Elsevier

Health Policy

Volume 120, Issue 9, September 2016, Pages 1095-1099
Health Policy

Copayments and physicians visits: A panel data study of Swedish regions 2003–2012

https://doi.org/10.1016/j.healthpol.2016.07.010Get rights and content

Highlights

  • The copayment for physician visits has no effect of significance on visits.

  • We estimate the “zero effect” of copayments with very high precision.

  • The effect of copayments seems to be vary by health care system and context.

Abstract

Objective

This paper analyzes how primary care physician visits are affected by the level of copayment in Sweden.

Data source

We use data between the years 2003–2012 from 21 Swedish health care regions that have the mandate to set their own level of copayment. The copayment per visit varies between €10 and €20 for these years and regions.

Study design

Our strategy to identify the causal effect and deal with unobserved endogeneity of price changes on physician visits is based on a panel data model using fixed effects to control for region and time and regional-variation in time trends.

Principal finding

We cannot reject that the copayment has no statistical or economic effect of significance, and we estimate the “zero effect” with very high precision.

Conclusion

In a setting with sub-national regions with autonomy to set co-payments the results points to that the copayment is not an important predictor for the number of health care visits. The result is in line with some previous studies on European data where the range of copayments used tends to be relatively low.

Introduction

Most OECD countries spend around 10% of GDP on health care but due to an increasingly elderly population coupled with the constant influx of new costly medical technology this number is believed to increase further over time [1]. One common suggestion to control costs and to fund an increasingly costly health care system is to increase the use of patient charges (copayments). The use of copayments is relatively common in several public sector areas to fund services, also in the health care sector. Among OECD countries the share of health care expenditures paid directly by households is about 20% [2].

If there is moral hazard, higher copayments will reduce over-utilization, which reduces costs and the welfare loss [3]. For tax-financed health care systems (e.g., the UK and Sweden) it may also be a more efficient means of raising budget income compared to general taxation, given that the latter is associated with significant welfare costs due to the distortionary tax effects [4]. But there are also some downsides of using copayments. If it leads to a reduction in demand this may worsen population health and increase costs even more in the long run. And, if demand drops more in low-income groups, as some previous literature indicates [5], it will increase income-related health inequalities in utilization. Additionally, if patients’ demand is completely price inelastic (insensitive) copayment increases will not lead to a reduction in health care utilization and there are no efficiency improvements to be made. There may also be offsetting effects, as found by Chandra et al. [6].

In this paper we estimate the effect of changes in the copayment level on primary care physician visits using Swedish regional data between 2003 and 2012. The 21 Swedish regions responsible for financing and providing health care have autonomy in setting the level of co-payment and there is significant variation between regions and over time.

Most of the previous empirical literature on copayments and primary care has been conducted in the US [7], and the most influential study is still the RAND health insurance experiment conducted between 1974 and 1982 showing a price elasticity of demand in the range of −0.1 to −0.2 [8], [9], [10]. However, the RAND experiment is very dated and its usefulness for policy makers in modern times and non-US contexts is unclear. In the US context recent quasi-experimental studies have found price elasticities close to the RAND estimates [6]. From a European context there are only a few studies with a quasi-experimental approach. For example, a study using a difference-in-difference approach evaluating an increase in patient charges in Belgium during the 1990s estimated price elasticities of demand between −0.03 and −0.13 [11]. And, two studies with a similar empirical approach in a French setting found no price elasticity at all (general population) for office General Practitioner (GP) visits and a very small effect for GP home visits among a general population [12] and no effect at all in a sub-set of the poorest households [13]. Finally, the 2004 reform in Germany that introduced a €10 copayment (for the first visit for each running three-month period) has been evaluated with the findings that there was no effect on utilization at all [14].

European studies tend to find very low or no price elasticity of demand for visits to primary care physicians, perhaps explained by the low European levels of copayments in comparison to copayments reported in some studies in the United States. The typical office-based out-of-pocket expenditure in the US is about $30, whereas in some of the larger European countries (e.g., the UK, Germany, and Italy) outpatient primary care is free of any charges at the point of care [1], [15].

