Why the Discretion Mix Matters: Understanding the Transformation of Long-term Care Services.

This study aims to extend the concept of discretion, ie, a certain degree of freedom in crucial decisions left to specific actors, to understand and examine the transformation of social care services in the era of aging and austerity. Although previous studies have reviewed and analyzed changes in care provision, they have been less concerned with who has the authority to make care decisions in the implementation process. We propose a new theoretical concept, the discretion mix, to understand the realignment of social care services beyond simply tracking institutional changes. Using a case study approach, this research investigates how the discretion mix of the Korean long-term care system has changed and the consequences of these changes; in addition, it discusses why the discretion mix can be a useful concept for analyzing the changing landscape of social care services.

What do we already know about this topic?

Long-term care reform is understood through the concepts of welfare mix, marketization, and decentralization.

How does your research contribute to the field?

The concept of discretion mix provides a better understanding of long-term care transformation by offering a framework that incorporates the different roles of the central government, local government, street-level bureaucrats, service providers, and users in daily-care decision-making.

What are your research’s implications toward theory, practice, or policy?

While much policy discussion pays attention to the provisional aspect of care, it is equally important to consider who should have the discretion to decide the details of care services.

Objectives

This study aims to extend the concept of discretion, defined as a certain degree of freedom in crucial decisions left to specific actors, to understand and examine the transformation of social care services in the era of aging and austerity. This study begins by providing a critical review of existing concepts related to the transformation of long-term care (LTC) services and argues that they are insufficient for understanding the nature of LTC transformation. In this article, we propose a new theoretical concept—the discretion mix—to aid in the understanding of the realignment of social care services and explore its conceptual validity by examining the discretion mix of the LTC scheme in South Korea (hereafter, Korea). We empirically examine how the particular form of discretion mix in Korea has developed and how this has led to problems in policy implementation. We argue that the discretion mix can be a useful concept for analyzing changes in territorial coordination and conflict and cooperation among the actors involved in the implementation process.

The Limitations of Existing Concepts

One of the most significant changes in contemporary welfare states has been in LTC services and pensions. Unlike pensions, LTC services have only been expanded to keep pace with rapid aging since the late 20th century, when austerity pressures intensified in developed countries. This has created a dilemma for advanced welfare states: reduced financial capacity and increasing social needs versus the increasing number of dependent elderly people.[1,2] Some countries—such as those in Scandinavia—have restructured and reduced the existing scope of service eligibility for the frail elderly.[3-5] Other countries—particularly late social service developers—have introduced new universal programs to respond to their growing aging population.[6,7] In this sense, LTC programs are often regarded as the main testing ground for addressing the increasing social needs in the era of austerity.[8]

Then, what is the nature of these changes and how can we understand the transformation of LTC services? Whether to expand or downsize is an important question in social policy reform, but the line between expansion and retrenchment in social care is not always clear-cut. Most LTC reforms have been in conjunction with the reform of meso-level governance with the aim of making LTC programs more affordable and efficient. One of the most prominent meso-level reforms is marketization, in which private actors become more influential in care arrangements. Competition among suppliers is perceived as a key mechanism in improving the efficiency of care service delivery. Cash-for-care schemes and quasi-cash payments such as vouchers have been widely introduced to increase service users’ level of control.

Another notable reform trend in the care sector is the reorganization of care governance through either decentralization or recentralization. In Europe, the importance of central government decreased during the post-war period and the responsibility for delivering social services shifted to the local level.[9,10] France and Japan, regarded as strongly centralized states, have pursued decentralization reform since the 1980s.[11,12] Meanwhile, recentralization has been observed in other countries. In Germany and Korea, the introduction of LTC insurance (LTCI) sparked significant changes such as nationalized care governance becoming tied to insurance management.

Several concepts have been proposed to explain these transformations in LTC governance. The concept of welfare mix identifies empirical differences between institutional combinations in different countries and highlights the transformation of the public–private mix in proportion to service ownership and funding.[13,14] The welfare mix is particularly useful for illustrating diversity and variation in the national arrangements for welfare provision, delivery, and financing beyond state welfare.[15] Although provisional and financial aspects are the most common conceptualizations of the welfare mix, some authors emphasized that it captures both the organization of different financial or provisional matrices and the organization of various norms and logics.[16,17]

Although the welfare mix approach has several advantages for studying changes in LTC governance, it also has certain limitations. First, it does not reveal what happens within the state. Decentralization and recentralization are important to the understanding of LTC transformation, but the welfare mix does not offer significant insights in this regard. Second, the descriptive and static nature of the welfare mix does not capture the dynamics of the different levels of government and different actors involved in LTC reform and the tensions between them. For example, the market in this approach has its own norms and logics but—depending on its autonomy from the state—its role could vary within an LTC system.

