1 Introduction

In 1994/95, a mandatory insurance scheme for long-term care (LTC) dependency, in other words, the extended need for assistance with daily living due to physical and/or mental impairments, was established in Germany. The implementation of this novel system not only constituted a landmark in national social policy but was also followed with interest by foreign observers. While the German system was not the first comprehensive public scheme focusing on LTC dependency, the introduction occurred at a time when debates around aging and care intensified in many high-income countries (e.g. Theobald and Kern 2011).Footnote 1 Consequently, the scheme attracted—as did other models of LTC—experts internationally. Two apparently similar LTC schemes were established in East Asia thereafter: The long-term care insurance (LTCI), Kaigo Hoken, in 1997/2000 in Japan; and the LTCI, No-in Jang-gi Yo-yang Bo-heom, in 2007/2008 in the Republic of Korea (hereafter Korea).

This chapter sets out to investigate the role of policy transfer for designing LTCI in Japan and Korea. Transfer describes a process by which knowledge about policies, past or present, is—voluntarily or coercively—used in another context (Dolowitz and Marsh 1996, 344). We focus on the spread of ideas from the German—and, in the case of Korea, also Japanese—LTC system. For this “chain” of introductions, the role of interdependent developments, in particular by means of transnational policy learning, is frequently stressed as having played an important part in the institutional set-up of the LTCI in Japan and Korea (e.g. Campbell et al. 2009; Sunwoo 2012; Maags 2020). However, it remains largely unclear as to what extent international exchange has taken place and influenced system features. In what follows, we ask, what evidence exists for policy transfer—including “positive” and “negative” learning—in the design of the Japanese and Korean LTC systems? To answer this question, we begin by considering the concept of policy transfer and the mechanism of policy learning, specifically, identifying “necessary” and “sufficient” criteria for empirical investigation. We then proceed to examine evidence in two steps: First, we explore the features of the three LTCIs to identify (dis)similarities that suggest transfer took place. Second, we review extant literature on the international travelling and influence of ideas. We conclude by reflecting on the strengths and limits of existing scholarship and suggest next steps for research.

2 Concepts

The role of ideas, knowledge and experience has been highlighted as one important factor in explaining policy-making (Maags 2020). Increasingly, attention has been drawn to exchanges beyond the confines of nation states, allowing the study of welfare policy to move past the so-called “methodological nationalism” characterising much research (Obinger et al. 2013). One concept that emphasises the role of transnational interdependencies in domestic social policy-making is policy transfer (Obinger et al. 2013; Theobald and Kern 2011). For our study, we want to stress two elements: Firstly, policy transfer entails the travelling of knowledge or ideas from country A to country B, for example via scientific experts or government officials. Secondly, the transferred knowledge is employed in country B in some way, indicated, for instance, by similarities in policy design.

Scholarship on transfer and related concepts has identified various types thereof. Accordingly, one important mechanism that can, amongst others, underpin transfer processes is policy learning (Obinger et al. 2013). Constituting a voluntary and intentional form of transfer, learning is an activity of drawing lessons from external information and examples for finding suitable policy solutions and minimising error in domestic policy-making. It is important to note that learning can be both positive and negative in kind (e.g. Dolowitz and Marsh 1996; Theobald and Kern 2011). That is, not only can actors from country B look abroad for best practices for adoption, but they can look to the experiences of country A and consciously choose non-adoption by drawing “negative” lessons. Based on extant studies, we hypothesise that the mechanism of lesson learning is of particular relevance for driving transfer in our cases.

It follows from these theoretical considerations that evidencing transfer entails establishing (1) that the introduction of systems occurred in chronological order; (2) that systems show similar features for demonstrating positive policy learning and differences in the case of negative learning; and (3) that foreign examples and discussions have been picked up by domestic actors and indeed influenced the policy outcome. We regard the first two criteria as necessary conditions for establishing transfer, while the latter can be seen as a sufficient condition for corroborating the role of learning as a mechanism driving transfer among our cases. We now turn to investigate empirically to what extent the three outlined conditions are met.

3 Same but Different: Comparing LTCI in Germany, Japan, and Korea

As indicated earlier, Germany, Japan, and Korea adopted compulsory social insurance LTC systems in succession of one another. The timing of adoptions therefore makes policy transfer a tenable notion. To establish whether the second necessary condition for policy transfer is also met by our cases, this section reviews key similarities and differences in financing, provision, and regulation of LTC (see Table 34.1).

Table 34.1 Comparison of LTCI schemes in Germany, Japan, and Korea

As social insurance systems, they are all underpinned by principles of social solidarity and share several characteristics. The three schemes rely, at least partly, on ear-marked contributions for funding and are managed by designated insurance bodies (Rhee et al. 2015). While remarkable differences in benefits do exist (to be discussed), in general all three systems provide a wide range of benefits across community, home, and institutional settings. Regarding the regulatory framework, the German, Japanese, and Korean LTCIs share formalised eligibility criteria in line with standardised dependency assessments, as opposed to, for instance, means testing (Campbell and Ikegami 2009; Chon 2012). Another common attribute is the relatively wide choice of services and providers, and high provider competition (Campbell et al. 2009; Maags 2020).

