Risk selection and heterogeneous preferences in health insurance markets with a public option

Transcription

1 Risk selection and heterogeneous preferences in health insurance markets with a public option Maria Polyakova December 27, 2015 Abstract Conventional wisdom suggests that if private health insurance plans compete alongside a public option, they may endanger the latter s financial stability by selecting good risks. In this paper, I argue that the degree of such risk cherry-picking depends crucially on the extent and nature of differentiation between the public and private options, as well as on the variance of consumer preferences. This paper explores an example of such differentiated insurance market empirically using the institutional setting in Germany. I propose a regression discontinuity approach to disentangle adverse selection and moral hazard between private insurance providers and the public option. The empirical strategy takes advantage of the German federal regulation that mandates individuals with income below an annually set threshold to enroll into the public system. Against conventional wisdom, I do not find compelling support for extensive cream-skimming by private insurers. Using a discrete choice model of demand for private insurance, I explore heterogeneous non-pecuniary preferences and long-term contract design of private insurers as potential explanations for this optimistic result about insurance design. JEL classification numbers: D12, I13, I18, G22, H44 Keywords: Health Insurance, Public Option, Adverse Selection, Individual Mandate Department of Health Research and Policy, Stanford University, and NBER, mpolyak@stanford.edu. This paper is a revised chapter of my MIT dissertation. First draft: May I am indebted to Amy Finkelstein and Stephen Ryan for their guidance throughout this project. I also thank the participants at the MIT Public Finance and Industrial Organization lunches, 15th IZA European Summer School in Labor Economics, MEA Seminar at the Max Planck Institute for Social Law and Social Policy, and Stanford HRP for their feedback. Data for this project - the Scientific Use Files of the German Socio-Economic Panel - were provided by DIW Berlin and Cornell Department of Policy Analysis and Management, which I gratefully acknowledge. 1

2 1 Introduction The ubiquitous feature of health insurance markets is that insurers costs depend on who their enrollees are and how they behave. This characteristic of selection markets raises concerns about the feasibility of efficiency-improving competition, and has served as a traditional rational for extensive government intervention in health insurance. Increasingly, public policies in health insurance attempt to strike a balance between selection and competitive efficiency by reorganizing purely public or purely private insurance systems into different mixtures of the two. A central question in such arrangements, where a private health insurance system exists in parallel to a public one, is whether private insurers may harm the public system by disproportinately enrolling good risks. This concern has lead policy-makers to set up elaborate risk adjustment schemes to counterveil the risk selection incentives of private insurers. The debate about the efficient design of mixed public-private insurance environments is not settled and has recently gained new momentum in the academic and policy discussion in light of extensive reforms of the health insurance landscape as part of the 2010 Affordable Care Act in the US. 1 Similar discussions have played a key role in the analysis of Medicare and Medicaid public programs that are both experiencing growing enrollment in private managed care plans that co-exist with the fee-for-service public option. Several European insurance systems have seen increasing discussions about the role of private health insurers in traditional social insurance domains. This paper offers empirical insights into the workings of an insurance system where private insurers compete alongside the public option and are allowed to fully underwrite risk. I use a fuzzy regression discontinuity design to decompose selection and moral hazard across the two systems. Surprisingly, I find limited evidence of risk selection across the two systems. I discuss the role of heterogeneous preferences for more convenience in healthcare consumption, as well as the long-term structure of private health insurance contracts as two potential forces that countervail the selection incentives. I estimate a discrete choice model of preference for private health insurance that empirically corraborates the presence of these two forces. My empirical setting is a unique institutional environment in Germany. Utilizing the German empirical setting offers several advantages. First, the discontinuity in the rule that determines access to private insurance allows for an effective way of separating adverse selection and moral hazard. Disentangling the latter two effects is a well known challenge to documenting selection between the private and of public insurance systems. Second, the na- 1 See, for example, Halpin and Harbage (2010) and Washington Post for the discussion of the public option as part of ACA Health Insurance Exchanges. 2

3 ture of differences between the public and the private insurance in Germany allows evaluating the role of nonpecuniary preferences in the choice of health insurance and healthcare. There are three key institutional features that characterize the German market. First, private and public insurers follow different pricing regimes. 2 In the so-called statutory system, there is guaranteed issue and premiums are equal to a percentage of pre-tax income set by the regulator at levels that would ensure the solvency of the system. Family members not in the labor force are covered at no extra charge in the statutory system. Statutory insurance is run by independent non-profit sickness funds and have to be self-solvent, as they do not directly depend on financing from the federal or state budgets. Private insurers, on the other hand, are mostly for-profit corporations, can reject enrollment, and are allowed to carry out practically unrestricted underwriting of individual risk (including family members) at the time of enrollment. Private insurers use long-term rather than annual contracts - the contracts are life-long and premium underwriting principles are similar to annuities. Second, public insurers offer very low, if any, levels of consumer cost-sharing. Private insurers typically offer contracts with higher cost-sharing levels, 3 but more comfortable hospital facilities, easier access to star physicians, and, anecdotally, shorter appointment wait times. Finally, the market is strongly influenced by a policy that mandates enrollment in the statutory system for all employees with income below an annually set threshold of about 50,000 USD. I extensively utilize this enrollment mandate in my empirical analysis. I start my investigation of cream-skimming in the German system with non-parametric evidence. I compare the demographic and health-related characteristics of the public insurance enrollees around the income threshold that mandates enrollment in the public system. I find no meaningful differences in individual characteristics to the left and to the right of the income cutoff, even though to the right of the cutoff, about 25% (at the cutoff) to 50% (overall) of individuals leave the public system. Comparing probabilities of chronic diagnoses, I find that individuals with private insurance are less likely to report having diabetes; however, there are no meaningful differences in incidence across seven other chronic conditions. Proceeding to more formal analysis, I compare the average healthcare utilization between the public and the private system and find that enrollees in the private system have lower utilization. This evidence, however, faces the classic challenge of the need to distinguish 2 Public insurance here refers to the system of sickness funds that are heavily regulated and can be considered as a unity for the purposes of analyzing selection on the extensive margin. 3 In addition to familiar cost-sharing methods such as deductibles and co-insurance, private insurers in Germany use a different way of combating moral hazard. Typically, individuals that pay for smaller expenses out of pocket and do not file claims are refunded a substantial fraction of annual premiums. 3