In this respect, the concept of subsidiarization proposed by Kazepov[10] has clear advantages over the welfare mix approach; it distinguishes between 2 processes of change: vertical subsidiarization with respect to “the territorial reorganization of regulatory powers,” and horizontal subsidiarization with respect to “the multiplication of actors.” In a related vein, Andreassen[18] proposed democratic consultation models concerning citizen involvement in health care and social care. Co-production is another concept that can be used to elucidate the transforming relationship between different actors, including service users. Hunter and Ritchie[19] defined this concept as a particular form of partnership between service users and service providers that can be applied in both planning and implementation. In this model, service users are no longer considered passive recipients[20] or mere consumers that simply choose between the different options they are offered.[21] Rather, co-production focuses on the active engagement of service users through dialogue or “intimate conversations.”[22]

The concepts of subsidiarization and co-production provide impressive breadth for understanding the transformation of LTC governance. However, they are less effective for providing an interactive picture of how meso-level reform can affect the discretion of various policy actors—from governmental actors to providers and users in the implementation process—and how these actors’ decisions interact to produce a particular type of care service. For instance, marketization both rearranges the roles of the state and the market and has ramifications in the reorganization of the roles of the central and local governments, while also giving private service providers and service users new roles with a certain amount of discretion in the policymaking process.

Furthermore, few existing concepts, including the welfare mix, have paid attention to the decision-making and discretionary aspects of LTC reform. Horizontal subsidiarization and co-production deal with these, but they are more concerned with explicit democratic participation in the decision-making process and are less interested in who makes the decisions in routine policy implementation. Thus, this research proposes a new conceptual framework that complements these existing concepts, the discretion mix.

The Discretion Mix

In classic theories on policymaking and implementation, most policy decisions are made within a legislative framework and public officials of various ranks are supposed to follow and enforce the rules in the command and control systems. From this perspective, confining, structuring, and monitoring discretion at the local level is a major concern,[23,24] and only the issue of compliance remains for bureaucrats and citizens who act as implementation agents. However, a group of scholars has challenged the separation of policymaking and implementation, contending that there is a divergence in local practices among lower-level policy actors[25] When particular legislation is enacted through the democratic process at a national level, the law itself outlines important principles and details and delegates the rest of “what to do” to other policy actors. These policy actors, who are given a substantial amount of discretion in their decision-making, play a much more important role than merely executing the law.[24,26-28]

This has led to considerable scholarly interest in investigating the vertical dimension of the policy process, in which each actor in the implementation process has a different range of decision-making powers. Many scholars argue that legislators delegate sufficient discretion to both upper-level executive authorities and lower-level actors to make and implement policy decisions. Here, we define discretion as the degree of freedom left to specific actors by an authority for crucial decisions in the routinized implementation and operation of specific policy measures. Many studies use discretion interchangeably with autonomy.[28] However, unlike autonomy, which is related to power and the notion of legitimacy, Carpenter[29] noted that discretion cannot be understood without rules and laws because an actor can only exercise discretion—ie, the leeway to interpret and act—within given bounds.

From a juridical perspective, discretion is delegated authority based on formal rules, whereas it is the freedom of judgment from a sociological perspective.[28] Some discretion is explicit in that it is designated by law, executive orders, or guidelines (ie, de jure discretion), while other forms of discretion are implicit in that actors need to actively interpret and implement the policy because it contains unclear or unwritten elements, ie, de facto discretion.[30] The definition of discretion in this study includes both de jure and de facto discretion.