This paper contributes to the literature by focusing on many different copayment changes over a more extensive time period, rather than evaluating a single price reform affecting a full country or region. In contrast to many of the previous studies we can thus detect not only short-term changes but also examine if they are sustained over time. A further benefit is that we evaluate marginal copayment changes, rather than evaluating e.g., a reform moving from a zero copayment to some positive level (such as the German reform in 2004), which may cause very different behavioral responses.

Section snippets

Institutional background

The main actors for funding and provision of health care in Sweden are the health care regions (county councils). The councils set the county council income tax rates in order to fund health care delivery and it varies between approx. 10.5 and 12% in the county councils. Within each county council primary care is generally provided at different primary care centers with multiple physicians and nurses employed within the same center. The default is that an individual is listed with the

Data

We have collected data from all 21 regional governments in Sweden between 2003 and 2012 on primary care physician visits, copayments, and a set of additional control variables; in total 210 region/year observations [16]. The data is aggregated at the regional level, thus ethical review of this study by the ethics committee was not necessary.

Descriptive statistics of all variables are shown in Table 1. Our dependent variable, general practitioner visits per capita, varies between 0.95 and 1.93

Empirical approach

The empirical challenge to overcome in estimating the effects of prices on visits is that we cannot observe what the number of visits would have been in a region if the price was different. We try to overcome this missing counterfactual problem using region fixed effects estimations. Thus, we are using each region when facing another price as that regions missing counterfactual. The main model that we estimate is:log(yjt)=β0Copaymentjt+γt+δr+δr×Time+θ0xjt+εjt,where log(yjt) is the log of the

Results

In Table 2 we present the results from our five different model versions (Model 1–5) as outlined above.

As seen, the choice of modelling specification (Model 1–5) does not affect the interpretation of our results; we cannot reject the null hypothesis of no impact of the level of copayment on the log of visits. The estimates are estimated with good precision; the coefficients are zero up to the fourth decimal in all models and the standard errors are very small. Even though there may be an

Discussion

We find that changes in the level of copayments do not affect the number of primary care physician visits, which is in contrast to earlier evidence in the US [8], [9], [19]. An important difference in our study is the relatively low dollar values of copayments. A number of European studies conducted in settings with relatively low absolute copayment levels (France [12], Belgium [11]) also documented a very small or absence of any effect on utilization from changes in the copayment. Thus, a

Conclusion

Our study of Swedish sub-national regions, which have autonomy to set co-payments, finds that co-payments are not a significant predictor of primary care physician visits. Our result is consistent with previous European studies where patient copayments tend to be relatively low.

Conflicts of interest

The authors have no conflicts of interest to report.

Financial support

Financial support from the Swedish Research Council for Health, Working Life and Welfare is gratefully acknowledged (grant number 2013-0573).

Acknowledgements

We are very grateful for the comments and suggestions by anonymous reviewers.

References (19)

There are more references available in the full text version of this article.

Cited by (14)

  • Equity in access to zero-fees and low-cost Primary Health Care in Aotearoa New Zealand: Results from repeated waves of the New Zealand Health Survey, 1996-2016

    2020, Health Policy
    Citation Excerpt :

    One of the arguments for user co-payments is that services that have no fee attached are not always used appropriately. A systematic review found that co-payments were associated with a lower rate of consultations [36], although this is not consistently found in studies examining high income populations with relatively low co-payments [37]. There is a balance between appropriate use of limited resources and ensuring that patients who need health care can access it.

  • Policy induced increases in private healthcare financing provide short-term relief of total healthcare expenditure growth: Evidence from OECD countries

    2019, European Journal of Political Economy
    Citation Excerpt :

    At the individual country level, hospital-financing reform is not found to have an effect on total HCE in Switzerland (Braendle and Colombier, 2016). Likewise, variation in co-payments in Sweden has no effect on the number physician visits (Jakobsen and Svenson, 2016). Colombo and Tapay (2004) conclude that increased opportunity to take out private health insurance generally increases total HCE relative to GDP.

View all citing articles on Scopus
View full text