4 selection from moral hazard. Put differently, individuals that end up insured in the private system are not a randomly assigned group and moreover, being in the private system may lead to a change in one s healthcare consumption. Thus, merely comparing healthcare utilization of the enrollees in the public and in the private systems will always confound the presence of the causal effect of private insurance ( moral hazard ) and the ex ante selection or a cream-skimming effect. In this paper, I address the challenge of distinguishing selection from moral hazard by using a fuzzy regression discontinuity design. The idea of this identification strategy is to estimate the moral hazard effect using an income-based mandate for enrollment into the statutory system, and then calculate the selection effect as the residual that explains the difference between the OLS estimates and the moral hazard estimates. This empirical strategy suggests that private insurers enroll individuals that are likely to incur more physician visits, while having private insurance causes individuals to significantly reduce the number of visits. My estimates cannot reject a reverse effect on inpatient admissions. The estimates are imprecise, but cast doubt on the prior that private insurers extensively cherry-pick low healthcare utilizers that would have likely been good risks in the public system. These findings may appear suprising given that private insurers are allowed to fully underwrite risks and reject enrollment. I discuss two possible (albeit certainly non-exhaustive) explanations for this result. First explanation is the presence of heterogeneous preferences for private insurance that are uncorrelated with health risk. Private insurance allows for higher cost-sharing and may thus be attractive to less risk-averse or less liquidity-constrained individuals. Moreover, anecdotally, private insurance is viewed as a luxury good that provides better service, although does not necessarily lead to better medical outcomes. I use a discrete choice model of demand for private health insurance to test for the presence of such preferences empirically. I find that individuals with higher income are more likely to choose private insurance. Moreover, conditional on income, individuals that employ household help, which I interpret as a reasonable proxy for valuing convenience and service, are also much more likely to opt out of the statutory system. Such preferences for convenience in healthcare consumption are not typically central in the literature on insurance contracts that are viewed purely as financial instruments. At the same time, the presence of convenience preferences may imply that plan features such as wait times and location of in-network physicians and hospitals may be the key drivers of individual choices of insurance. The presence of such non-pecuniary taste heterogeneity also introduces opportunities for horizontal differentiation across insurance plans that may help insurers soften price competition. The policy impli- 4

5 cation of these results, which is applicable beyond the specifics of the German institutional setting, is that allowing private plans that exist in parallel to a public option to provide products that are sufficiently horizontally differentiated from the public option, softens the selection concerns at the extensive margin between the two systems. The second hypothesis concerns the supply-side of the market. I argue that the lack of evident differences in risks across the two systems may be the outcome of incentives created by dynamic contracts of the private insurers. The annuity structure of these contracts implies that beneficiaries pay in equal monthly installments their expected lifetime spending on healthcare. 4 The insurers assess this expectation at the time of individual s enrollment and are not allowed by the regulator to re-classify risk or to drop coverage in response to information about the individual s health being revealed over time. Consequently, individuals have a strong incentive to enroll into the private system as early as possible in their lifetime to freeze their health risk at a point in time at which both the individual and the insurer have only very noisy information about individual-specific expected risks. Thus, in many cases, private insurers are likely to have relatively limited scope for underwriting and creamskimming. Indeed, this hypothesis is strongly supported by the existence of a market for options on private insurance contracts. Individuals that are not yet eligible to enroll because their income is too low, but expect to have higher income and be able to enroll with a private insurer in the future, can buy an option contract that freezes their health underwriting at the time of option purchase rather than at the time when they actually buy private coverage. This paper is related to several strands of literature. First, it is related to the broad literature that tests for the presence of adverse selection in insurance markets. Einav, Finkelstein, and Levin (2010) provide a recent survey. One strand of this literature has specifically focused on the question of selection between public and private health insurance. Brown, Duggan, Kuziemko, and Woolston (2014), Cabral, Geruso, and Mahoney (2014), and Newhouse et al. (2014) are recent contributions that explores the selection of risks between the Medicare fee-for-service and the Medicare Advantage program. Fang, Keane, and Silverman (2008) document evidence consistent with the presence of advantageous selection into Medigap, which is a private Medicare add-on insurance. Duggan (2004) studied the efficiency implication of having Medicaid provided by private insurers. A related strand of literature studies selection within the private markets that compete alongside a public option. Lustig (2011) studies the interaction of adverse selection and imperfect competition in Medicare 4 De facto, premiums do not stay constant over the life cycle, as they adjust to growing costs in the healthcare system overall. 5