Previous studies have focused on street-level bureaucrats when discussing the role of discretion. Some scholars argue that street-level bureaucrats’ discretion has been curtailed due to increasing managerial control and regulation[31] and the informatization and e-governance of public administration.[32] However, others insist that—despite the more elaborate rules and regulations in place—de facto discretion has not been reduced and in fact still plays an important role in the implementation process.[33,34] At the same time, other policy actors who exercise discretion have been overlooked and less attention has been paid to the characteristics of legal and institutional systems, which “can shed a differentiated light on the role of discretion.”[34] Indeed, the discretion of street-level bureaucrats can only be properly gauged in relation to that of other policy actors.

We assume that many policy actors are able to exercise discretion under a particular law that is the original source of their discretion. Specifically, we identify 3 levels of policy actor within executive bodies: the central government, local governments, and street-level bureaucrats. The central government often delegates a significant amount of power to specify policy content in accordance with the law and may exercise this discretion by producing executive orders or guidelines; then, central executives can delegate policymaking power to lower levels before deciding on the specifics of a policy. When a significant portion of the decision-making power is delegated in this manner, local governments can more readily adapt policies for their own context. In contrast, they may have a very marginal role if the law and/or executive orders are highly specific, thus relegating them to the role of administrative agents.

Street-level bureaucrats turn rules and regulations from upper-level decision bodies into daily routines. These bureaucrats may engage in the policy process by further embodying the law[27,35] or they may be strictly bound by a specific law or specific guidelines. In the former situation, the discretionary judgment of frontline workers is often regarded as a form of political decision-making and/or policymaking;[36] in the latter case, frontline workers are considered subservient agents who are not supposed to deviate from a predefined regulatory framework.

The delegation of discretion or decision-making power is not limited to governmental actors. Two other important policy actors are often granted discretion: service providers and service users. Both governmental and nongovernmental agencies tend to actively respond to changes in the law by exercising substantial discretion to reduce any negative effects for themselves[37] while still remaining constrained by the associated legal framework. Depending on the type of public–private partnership in care services, providers can enjoy autonomy from the government. As Bode[38] argued, recent public–private partnerships have tended to be more volatile and heterogeneous, as governments have delegated more power to private actors. However, the specific form of the public–private mix varies considerably between countries, particularly in terms of the use of discretion between levels.

It is also important to recognize that, unlike previous service provision systems, cash-for-care schemes significantly increase the influence of service users in terms of how services are produced and used. Although the range of their decision-making power varies between countries,[39] they can still influence the overall delivery system by becoming employers in relation to their care workers and service providers.[14,40] As mentioned earlier, if the legislative body decides to extend cash benefits to individuals rather than provide government- or market-led services, a large amount of discretion can be transferred directly from bureaucrats or service providers to care recipients.[41]

Thus, we propose a discretion mix, an extension of the concept of discretion, to understand the static and dynamic nature of social care services. This framework is based on the theoretical assumption that it is difficult to distinguish policymaking from policy implementation.[23] Policy actors at each level occupy distinct policy positions in the policymaking process depending on the associated law and institutional structure. This study defines the discretion mix as the established distribution of discretion between central-level, local-level, and street-level bureaucrats, service providers, and service users under the law. As with the welfare mix, the discretion mix is a neutral concept insofar as it does not imply a particular direction for policy changes, unlike decentralization and recentralization. In addition, it differs from co-production and subsidiarization in that it directly concerns actors’ choices regarding routine policy activities. In this article, we focus on the key aspects of care decisions, including the authority to determine eligibility (ie, to whom), the amount of care, the type of care service, and the choice of delivery (ie, from whom).

As seen in Figure 1, the law generally delegates discretion to other policy actors. If it did not, any minor change in care services would have to go through parliament. German law is similar; however, the law rarely specifies all details, instead leaving substantial discretion to the executive level. If the central government were to decide the specific details of each policy, other actors would enjoy only minor discretion. In contrast, legislation can explicitly afford local actors more discretion, as found in Nordic countries. Elsewhere, if there is no well-established or coherent national policy framework, as in Southern Europe, local governments are given more discretion to shape the system of care. If local-level governments do not specify the content of a policy, the decision-making power is delegated to street-level bureaucrats, which allows them to shape the policy in a manner that suits their context. If the law establishes a cash-for-care scheme, users can act as employers and are given strong autonomy without much intervention from central or local governments.

Figure 1.

Dynamics of discretion mix.