Significant differences also exist, especially, but not only, between the German and the other two systems. Firstly, while Germany allows for a minor role for substitutional mandatory private LTCI (Rothgang 2010), both Japan and Korea only have a social LTCI in place. Secondly, in contrast to Germany, the Japanese and Korean systems both employ age as an entitlement criterion (Maags 2020; see Table 34.1). Thirdly, the mix of financial sources differs: Japan and Korea rely on a wider mix of contributions, state budget and co-payments, while the German LTCI is formally financed by contributions exclusively (Rhee et al. 2015).Footnote 2 The final remarkable difference that sets the German system apart is its heavy reliance on cash allowances: While Japan provides in-kind benefits only, and Korea mainly, both in-kind and cash benefits are available and can be mixed in Germany (Campbell et al. 2009).

Differences between Japan and Korea exist, for instance, regarding the extent of regulation and concerning coverage and generosity (see Table 34.1). When comparing the share of LTC benefit recipients as a percentage of the total population for the three countries as indicated in Table 34.1, we see that it is much higher in Germany and Japan, both shortly after introduction of the schemes and today. One reason for this divergence is that the Korean government decided for financial reasons to limit LTCI benefits to the population with the highest need (Seok 2010). While this constitutes one way of controlling public LTC expenditure, Germany implemented a different, albeit also strict cost control mechanism by fixing benefits levels below cost-covering rates (Rothgang 2010). In contrast, both the Japanese and Korean systems use fixed individual co-payments by LTCI beneficiaries, with rates significantly higher for the latter (Maags 2020; Table 34.1). With a view to the similarities and differences of these cases, and given the role of timing, evidence suggests that policy transfer was possible. We now turn to review extant literature to examine whether sufficient evidence exists for the role of policy learning as a mechanism of transfer.

4 Travelling Ideas? Revisiting Evidence on Transnational Policy Learning

4.1 Japan

The adoption of the Japanese LTCI Act in 1997 was preceded by a decade of extensive public discourse on how to deal with the (rising) need for care among the elderly population. In addition to changing socio-economic structures, extended hospitalisations of elderly patients created “negative policy feedback” (Maags 2020, 9) through rising expenditure in the healthcare system (see e.g. Campbell 2000). In 1989 and 1994, the government presented strategies aimed at expanding the existing means-tested, state-run LTC scheme. As these did not provide sustainable solutions (Campbell 2000), policy-makers continued to search for alternatives throughout the mid-1990s. As Japan had experience with both state-based welfare schemes and social insurance programmes, both kinds of schemes were institutionally anchored and seen as potential paths for the new LTC system (Campbell 2000; Campbell and Ikegami 2009).

The literature highlights the role of several relevant domestic actors in the policy design of the novel LTC scheme: administrative officials, especially within the Ministry of Health and Welfare (MHW), political actors like parties and governments, and several interest/advocacy groups (e.g. Campbell 2000; Peng 2005; Campbell et al. 2009; Maags 2020). While we cannot go into detail on their roles here, for our purposes it is important to examine whether and to which extent they were exposed to “foreign” ideas. Campbell and Ikegami (2009), Campbell et al. (2009) and Theobald and Kern (2011) report cross-national exchanges between Japanese officials and German, but also Scandinavian LTC experts. An interesting observation is provided by Campbell (2000, 92–93) who describes that there were two camps of MHW bureaucrats in the 1990s, one “deeply influenced by British and Scandinavian social policy”, the other connected with social health insurance. While this point stresses the importance of transnational ideas or ideologies, it does not necessarily link them to LTC. In any case, it seems that ideas from Germany (and elsewhere) were known and studied within the social policy administration.

Most articles refer to the general model of social LTCI as the object of transfer without greater detail on the transferred features. Some authors (Campbell et al. 2009; Campbell 2000; Theobald and Kern 2011) also mention that monetary LTC benefits were explicitly not adopted, although they are a key aspect of the German system. Campbell et al. (2009, 72) point out that the proponents of cash allowances in Japan used the German example to legitimise their proposal. However, the role of domestic political considerations and (feminist) advocacy in the decision against cash transfers is highlighted much more in the literature (e.g. Peng 2005; Campbell 2000). It therefore remains unclear how important the German benefit model was as a reference point—negatively or positively—in the Japanese policy design.

All in all, what do researchers conclude about policy learning in the case of Japan? Campbell and others stress in several publications (Campbell 2000; Campbell et al. 2009; Campbell and Ikegami 2009) that the German LTCI mainly provided legitimacy for establishing the social LTCI in Japan but did not notably influence or form the choice(s). Others, like Maags (2020) and Theobald and Kern (2011), attribute a greater role to policy learning/adaption, but without showing to what extent foreign policies concretely became relevant for Japanese decision-making. We can thus conclude that knowledge transfer from abroad occurred. However, when it comes to naming precise policy choices or transfer channels, extant research does not adequately corroborate data for the sufficient condition of policy learning to be met.