6 Advantage. Kuziemko, Meckel, and Rossin-Slater (2013) explore whether health insurance companies compete on risk in the context of Medicaid managed care market. Bauhoff (2012) conducts an audit study in Germany, documenting how insurers within the statutory system screen risks. Second, it is related to the literature that has explored the role of heterogeneous preference for selection in insurance markets. Cutler, Finkelstein, and McGarry (2008) discuss the role of heterogeneous preferences in determining the degree and direction of selection in health insurance. The idea is that individual preferences, such as the degree of risk aversion, may reverse the relationship between the risk type and the selected level of coverage. Only few papers have explored the sources of heterogeneous preferences in health insurance empirically. Geruso (2013) focuses on the theme of heterogeneous preferences in employer-provided health insurance and finds that older individuals enroll in more comprehensive plans than younger individuals with the same healthcare expenditure risk. Ericson and Starc (forthcoming) study the implications of age-related heterogeneity in the context of the Massachusetts Health Insurance Exchange. Finally, this paper is closely related to the literature that has specifically studied the German health insurance system. Nuscheler and Knaus (2005) explore the issue of risk selection among different sickness funds within the statutory system, arguing that the observed differences in risk pools are due to the consumers switching costs rather than cherry-picking by plans. Hullegie and Klein (2010) use an RD design similar to the one I exploit in the current paper and also estimates that holding a private insurance policy decreases the number of doctoral visits, doesn t affect the number of hospital stays and improves self-assessed health. Grunow and Nuscheler (2013) study the issue of selection patterns between the private and statutory systems in Germany, arguing that the private insurers are unable to select good risks at the enrollment stage, but manage to return high-risk individuals back to the public system later. Hofmann and Browne (2013) study the German private insurance market from the perspective of the dynamic contracts theory with one-sided commitment, outlining several channels for selection within the private system. Bünnings and Tauchmann (2014) explore what determines the choice of enrolling with the private insurance system. The rest of the paper is structured as follows. Section 2 outlines the key market forces within the German institutional setting and describes the data. Section 3 presents the descriptive evidence on the allocation of risks across the two systems as well as the regression discontinuity analysis. Section 4 explores the potential explanations for the empirical results by documenting heterogeneous preferences for convenience in healthcare consumption and 6

7 the possibility that long-term insurance contracts increase informational uncertainty and thus decrease the opportunity for selection. Section 5 briefly concludes. 2 Data and economic environment 2.1 Institutional Environment Germany spends 11% of its GDP on healthcare, amounting to around $5,000 per capita or roughly $400 billion total in annual healthcare expenditures. 5 A large fraction of these expenditures - 58% - are paid by insurers that are part of the so-called Statutory Health Insurance (henceforth SHI) system. The SHI differs from conventional public coverage, as there are multiple independent non-profit mutual insurance funds operating within the system. The government does not direcly carry the actuarial risk or administer the health insurance plans. Similarly to a traditional public option, however, SHI insurers cannot deny coverage, cannot underwrite risk, and the amount of coverage they provide as well as the premiums they collect are largely dictated by local and federal governments. The next largest payer in the German healthcare sysem (after the out of pocket expenditures) is the private health insurance (henceforth PHI) that covers 8% of total spending. Independent commercial insurers that are part of the PHI system offer individual health insurance packages in a robust non-group market. These insurers are free to decide whether to enroll an individual and enjoy substantial freedom in their decision about the extent of coverage and premiums. There are several key public policies that shape the German health insurance setting. First, there exists an individual mandate policy, according to which employees with income below an annually set regulatory threshold (about $58,000 in 2015) 6 have to enroll in the SHI. Only a selected group of individuals may choose to enroll in the private system - primarily, employees with sufficiently high income, self-employed individuals, and civil servants. 7 Those choosing to forego the public option and enroll in the PHI, are restricted in their ability to return to the public system later. Second, the government sets redistributive premiums for the public option - premiums differ by individual s income, but not by risk. In 2015, enrollees paid 7.3% of their income to SHI. There is a cap to the amount of income that is subject to SHI premium withholding; as a result, individuals that are eligible to choose 5 Statistisches Bundesamt, 2013 data 6 The income threshold in 2015 was EUR 54, Civil servants have a portion (typically 50%) of their healthcare expenditures paid by the government directly, making PHI coverage of the residual expenditures especially economically attractive. 7