The discretion mix is not uniform between welfare states. Like the welfare mix, the specific form of a country’s discretion mix will reflect the nature of its welfare regime, which has been historically institutionalized. We can assume that private actors have more discretion in liberal welfare regimes, whereas a more central government will likely have more discretion in developmental states, such as in East Asia. However, the discretion mix is never static and its boundaries often shift in response to reform initiatives and changing contexts. For example, privatization could transfer a certain portion of the discretion of street-level bureaucrats to service providers.[42] Giving greater fiscal flexibility to local governments via block grants may also lead to lower-level government entities possessing greater discretion.[43]

Methodology

The analysis of discretion mix in the LTC sector in Korea is based on data collected in 2017 and 2018. Data were obtained from 2 sources: first, first-hand administrative documents produced for frontline agents by the government, which explains the procedures they should follow and numerous administrative documents that are available from Ministry of Health and Welfare (MOHW) websites. The second source is focus group interviews with service providers and in-depth interviews with relevant policy actors.

First, when analyzing administrative data, content analysis is used to track the changes in the discretion mix. Content analysis involves making inferences after systematically and objectively identifying specific characteristics of the target content.[44] Legal scholars tend to use content analysis without referring to a specific methodological approach; they often analyze a legal system by isolating other methodological tenets, which reduces complex legal information to several categories and concepts. In social sciences, content analysis has been used as a “systematic, replicable technique for compressing many words of text into fewer content categories.”[45] As the concept of discretion is strongly anchored to laws and regulations, we analyzed Korea’s LTC laws and related regulations. Specifically, we used the Act on Long-term Care Insurance for Senior Citizens, the Enforcement Decree for the Act on Long-term Care Insurance for Senior Citizens, the Enforcement Regulation for the Health and Welfare Services for Senior Citizens, and various guidelines for service providers from the MOHW. These guidelines are particularly important because they embody and specify the routine implementation of LTC services.

To analyze the different levels of discretion for each actor, we proposed 3 hierarchical levels of discretion: strong, weak, and none. Strong discretion means when a policy actor is highly autonomous as dictated by laws or regulations (de jure discretion) or when a higher authority only provides general principles without any specifics, thus leaving significant room for interpretation (de facto discretion). Weak discretion means that a policy actor only plays an additional role in care arrangements within a narrow range of options. Although a policy actor cannot transform specific rules or the principle framework, their de facto discretion could still have significant implications for care recipients. Finally, no discretion means that a policy actor is bound by highly specific laws and/or regulations and thus has no meaningful autonomy apart from limited inherent discretion. Based on these concepts, we have identified the discretion mix by qualitatively interpreting the laws and regulations.

Second, focus group interviews are used to complement the content analysis and investigate the actual implementation of the changed discretion mix after the marketization and its consequences in Korea. The first focus group interviews were conducted with different types of service providers: certified LTC facilities, uncertified LTC facilities, and home care service centers. The heads of the 36 service providers (the interview dates and number of participants are as follows: (1) August 31, 2017 [8], (2) September 14, 2017 [7], (3) September 5, 2017 [6], (4) September 1, 2017 [8], (5) September 12, 2017 [7]) were attended in 5 separate sessions in August and September 2017. In addition, we conducted 3 further focus group interview sessions with 6 staff members from the National Health Insurance Agency and 8 public servants in municipalities who were responsible for the LTC administration (The interview dates and number of participants are as follows: the National Health Insurance Scheme (NHIS) staff members: August 30, 2018 [6], municipality public officials: October 2, 2018 [4], and October 11, 2018 [4]) in August and October 2018. The former interview was implemented as part of a research project of the Seoul Welfare Foundation and the latter as a part of a research project for the Social Security Committee in Korea.

The overall purpose of the focus group interviews was to observe the specific experiences and roles of field managers in the private sector and administrators on the government side. Service providers who represented the typical managerial workforce in the LTC sector were selected by Seoul Welfare Foundation and the interviews with government personnel were arranged by the Social Security Committee. Each agency followed the organizational ethical process and obtained consent from interviewees while assuring them of their confidentiality and anonymity. Semi-structured interviews with interview guides were conducted and recorded. Each session lasted approximately 2 hours. While the interviewers questioned their perception of and experiences with the managerial and administrative process and the evaluation of LTC services, they also had information about their relationship with other key actors, including the central government and citizens.

