Per‐period Co‐payments and the Demand for Health Care: Evidence From Survey and Claims Data

Abstract

When health insurance reforms involve non-linear price schedules tied to payment periods (for example, fees levied by quarter or year), the empirical analysis of its effects has to take the within-period time structure of incentives into account. The analysis is further complicated when demand data are obtained from a survey in which the reporting period does not coincide with the payment period. We illustrate these issues using as an example a health care reform in Germany that imposed a per-quarter fee of €10 for doctor visits and additionally set an out-of-pocket maximum. This co-payment structure results in an effective 'spot' price for a doctor visit that decreases over time within each payment period. Taking this variation into account, we find a substantial reform effect-especially so for young adults. Overall, the number of doctor visits decreased by around 9% in the young population. The probability of visiting a physician in any given quarter decreased by around 4 to 8 percentage points.

Påvirker egenandeler ungdoms legebesøk?

“…Schreyögg og Grabka (2010) fant ingen effekt på antall legekonsultasjoner blant de som ble berørt da Tyskland i 2004 innførte en egenandel på 10 Euro for første konsultasjon per kvartal. Farbmacher og Winter (2013) mener at dette resultatet er misvisende fordi man ikke har tatt hensyn til den spesielle ikke-lineaere prisefThis article is downloaded from www.idunn.no. © 2016 Author(s).…”

Heterogeneous Effects of a Nonlinear Price Schedule for Outpatient Care

“…; Schreyögg and Grabka, ). Farbmacher and Winter () address measurement problems and nonlinearities that arise when using survey data to study the effects of a quarterly co‐payment, but restrict the analysis to homogeneous effects. Rückert et al.…”

An Econometric Model of Healthcare Demand With Nonlinear Pricing

“…The regression analyses control for these factors, but the differences still raise the question of the comparability of the two groups, and hence of the validity of the parallel trends assumption underlying the DiD identification strategy. A formal test of this assumption was conducted by Farbmacher and Winter (), using the same kind of data from the SOEP, and they could not reject the null hypothesis of parallel trends during the pre‐treatment years.…”

“…Because interviews typically do not take place at the end of a calendar quarter, the reporting period overlaps with two calendar quarters. As noted by Farbmacher and Winter (), the standard models are invalid in this case. By contrast, our model prescribes a method to deal with mismatch in a theory‐consistent way, owing to the derivation of the dynamic hurdle model from an underlying stochastic process, which the standard hurdle model lacks.…”

“…Farbmacher and Winter () recommend to base estimation on the subset of individuals who were interviewed close to the end of a calendar quarter. Even if one can treat the interview day as random, this approach takes a heavy toll in terms of sample size and thus precision.…”