8 between the SHI and PHI systems face the highest premiums in the SHI. Third, private insurers are allowed to fully underwrite individual s health risk at the time the individual enrolls with a PHI insurer for the first time; at the same time, the insurers have to offer renewable long-term contracts (similar to annuities) without the re-classification risk. 8 The SHI and PHI plans differ on many dimensions. In addition to the differences in premiums as outlined above, there are often substantial differences in cost-sharing within the plans. While SHI plans typically have low or even just nominal cost-sharing, PHI plans may offer significant deductibles and co-insurance levels, although consumers have a lot of choice in the deisgn of the PHI plans and can almost perfectly trade-off premiums and costsharing. PHI plans require more financial liquidity than SHI plans from their enrollees, as outpatient services are first paid by patients out of pocket and are later reimbursed by the PHI insurer if the patient submit the claim. SHI plans offer more generous family coverage for families with children or spouses that are not in the labor force - the latter two groups are covered at no extra charge under the SHI, while they have to pay full premiums under the PHI. At the same time, PHI plans often provide access to more convenience in healthcare consumption, covering more comfortable rooms for in-patient admissions, providing faster inpatient and outpatient appointments (at least anecdotally), and covering extra fees for seeing star physicians. Physicians can charge much higher fee for service rates to private insurers than they can in the SHI system, which may improve the care provided, but may also induce excess utilization in the PHI system. Given this multitude of differences between the two systems, different types of individuals may be considered good or bad risks by the private insurers and the public system. Consider the SHI. The employees with income above the income threshold, all pay the same fixed premium to the SHI, as their income is above the withholding cap. Thus, the good risks for the SHI are simply those individuals whose healthcare expenditures in a given year are lower than what they pay into the system. Let us call these individuals net payers and the individuals that are expected to spend more on their healthcare than they pay, net receivers. Then, we can define selection in this market as follows. There is adverse selection into the private system if the individuals that opt out of the public system would have been predominantly net payers. There is advantageous selection into the private market 8 About half of the premium costs in both the SHI and PHI systems are paid by the employer as an additional contribution to the employees gross income. SHI premium in 2015 was 14.6% of individual s income in premiums, half of this amount was paid by the employer. If an employee is enrolled with the PHI, employers pay half of the PHI premiums up to a maximum contribution that is equal to their maximum contribution in the SHI system. 8

9 if the individuals that opt out would have been net receivers in the public system. And finally there is no selection if the switchers are a random mix of risks. The conventional intuition suggests that competition alongside private insurers may harm the public option, because private insurers may disproportionally cream-skim net payers out of the public system. In Section 3, I proxy expected healthcare spending by healthcare utilization and test empirically whether the PHI system appears to cherry-pick relatively lower utilizers from the public system. 2.2 Data Throughout the empirical analysis, I use data from years 2004 to 2009 of the German household survey panel SOEP. The raw data records information on 28,693 individuals. The survey offers a collection of answers to a rich set of demographic, employment, and health-related questions for a representative sample of the German population. 9 The survey has multiple questions related to health and health insurance that I utilize in this paper. First, SOEP records whether an individual is enrolled within the statutory or the private health instance system, and whether the individual is the primary enrollee. For those individuals that are enrolled in the PHI, the survey asks for the level of monthly premiums. Second, SOEP contains several indicators related to health that provide a measure of healthcare utilization, as well as a rich set of private information that is not necessarily used for pricing by private insurers. The latter include questions about risk aversion in various domains, information about other insurance products that the household holds (e.g. life insurance), as well several variables that plausibly reveal individuals preferences for conveneince or time value. The baseline analytic sample imposes several restrictions on the raw data. First, I restrict the sample to include individuals of age This age restriction primarily excludes children, students, and retirees, who likely either do not make active decisions about their health insurance, or face a different set of incentives in their choices. Second, I restrict the sample to include individuals working full-time, either as employees or self-employed, with gross income above 400 EUR per month. This excludes those working part-time and civil servants, who may face a different set of incentives. This also excludes individuals that report being unemployed or out of the labor force, as they are typically not making their own insurance choices, being insured either as dependents or through welfare programs. These restrictions leave us with 37,552 individual-year observations on 10,725 unique individuals, 9 For more detailed information on the statistical properties of SOEP panel sample please see 9

10 out of whom 9,454 are employees and 1,271 are self-employed. Table 1 reports the summary statistics for the baseline sample. The individuals in the sample are on average 43 years old, 32 percent female, with average gross monthly income of 3,230 EUR. The average individual has 13 years of schooling and works 44 hours per week. 38% are not married. The survey respondents report visiting a doctor in an outpatient setting about 7 times a year, while only about 10 in a hundred have one inpatient admission. Individuals report being slightly overweight with an average BMI of 26; 33% report being smokers. 17% have high blood pressure, about 4% report asthma, cardiac conditions, depression, or diabetes. In the baseline sample, 14% have private health insurance. This fraction is much lower - at 8% - for the sub-sample of employees, who are only eligible to purchase PHI if their income is high enough. The sample without self-employed individuals has slightly lower average gross income at 3,030 EUR per month; other socio-demographic characteristics and health-related indicators are very close to the overall sample. 3 Empirical evidence: selection and moral hazard 3.1 Descriptive evidence I start my investigation of risk selection between public and private insurers in the German system with model-free evidence. If PHI disproportionately enrolled healthier individuals, we would expect this to be reflected in the compositional changes of SHI demographics, diagnoses, and utilization around the income threshold. For example, suppose PHI only accepted individuals that are younger than 40, then we would expect the average age of SHI enrollees to the right of the threshold to increase disproportionately relative to the average age in the SHI to the left of the threshold. Hence, we are interested in whether there are breaks in the observable demographics of SHI enrollees at the income threshold above which individuals may leave the SHI system and enroll with a private insurer. I compare the average age, the fraction of older enrollees, the average BMI, the fraction of smokers, the fraction of disabled individuals, and the self-reported risk attitude towards health of SHI enrollees around the income eligibility threshold. Figure 1 illustrates these comparisons. To detrend the development of the outcome variables from the relationship with income, the figures plot the residuals from the regression of the outcome variables on income using the sample to the left of the cutoff and calculating out-of-sample residuals to the right of the cutoff. The graphs uncover few differences in individual characteristics to 10

