D. (2000). Modeling female fertility using inflated count data models. Journal of Population Economics, 13, 189–203.
Neal, S., & Gaher, R. (2006). Risk for marijuana-related problems among college students: An application of zero-inflated negative binomial regression. American Journal of Drug and Alcohol Abuse, 32, 41–53.
Pauly, M. V. (1968). The Economics of Moral Hazard. American Economic Review, 58(3), 531–537.
Powell, D. (2014). Estimation of quantile treatment effects in the presence of covariates. Unpublished manuscript.
Powell, D., & Goldman, D.. (2016). Disentangling Moral Hazard and Adverse Selection in Private Health Insurance, NBER Working Paper No. 21858.
Ridout, M., Demetrio, C., & Hinde, J. (1998). ‘Models for count data with many zeros’. International biometric conference. Cape Town, 1998.
Riphahn, T., Wambach, A., & Million, A. (2003). Incentive effects in the demand for health care: A behavioral panel count data estimation. Journal of Applied Economics, 18, 387–405.
Rothschild, M., & Stiglitz, J. E. (1976). Equilibrium in competitive insurance markets: An essay on the economics of imperfect information. Quarterly Journal of Economics, 90(4), 629–649.
Vuong, Q. (1989). Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica, 57, 307–333.
Wolfe, J. R., & Goddeeris, J. H. (1991). Adverse selection, moral hazard, and wealth effects in the Medigap insurance market. Journal of Health Economics, 10(4), 433–459.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendices
Additional Tables
Model Specifications
To account for over-dispersion and deviations from E(Y i ) = V(Y i ) = μ i in the Poisson distribution, a new distribution is obtained by adding an individual unobserved effect (u i ) to the log of the mean of the Poisson model, ln(mean i ) = ln(μ i ) + ln(u i ). Thus, by defining different distributions for u i , new versions of the Poisson distribution are created. Table 21 presents a list of those distributions, known as standard distributions in this paper, with their variances. A Gamma distribution for u i , for example, gives a Negative Binomial 2 (NB2) distribution with mean μ i and conditional variance μ i + αμ i 2, with the constant parameter α controlling for heterogeneity or dispersion among individuals. The additional parameter p in the Power Negative Binomial (NB-P) distribution, introduced by Greene (2008), provides NB1 or NB2 distributions when p = 1 or p = 2 , respectively. Also, the Heterogeneous NB2 model allows the heterogeneity explained by α in the NB2 distribution to be a function of the individual’s characteristics (z i ), α = exp(z i γ). Thus, α can vary among individuals. In special case, where ϕ → 1, the variance of Famoye’s (1995) distribution approaches to that of the NB. The Waring Negative Binomial distribution introduce by Irwin (1968) converges to NB if \( k\to \frac{1}{\alpha },\rho \to \infty \). Also, if δ = 0, the GP distribution reduces to the usual Poisson distribution with parameter θ i . (See Hilbe (2011, 2014) for more details)
1.1 Zero-inflated Count Models
As Hilbe (2011) discuss, the framework of zero-inflated models are based on separating zero outcomes and positive ones. The probability of zero outcomes results from the group of individuals who are not the subject of an event (Q(0) for those who do not have physician to visit), and those who are the subject of the event but with zero outcome (P(0) for those who do not visit their physicians). The two part of the model is written as:
The probability of a zero outcome for the system is given byFootnote 8:
And the probability of a nonzero count isFootnote 9:
A Probit or logit model estimates Q(0) while one of the standared models in Table (1) estimates Pr(k), k = 0, 1, …n. The mixture model have more power in explaining over-dispersion in the data (see also Hilbe and Greene (2008)).
Table 22 presents different zero-inflated distributions that are used in the next sections for the purpose of estimation and comparison.
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG
About this paper
Cite this paper
Kavand, H., Voia, M. (2018). Estimation of Health Care Demand and its Implication on Income Effects of Individuals. In: Greene, W., Khalaf, L., Makdissi, P., Sickles, R., Veall, M., Voia, MC. (eds) Productivity and Inequality. NAPW 2016. Springer Proceedings in Business and Economics. Springer, Cham. https://doi.org/10.1007/978-3-319-68678-3_13
Download citation
DOI: https://doi.org/10.1007/978-3-319-68678-3_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-68677-6
Online ISBN: 978-3-319-68678-3
eBook Packages: Business and ManagementBusiness and Management (R0)