Health Insurance Competition in Germany – the Role of Advertising

Becker, Bettina; Uebelmesser, Silke Article Health Insurance Competition in Germany – the Role of Advertising Schmollers Jahrbuch – Journal of Applied Social Science Studies. Zeitschrift für Wirtschafts- und Sozialwissenschaften Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Becker, Bettina; Uebelmesser, Silke (2010) : Health Insurance Competition in Germany – the Role of Advertising, Schmollers Jahrbuch – Journal of Applied Social Science Studies. Zeitschrift für Wirtschafts- und Sozialwissenschaften, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 130, Iss. 2, pp. 169-194, https://doi.org/10.3790/schm.130.2.169 This Version is available at: https://hdl.handle.net/10419/292293 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Health Insurance Competition in Germany – the Role of Advertising By Bettina Becker and Silke Uebelmesser* Abstract In the 1990s, competition among public health insurance funds (‘sickness funds’) was introduced in Germany. As one means of competition, free choice of initial health funds and subsequent switching between them was made available to all insured. Since then, the number of funds has decreased substantially, and funds have had to engage in competitive strategies to remain in the market. In this paper, we want to analyse the funds’ advertising activities in the face of the changed competitive environment. This has not been possible to date due to a lack of data. We use two new datasets to get a first insight into the potential effects of competition on funds’ advertising strategies; one of the volume and cost of advertisements and one of their contents. Our results suggest that competition has been associated with an increase in the amount of advertising. As to the adverts themselves, we find that there was a decrease in the share of advertisements of a ‘general’ content in favour of advertisements of a more ‘fund-specific’ content. The data therefore indicate that once the market was open to switching of funds by the insured, funds’ advertising efforts changed to differentiating their own perceived strengths from those of competitor funds. These observations allow us to draw some tentative conclusions about the relevance of (attempts of) risk selection by health funds via advertisements and about the general success of the procompetitive legislation. Zusammenfassung Die Reformen des Gesundheitswesens führten in Deutschland in den 1990er Jahren zu deutlich mehr Wettbewerb zwischen den gesetzlichen Krankenkassen. Als ein wettbewerbliches Instrument wurde für alle Versicherten die Möglichkeit geschaffen, die Krankenkasse frei zu wählen. Seit dieser Zeit ist die Zahl der Krankenkassen deutlich gefallen, und die Kassen mussten wettbewerbliche Strategien ergreifen, um im Markt zu bleiben. In diesem Artikel analysieren wir die Werbeaktivitäten der gesetzlichen Krankenkassen vor dem Hintergrund des veränderten wettbewerblichen Umfelds. Dies war bis jetzt Schmollers Jahrbuch 130 (2010), 169–194 Duncker & Humblot, Berlin Schmollers Jahrbuch 130 (2010) 2 * We would like to thank Mathias Kifmann, Paul Ruud and Jürgen Wasem as well as participants at the European Conference of Health Economics and a seminar at the Universities of Munich and Augsburg for helpful comments and discussion. We would also like to thank Tanja Manhart, Ewa Puchala and Denitza Toteva for excellent research assistance. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 170 Bettina Becker and Silke Uebelmesser wegen fehlender Daten nicht möglich. Wir nutzen zwei neue Datensätze, um einen ersten Eindruck von den potentiellen Effekten von Wettbewerb auf die Werbestrategien der Krankenkassen zu erhalten: einen Datensatz zu Werbeumfang und -ausgaben und einen Datensatz zum Werbeinhalt. Unsere Ergebnisse legen nahe, dass die Intensivierung des Wettbewerbs mit einem Anstieg der Werbeaktivitäten einherging. Was die Anzeigen selbst betrifft, so zeigt sich, dass der Anteil der Anzeigen mit einem allgemeinen„ Inhalt zugunsten von Anzeigen mit einem kassenspezifischen“ Inhalt gefallen ist. Die Daten deuten also darauf hin, dass sich die Werbeanstrengungen der Kassen mit der Einführung der freien Kassenwahl und der Wechselmöglichkeit für die Versicherten geändert haben: die eigenen Stärken werden deutlicher herausgestellt, auch in Abgrenzung zu denen der Wettbewerber. Diese Beobachtungen erlauben es uns, einige vorsichtige Schlussfolgerungen zu ziehen, was die Relevanz von Werbung als potentielles Mittel zur Risikoselektion durch die gesetzlichen Krankenkassen und allgemein den Erfolg der pro-wettbewerblichen Reformen betrifft. JEL-Classification: I11, I18, G22, M37 Received: September 16, 2008 Accepted: March 16, 2010 1. Introduction The demographic evolution, in particular rising life-expectancies, as well as the technological progress in the health sector are largely seen as the main causes of the increase in health expenditures during the last decades. As a reaction to these developments, a number of pro-competition reforms were introduced in the German health sector in the 1990s to increase cost-efficiency and thus alleviate the financial pressure on the health system. 1 For the public health insurance funds (‘sickness funds’), major changes followed from the Health Care Structure Act (Gesundheitsstrukturgesetz) passed in 1992, which extended free choice of health fund to everyone from 1996 onwards, while free choice had previously been restricted to only a small group of insured. Since the passing of this law, the number of health funds has decreased substantially, and funds have had to engage in competitive strategies to remain in the market. One such strategy is advertising. 