The interviews were fully transcribed by a professional research company. Following the transcription, authors independently read the scripts. Then, we repeatedly read the transcripts to select relevant research materials relating to the discretion mix to identify key themes and categories. The analysis was inspired by Bacchi’s “What’s the problem?” approach in the sense that we searched for the problematic conditions that the strategies and approaches of policy actors aimed to solve.[46] For triangulation, in addition to the context analysis and the analysis of focus group interviews, we cross-checked the validity of the data with the existing literature.

The Changing Discretion Mix in Korean LTC Services

Discretion Mix of LTCI in Korea

Prior to 2000, Korea was not considered an aging society. Only about 7% of the population was older than 65 years and so demographic aging was not recognized as a serious social issue. During this period, the Korean government maintained a highly restrictive approach to social care. It only provided care services for those who did not have family members on whom they could depend. The number of recipients and the amount of social care spending were both negligible. The Older Persons Welfare Act and the Welfare and Social Welfare Services Act contained several clauses related to care services. These Acts offered both the central government and local governments significant de jure discretion. For example, in the 2005 Social Welfare Services Act, Article 41-2 noted that “the central and local governments may arrange home welfare service for the persons in need of protection . . .” Under the Acts, the central and local governments were officially granted the discretion to decide whether to provide a particular service and to decide who was eligible for such services, among other things.

However, the governments largely delegated service provision responsibilities to nonprofit welfare associations and did not engage in the management process. Before 2008, although the central government produced statutory guidelines for elderly care services, these guidelines were only about 15 pages long in 2001, later becoming about 400 pages in 2017.[47,48] The guidelines noted that “care services shall be given without charge to those who are not properly supported by family members.”[47] They also stipulate that the local government should decide who can receive services.[49] However, as there was no frontline care management function within local governments, users typically contacted nonprofit service providers directly. These service providers—specifically, licensed social workers within the nonprofit organizations—assessed care needs based on their professional judgment and ethics. Following this, clients had essentially no choice of service provider.[50] This model was largely maintained until 2008, when the MOHW provided general guidelines to local authorities, including contract specifications.

Since the early 2000s, rapid aging has become a significant social and economic issue in Korea. Experts warned about the inadequacy of social policies in terms of absorbing the risk of the increasing aging population under the legacy of past growth-oriented policies and the pressures of austerity. Meanwhile, the Korean government started by establishing the universal LTCI in 2007, which completely overhauled the elderly care system.[51] As services were extended to those with substantial care needs regardless of their income and the availability of family support, coverage expanded from below 1.0% of the elderly in 2007 to 6.5% in 2014.[52]

Second, to improve efficiency, the government deliberately embraced for-profit service providers and introduced a stricter regulatory framework. The government needed to build nationwide care infrastructure rapidly to provide services for those who were eligible. Thus, the previous reliance on only nonprofit organizations was no longer regarded as a viable option. Although nonprofit providers were generally considered reliable, the government was skeptical of whether the nonprofit and public sectors could meet the soaring social care needs, believing that they would be either too slow or too expensive. Thus, of the more than 14 000 home care providers in 2016, small-scale private providers accounted for more than 82%, whereas nonprofit organizations accounted for about 16% and public providers for less than 1%.[53]

The LTCI law remains open to interpretation; it contains basic principles but entrusts specific rules and regulations to the executive branch (ie, the MOHW). A clear illustration of this is the fact that the Korean LTC law is only 16 pages long, whereas the German LTC law is more than 100 pages long. In terms of the available benefits, Article 3 of the law notes that “Long-term care benefits shall be provided to the beneficiary in an amount that is appropriately within the scope of the beneficiary’s needs.” In addition, it states that the specifics for eligibility, benefit levels, and delivery methods “shall be determined by presidential decree” or “shall be determined by the regulations of the Ministry of Health and Welfare.”