4.2 South Korea

In Korea, the Elderly LTCI Law was passed in 2007 and implemented the following year, making Korea a comparably early adopter with regard to the country’s welfare state, economic and demographic development (Campbell et al. 2009). The policy process for establishing a new LTC system started in 2000 and was mostly driven by the government’s progressive welfare agenda (Seok 2010; Campbell et al. 2009). The public was generally positive about the establishment of a more comprehensive LTC system; the proposed policy features did not evoke much debate nor advocacy activities (Seok 2010; Maags 2020; Rhee et al. 2015). Consequently, an independent social LTCI model was agreed upon quite easily and deemed suitable for several reasons, among others the existing institutional infrastructure and experience with social insurance programmes (Chon 2012).

In terms of the present interest in transnational policy learning, evidence of exchange in the form of study trips and analysis of foreign systems are mentioned by Campbell et al. (2009, 75) and Chon (2012, 220). However, they do not provide details on the content of the exchange. All we learn is that Korean “experts”, “politicians” and “policy-makers” studied approaches elsewhere. Additionally, Sunwoo (2012, 49) points out that the emerging Korean academic discussion about LTC in the 1990s coincided with the implementation/drafting of the schemes in Germany and Japan, implying a relevance of these policy examples as well.

In the reviewed studies on Korea, only Germany and Japan are cited as concrete foreign models (e.g. Maags 2020; Campbell et al. 2009; Sunwoo 2012), with a somewhat stronger significance ascribed to the latter (e.g. Chon 2012). Authors attribute this to the Korean interest in social insurance. Regarding specific policy features transferred, the idea of a five-year basic LTC plan for revising the system regularly was brought from Japan (Sunwoo 2012, 59), as well as the needs and eligibility assessment instrument (Campbell et al. 2009, 78; Chon 2012, 223). While these specific policy adoptions indicate policy transfer resulting from positive learning, a third instance—the implementation of a cost control mechanism in Korea—points to negative learning. As mentioned earlier, the benefit coverage of the LTC scheme was limited in order to control expenditure. Chon (2012, 222) points out that this financial prudence was inspired by “Germany’s and Japan’s experiences of increasing deficits within a few years of their implementation of LTCI systems”. While more evidence on how these examples specifically exerted an influence on the Korean policy design is certainly necessary, this could be an instance of learning from negatively perceived foreign experiences.

In the literature reviewed, most authors (Chon 2012; Maags 2020; Seok 2010; Sunwoo 2012) attribute influence to foreign models, but base this assessment on similarities of policy features without specifying the sufficient condition of how ideas from abroad might have found a place in Korean LTCI design. In contrast, Campbell et al. (2009) are, as in the case of Japan, sceptical about a greater significance of foreign models. Furthermore, the availability of internationally circulating ideas is mentioned as a source of inspiration for Korean policy-makers (Campbell et al. 2009, 78; Maags 2020), but, again, without much explication of concrete ideas and “travel routes”. Consequently, while research highlights Korean experts’ interest in German and Japanese policies, uncertainty about concrete transmitters and influences remains.

5 Conclusion

The phenomenon of policy transfer and, by extension, the underlying mechanism of policy learning can be evidenced with respect to LTC by analysing several conditions. Firstly, the succession of introductory dates of LTCI systems in Germany, Japan and Korea allows for ideas and experiences of the forerunner(s) to be picked up by later adopter(s). Secondly, the comparison of system designs shows multiple resemblances, but also differences in benefit types and cost controlling mechanisms. Thus, the necessary conditions for policy transfer in terms of both timing and (dis)similarities are fulfilled.

Regarding evidence of actual lesson learning, the reviewed literature points out that in both Japan and Korea, foreign systems that could serve as a blueprint were indeed studied—by policy-makers and academics. In Japan, two schools of thinking could refer to foreign experience, the one relating to Britain and Scandinavia and the other to Germany. In the end, the decision for an LTCI rather than a tax-financed state-led system therefore seems to be the result of national politics, with European examples serving as a springboard for ideas and a source of legitimation. In the non-adoption of cash benefits, referencing for the purposes of legitimation seems also more likely than the occurrence of actual learning. In the Korean case, the insurance model was adopted without seriously discussing a tax-financed system. This might indicate the agenda-setting power of the Japanese example, which Korea follows more closely than the German design—but this is not yet proven. Both pre-adopted schemes might have served as a warning with forbidding budget deficits, although Korean domestic fiscal prudence might have ultimately been more influential for establishing tight cost control than foreign examples.

In sum, information retrieved from our review could not answer the question as to how influential the experiences from abroad were when it came to decision-making. Furthermore, most referencing of expert exchanges remains largely anecdotal without providing much detail about who was involved and how in studying foreign examples, especially in the case of Korea. Therefore, conclusive proof for the sufficient condition is lacking in existing scholarship. Future research on policy transfer and learning should therefore concentrate on the policy-making process and try to reveal how transnational knowledge travels and interacts with domestic institutions and power struggles in designing reforms. This can only be done in in-depth case studies on LTC politics that analyse the interplay of national constellations and transnational interdependencies. Besides the cases of Japan and Korea, more recent developments like the piloting of LTCI schemes in China (see chapter by Liu and Tian) also represent an interesting case to do so.