2 In a regulated competitive market such as the German health sector, advertising may also be used by the funds as one means of indirect risk selection: Instead of reducing costs through increasing efficiency, funds may opt to reduce costs through selective enrolment or ‘cream-skimming’ of low-risk indi- Schmollers Jahrbuch 130 (2010) 2 1For a discussion of the different reforms, see, e.g., Busse/ Riesberg (2004). 2Advertising strategies have been analysed in other contexts, e.g. financial markets (e.g. Cronqvist, 2005; Jain / Wu, 2000; Mullainathan / Shleifer, 2005) and the pharmaceutical industry (e.g. Avery et al., 2008, for a study on direct-to-consumer advertising, and more generally Scherer, 2000; Schweitzer, 2007), but not yet in the context of health insurance markets. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 171 viduals. As Van de Ven / Van Vliet (1992, 42) point out, “solving the problem of cream-skimming is a necessary condition for a successful implementation of a wide range of market oriented strategies in health care, which are being discussed these days in so many countries”. While direct risk selection in Germany is prohibited by law, there is anecdotal evidence that health insurance funds engage in indirect risk selection. Advertising in the German health insurance market may therefore have two faces: First, to simply remain in the market, and second, to increase the share of low-risk members. In this paper, we analyse the funds’ advertising activities in the changed competitive environment. This has not been possible to date due to a lack of data. We use two new datasets to get a first insight into the potential effects of these changes, one of the volume and cost of all advertisements placed by the main German health funds in all German newspapers and magazines, and one of the contents of the advertisements placed in the most advertising-intensive magazine Stern. Our results suggest that competition has been associated with an increase in the amount of advertising by German health insurance funds. In addition, we find that the introduction of competition has been associated with a decrease in the share of advertisements of a ‘general’ content in favour of advertisements of a more ‘fund-specific’ content. The data therefore indicate that once the market was open to switching of funds by the insured, funds’ advertising efforts changed to differentiating their own perceived strengths from those of competitor funds. The paper is structured as follows. Section 2 provides some background information about the German health funds sector. Section 3 presents the quantitative and qualitative data and examines the associations between the introduction of the competitive measures and the advertising activities. Section 4 evaluates the potential of risk selection by health funds in Germany and relates this to the observed advertising activities. Section 5 concludes. 2. Institutional Background Regulation of health insurance in Germany dates back to 1883 when the first Health Insurance Act was implemented. In 2007, about 88% of the German population were insured with a so-called ‘public’ health insurance fund with contribution rates related to wage income but not to individual risk. These insured include mostly employees, students, pensioners, unemployed and those not insured on their own but as a family member, for example children. 3, 4 Schmollers Jahrbuch 130 (2010) 2 3Of those who are not insured with a ‘public’ health insurance fund, 80% are insured with a private insurance. These are mostly the self-employed, civil servants, and employees with an income above a threshold level (A48.600 in 2009). In contrast to public OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 172 Bettina Becker and Silke Uebelmesser Initially, not everyone was eligible to join all health insurance funds. By the end of the 1980s, only about 60% of the insured had some choice (Buchner / Wasem, 2003). In general, everyone had access to the regional, or basic, funds (Allgemeine Ortskrankenkasse AOK). However, employees whose employing company or guild had founded a fund were confined to joining this company fund (Betriebskrankenkasse BKK) or guild fund (Innungskrankenkasse IKK). The so-called substitute funds were available for blue-collar workers (Ersatzkasse für Arbeiter EAR) as well as for white-collar workers (Ersatzkasse für Angestellte EAN), where ‘substitute’ refers to the fact that membership of these funds was a substitute for membership of the AOK, BKK and IKK. The regional insurance funds faced higher average risk portfolios compared with the BKK, IKK and the substitute funds as the share of low-income insured, for example social aid recipients, unemployed and pensioners, was relatively high. 5 As the higher cost, which followed, required the regional funds to charge higher contribution rates, the risk structure as well as the contribution rates differed widely across funds. Furthermore, many of their insured had little or no possibility to switch to a fund with a lower contribution rate. This inequality was considered ‘unfair’, and as a consequence, the German health sector saw a number of pro-competition reform measures in the 1990s. Beside increasing (cost-)efficiency, the reforms were intended to remedy the unequal eligibility of different groups of insured to switch their health funds. The Health Care Structure Act passed in December 1992 marked a major step in that direction. From 1996 onwards, every insured was to have free choice between all open health insurance funds on a yearly basis. Up to 2001, switching funds was possible on an annual basis at the end of each calendar year, while since 2002 switching