11 the left and to the right of the income cutoff for any of the observable characteristics of the individuals, even though to the right of the cutoff, about 25% (at the cutoff) to 50% (overall) of individuals leave the public health insurance system. Figure 2 reports the outcomes of a similar exercise, looking at the probability of various diagnosis rather than the demographic factors, for eight diagnoses - asthma, cancer, stroke, migraine, depression, diabetes, high blood pressure, and cardiovascular conditions. I compare the probabilities of having these diagnoses for public health insurance enrollees around the income thresholds. Visually, it appears that the fraction of individuals with diabetes increases around the threshold, although the difference is not statistically significant in a linear specification. There are no discernible breaks in the probability of other conditions. Overall, the descriptive evidence thus suggests that individuals that leave the SHI around the income threshold have similar observable characteristics as those that stay. 3.2 Disentangling adverse selection and moral hazard The key challenge for the empirical identification of selection between the two insurance systems is the need to disentangle the ex ante selection into the PHI system from the ex post causal effects of PHI enrollment, or moral hazard. To address this identification challenge, I rely on a combination of OLS and IV estimates. The idea is that an ordinary least squares regression of health care utilization on the indicator of insurance type combines the treatment (i.e. PHI changing individuals utilization) and the selection effect (i.e. PHI selecting or being selected by lower risk individuals). An instrumental variables strategy - in this case based on a fuzzy regression discontinuity design - allows us to estimate the treatment effect, or the moral hazard component. The difference between the OLS and IV estimates should then allow us to capture the selection effect of interest. This approach is similar in spirit to the ideas in Chandra and Staiger (2007); McClellan et al. (1994). I proceed in three steps. First, I estimate an OLS regression that captures both selection and moral hazard. Second, I use a instrumental variables specification to estimate the extent of moral hazard. And third, I compare the two sets of estimates, to quantify the extent of selection by subtracting the instrumental variables coefficients from the OLS results. 11

12 Healthcare utilization and insurance type: OLS I first use an OLS specification to estimate the relationship between the insurance type and the observed healthcare utilization: E[Yit outcome.] = αp HI it + βx it The outcome variable Y outcome is one of the following outcome variables observed in the data: the unconditional annual number of inpatient visits, the number of outpatient visits, the number of inpatient and outpatient visits conditional on having at least one visit, as well as the probability of having at least one inpatient or outpatient visit. P HI is an indicator variable that is equal to one if the individual has private insurance. The set of control covariates X includes age, gender and income. Table 3 reports the results of this regression on the full baseline sample of employees for all six outcome variables. Most coefficients are not different from zero at a 5% confidence level with point estimates close to zeros relative to the mean of the outcome variables in the data. The estimates are more precise for the inpatient admissions outcomes, driven primarily by the negative correlation between having a private insurance and reporting fewer hospital stays conditional on having had at least one. For this outcome variable, individuals are likely to have on average 0.15 fewer (95% CI [-0.32, 0.02]) hospital stays, while the mean number of hospital stays conditional on having any is 1.3. The probability of having any hospital stay is not meaningfully larger or smaller for individuals with private insurance - the point estimate is , as compared to the mean probability of inpatient admission in the sample of Overall, these results suggest that there is little if any difference in the frequency of outpatient visits for PHI-insured individuals and there is also little if any difference in the probability of experiencing a hospital admission. At the same time, privately insured individuals have fewer hospital stays in a year conditional on having been hospitalized at least once. The estimated correlation between the PHI enrollment and the utilization of healthcare includes the effects of selection and moral hazard that we try to disentangle in the next step. Before moving to this next step, it is important to put these results into the institutional context. Given the multiplicity of differences between the PHI and the SHI, both the causal or moral hazard and the selection effects in this setting themselves include a multitude of potentially countervailing forces. The selection effect could be a combination of strategic cream-skimming by the private insurers, as well as selection on individual preferences that lead different individuals to apply for PHI contracts. The causal effect of the PHI may 12

13 include the classic moral hazard argument, according to which the higher degree of costsharing should decrease the demand for healthcare. At the same time, the causal effect of the PHI could include the physician-induced demand argument, whereby physicians, whose remuneration is substantially higher under the PHI, induce more demand from patients. Yet a third causal channel could arise if PHI-insured are treated better and thus need fewer healthcare services. Lastly, if PHI patients face shorter waiting times and more convenient service, they could be inclined to more utilization of healthcare. The available data will not allow me to cleanly disentangle any of these forces; therefore, it is useful to keep in mind that my empirical findings of selection and moral hazard will necessarily reflect the net of all these channels. Measuring moral hazard To identify the causal effects of having a private insurance plan on healthcare utilization, I exploit the regulatory break in the PHI eligibility as an instrument for private insurance enrollment. Only employees whose income crosses an annually set eligibility threshold may choose to opt out of the SHI system in favor of a private insurance plan. Hence, we would expect a change in the probability of enrolling into the PHI at the income eligibility cutoff. This set-up corresponds to a fuzzy regression discontinuity design, where I use the change in the probability of treatment as the instrument for the treatment status. The discontinuity design is fuzzy, since the crossing of the eligibility threshold only gives the individual the choice to take up the PHI treatment, rather than imposing a switch to the PHI. The key identifying assumption in this setting is that individuals cannot precisely manipulate which side of the cutoff they are on. To explore the plausibility of this assumption, I plot two histograms of income distribution in Figure 3. The histograms zoom in to 1000 EUR income around the cutoff. Given that some income observations may be reported with measurement error the lack of heaping in the histograms around the cutoff should be interpreted with care. At the same time, since employees may report wages with rounding or since employers tend to set rounded wages or use the insurance income cutoff as a wage benchmark, bunching around the cutoff may not necessarily present evidence of manipulation. The histograms in Figure 3 do not provide any strong visual evidence of bunching at the cutoff. Zooming in around the cutoff, there appears to be some bunching around zero in the histogram that uses the centered income variable. The analogous histogram that uses the levels of income rather than the deviations from the cutoff, however, suggests that the bunching occurs there only for years where the threshold was a round number. The income 13

