Health expenditure estimation and functional form: applications of the generalized gamma and extended estimating equations models

Abstract

Health-care expenditure regressions are used in a wide variety of economic analyses including risk adjustment and program and treatment evaluations. Recent articles demonstrated that generalized gamma models (GGMs) and extended estimating equations (EEE) models provide flexible approaches to deal with a variety of data problems encountered in expenditure estimation. To date there have been few empirical applications of these models to expenditures. We use data from the US Medical Expenditure Panel Survey to compare the bias, predictive accuracy, and marginal effects of GGM and EEE models with other commonly used regression models in a cross-validation study design. Health-care expenditure distributions vary in the degree of heteroskedasticity, skewness, and kurtosis by type of service and population. To examine the ability of estimators to address a range of data problems, we estimate models of total health expenditures and prescription drug expenditures for two populations, the elderly and privately insured adults. Our findings illustrate the need for researchers to examine their assumptions about link functions: the appropriate link function varies across our four distributions. The EEE model, which has a flexible link function, is a robust estimator that performs as well, or better, than the other models in each distribution.

The influence of obesity and overweight on medical costs: a panel data perspective

“…Notice this alternative model is much poorer in terms of goodness of fit and predictive capacity. 21 This finding is compatible with Hill and Miller (2010) that sustain that OLS of log (costs) models tend to perform poorly in terms of their bias and predictive accuracy, making the GLM more attractive for the second part of the two-part model. Cawley and Meyerhoefer (2012) follow the same strategy when estimating their models.…”

“…6 A finding that emerges from the literature that compares the performance of several models for positive expenditures in terms of consistency and precision (Manning and Mullahy, 2001;Buntin and Zaslavsky, 2004;Manning et al, 2005;Baser, 2007;Hill and Miller, 2010) is that no one method dominates the other and there are important trade-offs in terms of precision and bias, mainly when different subgroups of population or types of medical costs are analysed (Hill and Miller, 2010;Jones, 2010). Notwithstanding, Mihaylova et al (2011) literature review confirms that 2PM models perform better.…”

Extended Office Hours and Health Care Expenditures: A National Study

“…In addition to using logarithm of expenditures as the dependent variable (which excludes those with no expenditures), we also implemented 2 generalized linear model analyses using a logarithm link and a Poisson distribution. 17 A fi nal analysis also included smoking status (smoker or not) and body mass index category (<20, 20 to <25, 25 to <30, and ≥30 kg/m 2 ), to adjust for health risk behaviors. These health risks were not included in the main analyses because 4% of the analytic sample had missing data for these variables.…”

The Medical Care Costs of Obesity: An Instrumental Variables Approach

“…We rejected the null hypothesis that the decile coefficients are jointly equal to zero for only three out of sixteen models, which suggests that the gamma model with log link is broadly appropriate. In addition, Hill and Miller (2009) found that this specification performed relatively well on the 1996-2003 sample of non-elderly MEPS respondents with private insurance.…”