has been facilitated by allowing for changes on a monthly basis subject to a two-months notice period. However, once a fund is changed, further changes within the following 18 months are permitted only if the insurance fund increases the contribution rate. In order to promote ‘fair’ competition, the Health Care Structure Act subjected funds to ‘open enrolment’ (Kontrahierungszwang), which requires them to insure every applicant, and to ‘community rating’ (Diskriminierungsverbot), which prevents them from charging different premia for different risk Schmollers Jahrbuch 130 (2010) 2 funds, the private funds’ premia are related to individual risks but not to wage income. In this paper, we consider public health insurance funds only. 4In legal terms, there is a distinction between contributing members and so-called family members, i.e. those not insured in their own right but through an insured member of the family. In addition, contributing members can be obligatory or voluntary members depending on whether their wage income falls short or exceeds the threshold level which allows them to choose a private fund (cf. Footnote 3). We refer here to the different groups together as insured or members. 5It is a well-documented empirical observation that income and health are positively related (e.g., Ettner, 1996). OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 173 types. In addition, 95% of the benefits packages are equalised between health insurance funds, as determined by the Social-Code-Book V (Buchner / Wasem, 2003). 6 In order to prevent the selection of low risks, or ‘cream-skimming’, which is prohibited by law, the ‘Risk Equalisation Scheme’ (Risikostrukturausgleich) was implemented in 1994 as part of the Health Care Structure Act. It was meant to compensate health funds for a relatively adverse risk portfolio by reallocating monetary funds between them according to their relative risk structure. The re-allocation is based on the so-called risk adjusters age, gender, disability and sickness allowances entitlement. Income is also taken into account as far as this affects the revenues rather than the costs of the health insurance funds. It is, however, only equalised to 92% across funds. The idea here was that different contribution rates should then reflect only differences in cost efficiency for a standardised risk structure of the insured (Buchner/ Wasem, 2003). The Risk Equalisation Scheme was reformed when the enrolment in disease management programmes was introduced as a further risk adjuster and when a risk pool was established in order to better share the financial risks related to high-risk individuals. 7 From 2009 onwards, the Risk Equalisation Scheme has also included morbidity as laid down in the Health Insurance Competition Strengthening Act (GKV-Wettbewerbsstärkungsgesetz) of 2007. As the new legislation was implemented, the German health insurance market experienced some major changes.The number of health insurance funds decreased by more than 80% between 1991 and 2009, from 1209 to 202 (Figure 1). This concentration process was accompanied by a tendency towards convergence of the major German health funds’ contribution rates, albeit upward rather than downward (Figure 2). Moreover, in the wake of the improved switching possibilities, some of the main funds saw the number of their insured change substantially. Between 1996 and 2008 the AOK and EAN lost 20% and 10% of their insured, respectively (the bulk of that, 16% and 15%, respectively, between 1996 and 2004) (Bundesministerium für Gesundheit, 2009). This reduced their market shares from 43.0% to 34.4% (AOK) and from 35.1% to 31.7% (EAN). While the EAR saw a slight increase in its market share from 1.9% to 2.3%, the IKK and the BKK experienced the largest increase with respect to both the number Schmollers Jahrbuch 130 (2010) 2 6Only since 2004 has it been possible for health funds to offer bonus programmes to their insured. Limiting our analysis to the period 1990 to 2003 enables us to analyse the effect of increased competition on advertising within an otherwise rather stable environment. 7‘Low-risk’ and ‘high-risk’ here refer to the expected expenditures for each type of individuals under the Risk Equalisation Scheme. It is possible that an old person with serious health problems is still more attractive for an insurer than a young person with only minor health problems if the costs assumed in the risk scheme are even higher for the former, while they are lower for the latter. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 174 Bettina Becker and Silke Uebelmesser of their insured and their market share: The IKK gained 47% in terms of numbers of insured and 45% in terms of market share (from 6.9% to 8.7%) and the BKK was able to boost the number of insured by 86% and its market share by 85% (from 10.3% to 19.1%). Source: Bundesministerium für Gesundheit (2001, 2009), BKK Bundesverband (2007). Figure 1: Number of health insurance funds in Germany Note: From July 2005 onwards, the ‘general contribution rate’ was reduced by 0.9% and a ‘supplement rate’ of 0.9% on employees only was re-introduced. Source: Bundesministerium für Gesundheit (2009). Figure 2: Convergence of contribution rates Based on data from the German Socio-Economic Panel (SOEP), Nuscheler / Knaus (2005) conclude that among the 25 to 54 year old obligatory and voluntary members, the percentage of switchers increased from 6.5% to Schmollers Jahrbuch 130 (2010) 2 0 500 1000 1500 2000 1970 1980 1990 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 Germany West Germany Health Act passed Competition begins 11 12 13 14 15 1991 1993 1995 1997 1999 2001 2003 2005 2007 Average AOK BKK IKK EAR EAN Risk adjustm. begins Competition begins OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 175 10.1% between 1995 and 1999. 