14 thresholds for years are illustrated with vertical bars and they are all close to the 4,000 EUR mark. Looking closer around the vertical bars that represent the exact cutoff amounts, we see no evidence that bunching occurs at those levels if they are slightly further away from 4,000 EUR. Overall, the histogram analysis of the density of the running variable does not seem to present evidence for a systematic manipulation of the running variable. Further covariate balance checks around the cutoff do not reveal any differences along the non-income observables between individuals below and above the cutoff. Table 2 records average age, gender, health status, healthcare utilization, BMI, smoking propensity, disability and risk attitudes for individual-years observed 250 EUR below and 250 EUR above the cutoff. None of these observables have statistically significant differences in means across the two samples. By definition of the cutoff, income is significantly different across the two groups, with monthly income averaging at 3,806 EUR below the cutoff, and 4,009 EUR above the cutoff. The smoothness of the non-income observables, even of those that we would expect to be correlated with income, such as gender and age, corroborates the plausibility of assuming that whether individuals end up slightly above or slightly below the public insurance mandate is as good as randomly assigned. I continue with the estimation of a first stage relationship that tests for the existence of a strong link between the instrument and the PHI enrollment. I use a linear specification that allows for a break in levels at the cutoff and for different slopes before and after the threshold. The income running variable is centered at the cutoff, which allows combining observations from different years that had different threshold levels. The outcome variable is the indicator of whether an individual has PHI: E[P HI it.] = γ 1 + γ 2 Above it + γ 3 (Income it Cutoff t )+ + γ 4 (Income Cutoff t ) Above it + γ 5 X it Figure 4 presents a graphical illustration of the first stage. The first scatterplot uses the baseline analytic sample of full-time employees. We see a clear change in the probability of having the PHI right after individuals cross the income threshold marked with a vertical line. The second scatterplot also plots the probability of enrolling with the PHI at different income levels, but it uses a sample of self-employed individuals only. These individuals do not fall under the SHI mandate at any income level and thus we would not expect a first stage for this sample. Indeed, there is no visual break in trend or jump in the probability of 14

15 PHI enrollment for the self-employed at any income level. 10 The regression results for the first stage are reported in Table 4. The probability of PHI enrollment is estimated to be 24 percentage points higher after the threshold. As we have seen in the graphical representation, the jump is stark and thus the estimates are very precise, with the F-statistic of 172 in the specification with demographic controls X that include age and gender. Having established the presence of a first-stage relationship, I proceed with the analysis of the reduced form specification. Figure 5 provides a graphical representation of the reduced form for six outcomes of healthcare utilization: total outpatient and inpatient visits; probability of having at least one inpatient or outpatient visit; and the number of visits conditional on having had at least one. The graphical representation shows no strong evidence of a discontinuity or change in trend for all six outcome measures. I test for the presence of a statistically significant discontinuity formally using the following linear specification. The specification is similar to the first stage - the income variable is centered at the cutoff and it allows for different income trend slopes before and after the cutoff, and I include age and gender as demographic controls in X: E[Y outcome it.] = α 1 + α 2 Above it + α 3 (Income it Cutoff t )+ + α 4 (Income it Cutoff t ) Above it + α 5 X it Table 5 summarizes the reduced form coefficients. The estimates are imprecise, but broadly confirm the intuition from the graphical evidence that there is no economically meaningful jump in the average utilization of the healthcare services below and above the public insurance mandate income cutoff. Table 5 reports the coefficients of a 2SLS specification that is similar to the reduced form regression, except that I instrument for PHI enrollment with an indicator for being above the income threshold. The point estimates suggest that PHI induces individuals to have 2.3 fewer outpatient visits per year from the 10 The results remain similar in specifications with higher order polynomials; I do not test a non-parametric specification in a small bandwidth around the cutoff due to scarcity of observations right aroud the threshold. Moreover, considering the potential measurement error in income, local results right around the cutoff may be misleading, since the observations around the cutoff may have been misclassified. Note that the graphical evidence suggests that there are a number of observations very close to the cutoff that have a fairly high probability of PHI enrollment, even if their income is reported to be below the eligibility level. The first reason for such observations may be a measurement error in income that leads me to misclassify the individual s eligibility. Secondly, the German health insurance regulation allows individuals that opted out to the PHI at some point and then their income dropped below the current eligibility threshold, to sign a waiver for the re-entry of the SHI. 15