8 We find that in the subsequent five-year period 2000 to 2004, the switching rate for obligatory members remained relatively stable, fluctuating between 9.0% and 11.1%. The rate in Germany exceeded that in countries with a comparable institutional setting, such as Switzerland, Belgium, Israel and the Netherlands (Laske-Aldershof et al., 2004). In addition, of course, the effect of the threat of switching should not be underestimated. 9 3. Advertising by German Health Insurance Funds In order to gain a first insight into the impact of the changes in the competitive market structure on the advertising activities of the health insurance funds, we analyse a new dataset of advertisements placed by the main German funds over the period 1990 to 2003, provided by Nielsen Media Research (2005), and complement this with data on advertising contents we collected from the most advertising-intensive magazine Stern for the period 1992 to 2003. 3.1 Data on the Number and Costs of Advertisements We have obtained the data on the number of advertisements placed in newspapers and magazines by the main German health insurance funds, i.e. the federal associations of the AOK and BKK as well as the large individual funds Barmer, TK, and DAK (all belonging to the group of substitute funds), along with the data on the costs of these adverts, from Nielsen Media Research (2005), a company specialised in the collection of data on advertising. Figure 3 shows the number and costs of these advertisements (for some summary statistics, please see Appendix I). In total, the five funds placed more than 54,000 advertisements for A225mn in all German newspapers and magazines between 1990 and 2003. There was some advertising in the early 1990s when some limited competition already existed (cf. Section 2). The health funds had probably also known about the forthcoming 1992 law and may have begun to increase their advertising efforts before the law was passed formally. The data then show a further substantial increase in the total number of advertisements since around the time of the passing of the Health Care Schmollers Jahrbuch 130 (2010) 2 8Andersen / Schwarze (1998) and Schwarze / Andersen (2001) come to similar conclusions as the increase of switching is concerned, although they find lower switching rates in the range of 4% in 1997 and 5% in 2000. These differences might be due to different definitions of the switching variable and different sub-samples used. 9A number of recent studies has analysed switching among German health insurance funds (see, e.g., Andersen/ Grabka, 2006, Andersen et al., 2007, Tamm et al., 2007 as well as Nuscheler / Knaus, 2005). As our main focus here is, however, on advertising activities of health funds without explicitly considering the switching response, we abstract from a more detailed discussion of these studies. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 176 Bettina Becker and Silke Uebelmesser Structure Act until just before the introduction of free choice in 1996. This development suggests that health funds used the period 1992 to 1995 for increased advertising as a strategy to defend their market position once insurance-switching would be allowed from 1996. The subsequent short decline in advertising up to 1997 may indicate a period of ‘wait-and-see’ which the funds used to observe whether their advertising efforts would show any success. The change from annual open enrolment to monthly open enrolment in 2002 was associated with another increase in advertising after a somewhat reduced activity level since 1996. Source: Nielsen Media Research (2005) (see Appendix I.1 and I.2). Figure 3: Volume of advertisements in newspapers and magazines (1990 to 2003): Main German health insurance funds The breakdown by health insurance funds conveys further interesting insights. From 1990 to 1994, the AOK and DAK were by far the most advertis- Schmollers Jahrbuch 130 (2010) 2 0 500 1000 1500 2000 2500 3000 3500 4000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 0 1000 2000 3000 4000 5000 6000 7000 AOK Barmer BKK DAK TK Numbers Total (right axis) Numbers 0 2000 4000 6000 8000 10000 12000 0 5000 10000 15000 20000 25000 30000 AOK Barmer BKK DAK TK In 1000 Euro Total (right axis) In 1000 Euro 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 183 introduced into the German health insurance market, in contrast to general advertisements. Comparison with survey data suggests that the contents have changed towards the factors that matter most in individuals’ decisions of whether or not to switch health funds. 4. Risk Selection, Advertising and Competition of Health Insurance Funds We are now in a position to discuss the results of our analysis in the light of the intended outcome of the pro-competitive legislation. We are in particular interested in analysing whether competition among health insurance funds works as intended. Does it lead to more cost-efficiency, or do funds try to attract good risks, i.e. do they engage in risk selection? 13 It is useful to consider the objective function of health insurance funds when they are subject to regulation such as in the German health market. We conjecture here that the objective of health funds, being non-profit organisations, is to increase their size in terms of the number of insured, as one strategy to remain in the market. 14 According to a survey of health fund managers, guaranteeing the continuity of the fund ranks first on their agenda (Haenecke, 2001). Besides, it is common practice that the contracts with the fund management contain clauses according to which bonuses are related to the growth of the fund (Höppner et al., 2006) while growth itself is linked to an increase in the reputation of the management. From the discussion of the institutional framework of the German health market in Section 2, we know that benefits are largely determined by law and therefore largely equal across funds. It is often claimed that competition then takes place in terms of the price, i.e. the contribution rate, rather than in terms of the quality of service (Lauterbach / Wille, 2001; Greß, 2002). 15 Schmollers Jahrbuch 130 (2010) 2 13 A related aspect is the question whether health funds, which resort to indirect risk selection, are successful, i.e. whether via their advertising strategy they manage to affect the switching behaviour and through this the risk structure of their insured. See Becker / Uebelmesser (2010) for an econometric analysis of this question, and Becker / Hole / Uebelmesser (2010) for an analysis of the heterogeneity of households’ preferences for individual health funds. 14 Hart (1983) considers the case of firms run by so-called ‘satisficing’ managers who do not value profits per se but gain private benefits from keeping their job by maintaining the firm afloat. This may hold for managers of a non-profit organisation such as the German health insurance funds. In the model by Hart (1983), an increase in competition may then induce otherwise reluctant managers to increase their efforts to reduce costs in order to avoid bankruptcy. Cost reduction via increases in efficiency would be one way for health funds to reduce their contribution rate to attract new members, cost reduction via a lower risk portfolio would be an additional or an alternative way, and possibly a less costly one in terms of effort involved. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 184 Bettina Becker and Silke Uebelmesser A lower price implies a competitive advantage. With perfect risk compensation, a fund would be the cheaper the more efficiently it operated. With imperfect risk compensation, the risk structure of the insured becomes important. It is then well possible that funds benefit from successful risk selection and a favourable risk structure. Was there any incentive in the German health care market for health insurance funds to engage in risk selection for the period under consideration? As has been shown by different studies for Germany (Breyer /Kifmann, 2001; Jacobs et al., 2002; Lauterbach / Wille, 2001), the number of risk adjusters under the current Risk Equalisation Scheme was far from sufficient to adequately reflect individual risks. For example, as discussed in relation to the characteristics of the Stern readers in Section 3, the educational level was not part of the Risk Equalisation Scheme even though education is likely to be positively correlated with the health status. Buchner / Wasem (2003) also demonstrate that risk compensation was less than perfect: Low-cost health funds with a less than 100% ratio of actual to standardised expenditures have grown fast in recent years, while funds with a ratio above 100% have lost members. Together with open enrolment, these shortcomings of the risk scheme have created incentives for insurers to engage in risk selection so as to either achieve or maintain a low risk profile. 16, 17 There is so far only anecdotal evidence that health funds engage in risk selection (e.g., Van de Ven et al., 2003; Buchner / Wasem, 2003). Glazer / McGuire (2006) conclude that it is not possible to evaluate the significance of the problem as there is no reported evidence on its prevalence. Nuscheler / Knaus (2005) indirectly test for risk selection of BKK funds. They do not find evidence for risk selection of BKKs when comparing the health characteristics of individuals who switch to a BKK to the characteristics of those who switch to a non-BKK. 18 While the authors analyse the possible outcome or output of (successful) risk selection, we in this paper have the data to focus on a potential instrument of, or input to, risk selection, namely advertising. Schmollers Jahrbuch 130 (2010) 2 15 For empirical analyses that show that the contribution rate is a significant factor of an individual’s probability to switch his fund, see for example Andersen / Schwarze (1998), Schwarze /Andersen (2001), as well as Nuscheler / Knaus (2005) for Germany and Buchmueller / Feldstein (1997) as well as Strombom et. al. (2002) for analyses for the US. 16 See Höppner et al. (2006) for an overview of possible risk selection strategies and Van de Ven / Ellis (2000) for a discussion of several welfare-decreasing effects of risk selection. 17 Observing that mainly the young and healthy switch funds is, of course, not proof of risk selection as switching costs might be lower for them (see Cutler /Zeckhauser, 2000; Nuscheler / Knaus, 2005). 18 The analysis is based on the assumptions that non-BKKs do not engage in risk selection and can thus be taken as a benchmark, and that BKKs and non-BKKs are sufficiently homogeneous otherwise. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 185 How can these reflections about potential risk selection incentives be related to our analysis of the advertising activities of health insurance funds? For this, we resort to our content analysis. We consider first advertisements which focus on general topics loosely related to health issues. By placing advertisements of this type, the funds might aim at providing incentives for the insured to improve their health status and thus lower the funds’ health expenditures while at the same time accepting that an advertisement of general content may also have positive spill-over effects on members of rival funds. One might, therefore, expect that general advertisements due to their public-goods character are mainly placed by large health funds and that the number of these advertisements falls when competition is introduced, if competition is assumed to reduce the size of individual funds. 