16 mean of 7.2. Most of this effect appears to stem from individuals having fewer visits conditional on having had at least one. Both effects are not estimated precisely at 5% level, but the confidence intervals suggest a substantial negative effect. The probability of having at least one visit is only slightly smaller for the PHI enrollees. The measures on inpatient admission effects are close to zero, but are also quite imprecise. Taken at face value, the point estimates suggest that enrolling with PHI leads individuals to have a slightly higher probability of experiencing an inpatient admission. These patterns may be symptomatic of several underlying mechanisms. For instance, they would be consistent with PHI offering nicer hospital facilities and thus induces less deterrence of inpatient treatment, or inducing inpatient demand through higher physician reimbursement. Overall, the data suggest that PHI causes fewer number of outpatient visits, while the effect on inpatient admissions is likely to be close to zero or slightly positive. Measuring selection In the last step, I combine the OLS and IV estimates to learn about the extent of selection between the public and the private insurance systems. The idea is to substact the IV estimates of moral hazard from the OLS estimates that combine selection and treatment effects. While some of the OLS and 2SLS estimates are imprecise zeros, we can still use the confidence intervals to bound the extent of risk selection. Consider first the outpatient visit frequency as a measure of utilization. In column (1) of Table 3, I estimate that individuals with private insurance report 0.16 fewer physician visits than SHI-insured individuals. The confidence interval on this estimate is [-0.857, 0.544]. The moral hazard component in outpatient visits is estimated at with the confidence interval of [-4.970,0.319] as reported in the first column of Panel B in Table 5. Using the point estimates, we arrive at the implied extent of selection of ( 2.325) = or 30% of the mean number of physician visits. The treatment effect estimate is quite large in this case and implies that individuals who selected into the PHI would have had 2.2 more outpatient physician visits than the individuals that stayed in the public system. In other words, these point estimates suggest the selection of higher outpatient care utilizers into the PHI system. To take into account the imprecision of the point estimates, I next use the confidence intervals to bound the maximum amount of adverse selection that the data would be consistent with. The maximum level of adverse selection occurs at the left hand side of the OLS confidence interval together with the right hand side of the moral hazard confidence interval, leading to a selection effect of ) = of outpatient visits per 16

17 year. We next consider the inpatient admissions. The OLS results for the total number of annual hospital stays, reported in Column 4 in Table3, suggest a combined effect of selection and moral hazard at with a confidence interval of [ , ]. The moral hazard effect of having private insurance is reported in Table 5 as being more visits. The 95% confidence interval for the moral hazard estimate is between and Applying the same logic as in the previous paragraph and subtracting the IV estimates from OLS, we conclude that the extent of selection on the inpatient admissions dimension is around = This suggests adverse selection of individuals with expected 0.04 fewer hospital admissions into the private insurance off the mean of 0.09 admissions per year. Thus, overall the data is consistent with advantageous selection of worse risks into the PHI on the outpatient utilization dimension, and adverse selection of better risks on the inpatient utilization dimension. At the same time, we cannot reject that the selection effects of the PHI are zero, on both the inpatient and outpatient dimensions. These results cast doubt on the prior that private insurers manage to select individuals with substantially lower expected healthcare utilization, who would have very likely been good risks in the public system. In the next section I discuss several possible explanations for this result. 4 Preferences for private health insurance Many factors may be limiting the extent of risk selection between the PHI and the SHI. One possible factor is the presence of taste preferences for private insurance that are either unrelated to the risk profile of individuals or are negatively correlated with risk. The idea of heterogeneous preferences for insurance has been discussed in Hemenway (1990) and empirically corroborated in the context of annuity insurance in Finkelstein and McGarry (2006), as well as in the context of Medicare supplementary insurance - Medigap - in Fang, Keane, and Silverman (2008). More recently, Geruso (2013) has documented age-specific preferences for insurance that go beyond the predicted age-specific health risk in the contenxt of US employer-sponsored insurance plans, while Shepard (2015) has documented the importance of selection across insurance plans based on individuals preferences for having access to star hospitals. In the German institutional environment, the PHI provides not only a different financial product with different premiums and cost-sharing, but also renders access to more convenience in healthcare (shorter waiting times, single hopsital rooms, etc.) and potentially easier access to star physicians. Such convenience preferences are not 17

18 necessarily correlated with risk, and hence, there is substantial scope for choices between the PHI and the SHI on non-risk-related dimensions that could lead to the kinds of patterns we have observed in Section To empirically test for the presence of such heterogeneous preferences for private health insurance, I estimate a discrete choice model of demand for private health insurance. The model takes advantage of the survey data, which allow observing many characteristics of the individuals that are not related to their consumption of healthcare and thus would not typically be observed in more common claims data. We let the utility of individual i from choosing insurance j take the following form: u ij = α i p ij + β i φ j + ɛ ij (1) where p ij is the premium that an individual i pays for choosing insurance option j, while φ j are the characteristics of the insurance choice. ɛ ij is a Type 1 extreme value that accounts for the unobservable part of utility. Individual i chooses insurance j that maximizes her utility. Since in our case j is binary (private or public insurance), the model simplifies significantly. The characteristics termφ j reduces to a insurance-system specific constant that captures the average valuation for each type of insurance. To incorporate preference heterogeneity into the model, I let the coefficient on premium α i that measures the marginal utility of income to depend on whether individuals are employees or self-employed. 12 The coefficient on insurance system choice β i depend on individual characteristics as follows: β i = β 0 + β 1 X i X i are especially interesting in our setting, as they allow to test whether there are observable characteristics of the individuals - for example, political preferences - that are plausibly not directly related to their demand for healthcare, but increase the probability of choosing private health insurance and thus capture some underlying preferences for this type of coverage. X i also includes demographic factors - for example, age, gender, sport affinity, smoking, BMI - that we would expect to be directly associated with health outcomes Baicker et al., 2013 discuss the ideas for the potential of such basic versus more generous coverage in the context of Medicare, as way to efficiently sort beneficiaries according to their willingness to pay for more conveneince or less cost-effective treatments. 12 I have tried richer specifications of the model that allow the marginal utility of income to vary across multiple demographic factors; they do not uncover substantial additional heterogeneity. 13 Given the limitations of the data, the model simplifies the choice problem to a static decision. In reality, individuals face a dynamic decision, since in most cases they can only opt out of the SHI once, without 18