19 This is indeed what we observe (cf. Figure 5). In contrast to general advertisements, the number of fund-specific advertisements has increased since 1994 / 1995 (cf. Figure 6). The change in the relative importance of both types of advertisements as shown in Figure 7 can result from two different strategies. On the one hand, the traditional theory of advertising postulates that advertisements communicate objectively useful information (Stigler, 1961 and in particular Nelson, 1970, 1974) which consumers use to rationally update their beliefs before making their choices. As individuals need information about the main characteristics of the health funds to be able to make a well-founded switching decision, one would expect that more competition would lead to an increase of the number of advertisements which convey ‘fundamental’ information about fund-specific facts, such as the contribution rate of a health fund, as the rational consumer’s demand for this type of information should increase. If advertisements indeed contained this information, this would indicate that competition works as intended. It pays for the funds to compete along the lines stipulated by law in contrast to pursuing risk selection activities. On the other hand, an increase of fund-specific advertisements might also follow from an incentive of the health insurance funds to create a subjective image, which can help in competing for the insured in two ways (Bagwell, 2008): First, it would allow funds to artificially differentiate themselves from competitors despite the a-priori homogeneity imposed on the funds by law, which might justify a mark-up on contribution rates ceteris paribus relative to competitors. Second, an image that would be appealing particularly to good risks could then enable the fund to reduce its contribution rate, thus becoming even more attractive to consumers. If funds followed this behavioural strategy, the introduction of competition should, similarly to the traditional theory, in- Schmollers Jahrbuch 130 (2010) 2 19 This line of reasoning could provide another reason why the funds advertised already before the passing of the Health Care Structure Act (cf. Figure 3). OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 186 Bettina Becker and Silke Uebelmesser crease the number of fund-specific advertisements. 20 Competition would then, however, be associated with more risk selection. Note: There was only one advert in 1997 mentioning costs and contributions. Source: Own data (see Appendix). Figure 8: Advertisements mentioning costs and contribution rates: shares of those with informative and non-informative contents Hence the conclusions regarding potential risk selection and therefore regarding one effect of the introduced competition, will differ substantially depending on which advertising strategy the funds have been following. Identification of the strategy requires a careful analysis of the informational contents of the advertisements, and our data enable us to do this. We focus here on the subgroup of fund-specific advertisements which mention costs and contributions (ignoring those mentioning benefits and services) as they should best allow a distinction between informative and non-infor- mative contents. Only if advertisements refer to costs or contribution rates by giving precise information (e.g., total expenditure, expenditure per insured, contribution rate) are they labelled ‘informative’. When they only generally mention costs or contribution rates, they are labelled as ‘non-informative’. Figure 8 provides some details about the development of both types of advertisements. The data suggest, therefore, that health funds may not primarily use advertising to communicate useful fact-based information. 21 Rather, there is some Schmollers Jahrbuch 130 (2010) 2 20 Of course, also general advertisements might be placed in order to create an image. But as this type of advertisement is no longer very important, we abstract here from further discussing it. 21 Except for the size of the funds the advertisements hardly contain any detailed information. But as we have argued above the size is to a large extent determined by the 0% 20% 40% 60% 80% 100% Non-informative ads 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Informative ads OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 187 evidence that advertising may be one instrument which funds employ to try to attract good risks. This would, of course, be at odds with the intended effect of the pro-competitive legislation. 5. Conclusions We analyse the associations between the pro-competition measures recently introduced into the German health sector and the advertising activities of the major health insurance funds. We use two new datasets in this paper, one of the volume and costs of all advertisements placed by the main German health insurance funds in all German newspapers and magazines, and one of the advertisements placed in the most advertising-intensive magazine Stern. Our results suggest that competition has gone hand-in-hand with an increase in advertising. Although the amount of fund-specific information in advertisements has increased, we have seen that even these advertisements are still relatively little informative. In line with this, a substantial number of survey respondents who did not even consider switching said that they did not see any significant differences between the various funds. This suggests that it may be important for health funds to create an insurance-specific image, with which to generate spurious differentiation and increase consumers’ perception of the fund. This would then further increase the evidence in favour of the behavioural model of advertising and run contrary to the goals of the pro-competition reforms. In future research it would be interesting to see how far our tentative results for Germany may be applied to other countries with a similar institutional setting. References Andersen, H. H. / Grabka, M. M. (2006): Kassenwechsel in der GKV 1997 – 2004: Profile – Trends – Perspektive, in: D. Göpffarth / St. Greß, K. Jacobs / J. Wasem (Hrsg.): Jahrbuch Risikostrukturausgleich 2006, Sankt Augustin, 145 – 189. Andersen, H. H. / Grabka, M. M./Schwarze, J. (2007): Premium rates, competition among health insurance funds and the health care reform 2007, Jahrbücher für Nationalökonomie und Statistik 227, 429 – 450. Andersen, H. H. / Schwarze, J. (1998): GKW‚ 97: Kommt Bewegung in die Landschaft? Eine empirische Analyse der Kassenwahlentscheidung, Arbeit und Sozialpolitik 9 / 10, 11 – 23. Avery, R. J. / Kenkel, D. / Lillard, D. R. / Mathios, A./Wang, H. (2008): Health disparities and direct-to-consumer advertising of pharmaceutical products, in: L. Helmchen /R. Kaestner / A. L. Sasso (eds.): Advances in Health Economics and Health Services Research 19, 71– 94. Schmollers Jahrbuch 130 (2010) 2 institutional restrictions of the past and thus less of an indicator of the present quality of a fund. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 188 Bettina Becker and Silke Uebelmesser Bagwell, K. (2008): The economic analysis of advertising, in: M. Armstrong / R. Porter (eds.): Handbook of Industrial Organization, Vol. 3, Amsterdam, 1701 – 1844. Becker, B. / Hole, A. / Uebelmesser, S. (2010): Heterogeneity of household preferences for health insurance companies, mimeo, Loughborough University, University of Sheffield and University of Munich. Becker, B. / Uebelmesser, S. (2010): The impact of print media advertising on public health insurance switching, mimeo, Loughborough University and University of Munich. BKK Bundesverband (2007): Entwicklung der Anzahl der Krankenkassen in der GKV, Berlin. Braun, B. / Greß, St. / Höppner, K./Marstedt, G./Rothgang, H. / Tamm, M. / Wasem, J. (2006): Barrieren für einen Wechsel der Krankenkasse: Loyalität, Bequemlichkeit, Informationsdefizite?, in: J. Böcken / B. Braun / R. Amhof / M. Schnee (Hrsg.): Gesundheitsmonitor 2006, Gütersloh, 11 – 31. Breyer, F. / Kifmann, M. (2001): Optionen der Weiterentwicklung des Risikostrukturausgleichs in der GKV, DIW Diskussionspapier Nr. 236, Berlin. Buchner, F. / Wasem, J. (2003): Needs for further improvement: Risk adjustment in the German health insurance system, Health Policy 65, 21 – 35. Bundesministerium für Gesundheit (2001): Daten des Gesundheitswesens 2001, Bonn. Bundesministerium für Gesundheit (2009): Daten des Gesundheitswesens 2009, Berlin. Burda Advertising Center GmbH (2008): Reichweitenanalyse (ma 2008) Pressemedien I, http://ma.bik-gmbh.de/burda/(20 August 2008). Busse, R. / Riesberg, A. (2004): Health care systems in transition: Germany, WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies: Copenhagen. Cronqvist, H. (2005): Advertising and portfolio choice, Working Paper CeRP N°44 / 05, Moncalieri. Cutler, D. M. / Zeckhauser, R. J. (2000): The Anatomy of health insurance, in: A. J. Culyer/ J. P. Newhouse (eds.): Handbook of Health Economics 1A, Amsterdam, 563– 643. Ettner, S. L. (1996): New evidence on the relationship between income and health, Journal of Health Economics 15, 67 – 85. Glazer, J. / McGuire, Th. G. (2006): Contending with risk selection in competitive health insurance markets, Perspektiven der Wirtschaftspolitk 7, 75– 91. Greß, St. / Groenwegen, P. / Kerssens, J. / Braun, B. / Wasem, J. (2002): Free choice of sickness funds in regulated competition: Evidence from Germany and The Netherlands, Health Policy 60, 235 – 254. Greß, St. (2002): Freie Kassenwahl und Preiswettbewerb in der GKV – Effekte und Perspektiven, Vierteljahresheft zur Wirtschaftsforschung 71, 490 – 497. Haenecke, H. (2001): Unternehmensziele von Krankenkassen – Eine empirische Analyse, Arbeit und Sozialpolitik 1 – 2, 27 – 34. Schmollers Jahrbuch 130 (2010) 2 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 189 Hart, O. (1983): The market mechanism as an incentive scheme, Bell Journal of Economics, 14, 366 – 382. Höppner, K./Greß, St. / Rothgang, H./ Wasem, J. (2006): Instrumente der Risikoselektion – Theorie und Empirie, in: D. Göpffarth / St. Greß / K. Jacobs / J. Wasem (Hrsg.): Jahrbuch Risikostrukturausgleich 2006, Sankt Augustin, 119 – 144. IVW (2010): Auflagenzahlen Print, http://www.ivw.de/index.html (1 March 2010). Jacobs, K. / Reschke, P. / Cassel, D. / Wasem, J. (2002): Zur Wirkung des Risikostrukturausgleichs in der gesetzlichen Krankenversicherung – Eine Untersuchung im Auftrag des Bundesministeriums für Gesundheit: Endbericht, Schriftenreihe des Bundesministeriums für Gesundheit, Bd. 140, Baden-Baden. Jain, P. / Wu, J. S. (2000), Truth in mutual fund advertising: Evidence on future performance and fund flows, Journal of Finance 55, 937 – 958. Laske-Aldershof, T. / Schut, F. / Beck, K. / Greß, St. / Shmueli, A./Van de Voorde, C. (2004): Consumer mobility in social health insurance markets: A five-country comparison, Applied Health Economics and Health Policy 3, 229 – 241. Lauterbach, K. W./ Wille, E. (2001): Modell eines fairen Wettbewerbs durch den Risikostrukturausgleich. Sofortprogramm „Wechselkomponente und solidarische Rückversicherung“ unter Berücksichtigung der Morbidität (Abschlussbericht). Mullainathan, S. / Shleifer, A. (2005): Persuasion in finance, mimeo, Harvard University. Nelson, Ph. (1970): Information and consumer behavior, Journal of Political Economy 78, 311 – 329. Nelson, Phillip (1974): Advertising as information, Journal of Political Economy 82, 729 – 754. Nielsen Media Research (2005): Data on ads placed by AOK, Barmer, BKK, DAK, TK in newspapers and magazines, Nielsen Median Research GmbH (acquired by purchase). Nielsen Media Research (2009): Data on the development of gross advertisement expenditures 1990 – 2008 (Brutto-Werbemarkt), http://www.ip-deutschland.de/ipd/ forschung_und_service/mediaforschung/werbemarktdaten/entwicklung_des_werbe marktes.cfm (25 February 2010). Nuscheler, R./ Knaus, Th. (2005): Risk selection in the German public health insurance system, Health Economics 14, 1253 – 1271. Scherer, F. M. (2000): The pharmaceutical industry, in: A. J. Culyer / J. P. Newhouse (eds.): Handbook of Health Economics, vol. 1A, Amsterdam, 1297 – 1336. Schwarze, J. / Andersen, H. H. (2001): Kassenwechsel in der Gesetzlichen Krankenversicherung: Welche Rolle spielt der Beitragssatz?, DIW Diskussionspapier 267, Berlin. Schweitzer, S. O. (2007): Pharmaceutical Economics and Policy, New York. Stern (1992 – 2003): Weekly magazine, Gruner + Jahr. Stigler, G. (1961): The economics of information, Journal of Political Economy 69, 213 – 225. Stigler, G. (1987): The theory of price, New York. Strombom, B. A. / Buchmueller, Th. C. /Feldstein, P. J. (2002): Switching costs, price sensitivity and health plan choice, Journal of Health Economics 21, 89 – 116. Schmollers Jahrbuch 130 (2010) 2 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 190 Bettina Becker and Silke Uebelmesser Tamm, M. /Tauchmann, H. / Wasem, J. / Greß, St. (2007): Price elasticities and social health insurance choice in Germany – A dynamic panel data approach, Health Economics 16, 243 – 256. Van de Ven, W. P. M. M. / Beck, K./ Buchner, F. / Chernichovsky, D. / Gardiol, L. / Holly, A./ Lamers, L. M./ Schokkaert, E. / Shmueli, A. / Spycher, St. / Van de Voorde, C. / van Vliet, R. C. J. A. / Wasem, J. / Zmora, I. (2003): Risk adjustment and risk selection on the sickness fund insurance market in five European countries, Health Policy 65, 75 –98. Van de Ven, W. P. M. M. / Ellis, R. P. (2000): Risk adjustment in competitive health plan markets, in: A. J. Culyer /J. P. Newhouse (eds.): Handbook of Health Economics, Vol. 1A, Amsterdam, 755– 845. Van de Ven, W. P. M. M. / van Vliet, R. C. J. A. (1992): How can we prevent cream skimming in a competitive health insurance market?, in: P. Zweifel / H. E. Frech III (eds.): Health Economics Worldwide, Dodrecht, 23 – 46. Appendix Data on number and costs of advertisements a) Number of ads Funds Obs. Mean Std.Dev. Min Max AOK 14 2532.50 641.86 1591 3649 Barmer 14 143.00 105.44 7 307 BKK 14 414.50 269.14 43 956 DAK 14 515.29 506.45 9 1631 TK 14 258.79 248.87 32 935 Year Obs Mean Std.Dev. Min Max 1990 5 343.8 697.65 8 1591 1991 5 389.4 799.53 7 1819 1992 5 620.8 1220.75 31 2803 1993 5 745.2 1294.03 34 3046 1994 5 953.6 1527.37 49 3649 1995 5 1193.8 1407.85 154 3606 1996 5 1128.2 1204.56 253 3028 1997 5 665.4 702.21 134 1889 1998 5 741 739.29 213 1956 1999 5 700.2 847.09 166 2194 2000 5 1056.2 917.21 242 2561 2001 5 687 1005.47 92 2458 2002 5 895 1025.13 39 2639 2003 5 699.8 859.96 170 2216 Schmollers Jahrbuch 130 (2010) 2 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 Health Insurance Competition in Germany 191 b) Costs of ads (in Euro) Funds Obs. Mean Std.Dev. Min Max AOK 14 6711.07 1399.68 4891 10217 Barmer 14 1155.28 1506.14 9 3771 BKK 14 2426 1945.09 158 6843 DAK 14 4315 3475.10 4 10466 TK 14 1441.35 1059.63 168 3097 Year Obs Mean Std.Dev. Min Max 1990 5 1260.6 2576.25 4 5865 1991 5 1301.8 2716.93 9 6160 1992 5 2593 3717.93 29 8602 1993 5 2970.4 3747.91 33 7238 1994 5 3868 4925.73 200 10217 1995 5 4962.4 4187.31 303 10466 1996 5 4092.2 3736.59 308 9590 1997 5 2459.4 2450.13 160 6183 1998 5 3425 1609.97 2279 6226 1999 5 3629.6 2144.26 1673 7307 2000 5 4320.6 2094.54 1768 7586 2001 5 2563.2 1606.91 979 5196 2002 5 3779.8 2556.80 112 6471 2003 5 3710.4 818.24 2582 4891 Source: Nielsen Media Research (2005) Data on the contents of advertisements a) Number of ads with general content (food, fitness and other) Funds Year Total 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 AOK 11 11 7 2 1 32 Barmer 2 2 6 10 BKK 3 3 DAK 101111 8 1 41 TK 4 4 Total 21 22 18 11 1 2 2 2 11 90 Schmollers Jahrbuch 130 (2010) 2 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00 192 Bettina Becker and Silke Uebelmesser b) Number of ads with fund-specific content (benefits, services, costs, contribution rates) Funds Year Total 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 AOK 4 4 3 3 14 Barmer 5 8 7 3 12 35 BKK 8 6 1 9 10 6 10 13 9 72 DAK 2 2 1 14 6 7 3 2 15 19 71 TK 2 3 8 8 21 Total 2 2 5 8 20 5 25 28 16 18 36 48 213 Schmollers Jahrbuch 130 (2010) 2 c) Ads with fund-specific content (only costs, contribution rates = CC) – differentiated acc. to content (non-informative vs. informative) Year Total 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 CC ads (noninfo) in % of total CC ads 0 83 100 78 60 60 60 CC ads (info) in % of total CC ads 100 17 0 22 40 40 40 Total CC ads 0 0 0 0 0 1 12 15 9 10 5 15 67 d) General and fund-specific ads in % of total number of ads (per year) Year Total 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 General content 100 100 78 42 407080422 Fund-specific content 10 9 22 31 77 100 86 97 64 86 77 98 Fund-specific content incl. size of the fund 10 9 22 69 81 100 86 97 72 95 81 98 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.130.2.169 | Generated on 2023-01-16 13:36:00