19 The key ingredient in the individual s choice between the statutory and the private insurance systems is the relative price that the individual faces on both markets. In the data, however, I cannot observe SHI premiums and can only observe PHI premiums for those individuals, who chose to enroll into the PHI. Thus, before proceeding with the estimation of preferences, I simulate public and private insurance premiums for all individuals that were eligible to choose PHI. First, I calculate SHI-premiums for both the SHI and the PHI-insured using the regulatory income percentages that determine premiums as a function of income. In the next step, I use the prices reported by the individuals that chose private insurance to run a hedonic pricing regression. Private health insurance underwriting is legally allowed to use information on the individual s age, gender, and health. The real premium calculations are complex and reflect the individual s lifetime expected spending spread equally over the expected lifespan. To construct hypothetical private insurance prices for publicly insured individuals, I use a linear specification to approximate the conditional expectation function of premiums that individuals face for a given combination of age, gender and health status. E[Pi P HI θ i ] = α 0 + α 1 female i + α 2 age i + β HealthStatus i + γ F E i where the P P HI refers to the reported monthly private insurance premiums in EUR for the years , female equals one if the observation is for a woman, age has age in years, HealthStatus refers to a set of measures that captures chronic conditions and healthcare utilization. Table 6 reports the key coefficients for several specifications of the pricing regression. The results imply that a 40-year old male that is a full-time employee would pay about 400 EUR a month for his private health insurance. Reformulating the regression in percentage terms by doing a log-transformation of the price variable, we get that each additional year of life increases the premium by about 2%. Pricing is also sensitive to the individual s gender. It appears that women have both a different level and a different age slope in the PHI pricing. Overall, the linear approximation of the PHI pricing accounts for a substantial amount of variation in the prices, with R 2 equal to 0.3 in the log-specification with diagnosis-specific controls. Using this approximation of the average PHI prices, I calculate the counterfactual PHI premiums for the PHI-eligible individuals that chose SHI in the years This counterfactual calculation is based on a set of non-trivial assumptions. One of them is an assumption that conditional on having the same observable characteristics, individuals who the option of returning to the SHI coverage later. See the discussion of the incentives stemming from the long-terms contracts below. 19

20 didn t switch to the PHI would have faced the same prices as the individuals that did switch. With the estimates of p ij in hand, I proceed to the estimation of the choice model. The estimation is done on the individuals in the baseline sample that face or had faced the choice of enrolling into the PHI. In other words, I consider only employees with income above the income thresold and the self-employed. the choice model. 14 Table 7 reports the marginal effect estimates of I report several specifications that vary on the number of explanatory demographic and health-related variables that are included in the utility function. richest specification includes a full set of demographics, as well as diagnoses indicators. All specifications suggest that conditional on prices, there is significant taste-heterogeneity in the preferences for the private insurance system. For example, older, self-employed, higherincome individuals are more likely to get private health insurance conditional on prices. Since private insurance covers family members not in the labor force at extra premiums (as opposed to free coverage under the public insurance ), it is not surprising to find that single individuals, those with fewer children, and those whose spouse works full-time are more likely to enroll into the PHI. The elasticity of demand towards the price differential conditional on the demographics increases with income levels. At the same time, there appears to be no specific tastes for private insurance based on gender, BMI and risk aversion - individual charactersitics that we would expect to be closely linked to healthcare spending levels. Individuals with disability are less likely to enroll into the PHI - this may be a reflection of tastes (e.g. aversion towards deductibles) as well as rejections by the PHI companies. Having diabetes is associated with a lower probability of choosing the PHI, while other chronic conditions do not appear to be correlated with the PHI choice, which is consistent with our findings in the previous section. Interesting taste heterogeneity is captured by the indicators for whether individuals employ household help and whether they have strong center-left political views. Supposing that household help variable may capture preferences for convenience, the large magnitude and statistical significance of this coefficient would suggest the importance of purely nonpecuniary and non-risk related preferences in the choice of the insurance plans. At the same time the residual impact of the convenience preferences, which are proxied by whether or not the individual employs household help, fades at higher income levels. Overall, I conclude that the data provides strong support for the presence of taste pref- 14 The non-linear logit transformation of the choice model in theory makes the inclusion of the simulated price variable problematic. Hence, I repear the same specifications using a linear probability model that is robust to the inclusion of the simulated variable in the set of regressors. The results are in practice very close across the two models. The 20