Accordingly, the central government has the strong authority to specify the details of the LTCI law. Instead of delegating de facto authority to nonprofit organizations, it fully exercises its discretion. The MOHW specifies the eligibility requirements, the level of benefits, the wage levels of care workers, the contribution rate, and the licensing requirements for providers in detail through presidential decrees and guidelines. An individual’s eligibility and the benefits for which they qualify are standardized on a graded scale of 1 to 5 according to their dependency score. The MOHW provides a checklist that consists of 52 items and care investigators assess the level of dependency using this checklist. The care investigators do not have the discretion to write letters or choose documents that might influence the board’s decision such as summary statements or recommendations.[54]

The law and the MOHW do not give much room for discretion to local governments or street-level bureaucrats. The local governments participate in the Grade Assessment Committee to review and confirm care investigators’ assessments. However, according to the MOHW,[55] the Committee has only changed only around 1% of all grade assessment decisions. Besides, although local governments have some de jure discretion to regulate the local care market by issuing and canceling service providers’ licenses, they have hardly been an important policy actor in reality. As the MOHW provides fully specified standards, it actually allows little discretion to local inspectors as the national standards do not allow for a significant degree of interpretation for inspectors.

Unlike the previous system, the discretionary role of social workers in care management has completely disappeared. Their role has been replaced by that of care investigators from the National Health Insurance Corporation (NHIC), who strictly follow detailed guidelines from the MOHW. Service providers are expected to comply with these rules. After needs assessment, service users can choose between residential and domiciliary care services based on their dependency grade and they can choose a service provider. In addition, the government strictly bans care services provided by family members because of concerns about benefit fraud.[51] If elderly people want to receive care services from their own family, a family member should obtain an official care worker license. However, even in this case, care users are entitled to a lower level of benefits and limited services compared with standard care. Table 1 summarizes the discretion mix following the 2008 reform.

Table 1.

Discretion Mix for LTCI in Korea.

Decision-making level	Eligibility	Amount of care	Benefit type	Delivery choice
LTCI Act	1. Aged 65+ (§2)2. NHI subscribers (§12)Other conditions are delegated to the MOHW	Decisions are delegated to the MOHW (§28)	Decisions are delegated to the MOHW and service users (§24, §25)	Decisions are delegated to MOHW and service users (§31.2)

Note. LTCI = long-term care insurance; NHI = National Health Insurance; MOHW = Ministry of Health and Welfare; NHIC = National Health Insurance Corporation.

Reality and Consequences of the New Discretion Mix

With the introduction of LTCI, the government had to universalize LTC services without creating a significant financial burden in the face of a rapidly aging population. The centralization of regulation and the marketization of care provision are regarded as 2 essential pillars for achieving and managing efficient expansion. In this regard, the central government has developed very detailed guidelines for for-profit providers in the delivery system. Statutory ordinances specify the operational requirements for care homes, such as the frequency of staff meetings, minimum attendance rates, and the need to document meetings.[56] Furthermore, the ministry provides care providers with a very detailed manual for sterilization procedures for the various equipment types in care facilities.

However, quickly establishing a viable care market requires that the central government establishes very loose entry requirements for service providers without granting discretionary authority to local governments.[57] For example, a care worker requires only 240 education hours to obtain a license, which is much lower than the nearly 2000 hours required in Japan. In addition, a home care company can operate if it employs a care worker and has an office larger than 16.5 m2. The loose entry requirements are also related to the marketization trend that has developed since the 1990s.[2,8] During the period of welfare developments over the last 2 decades, market principles have been strengthened in LTC policies such as by promoting competition, outsourcing services to for-profit providers, and transforming older adults in care into consumers. Indeed, the government successfully created an LTC market by attracting profit-seeking providers, which led to steep competition. The number of home care companies in 2008, the first year for LTCI, was around 9900, but this rose to around 19 000 in 2009.[58]

Unfortunately, this strategy also immediately led to issues regarding the quality of care and provider misconduct. According to a survey by the MOHW, fraudulent payments to LTC providers increased from $9.1 million in 2011 to $16.7 million in 2014.[59] The major reason for this was that some providers misreported their number of personnel, which resulted in higher reimbursements from the government. In addition, it was reported that they distorted the rule-makers’ policy intentions and operated beyond their granted discretion. For instance, they intervened in decisions regarding user eligibility or benefit levels by advising applicants on how to deceive care assessors or how to attract customers through co-payment fee exemptions or gifts.[60]

To combat this, the financial penalties for provider misconduct were strengthened in a 2013 revision of the LTCI Act. The MOHW also announced new rules for financial and accounting management for private service providers in 2016, in which providers were required to report their financial and accounting records to the government using a public information communication technology (ICT) system.[54] In conjunction with this, the government strengthened the evaluation of LTCI institutions to ensure the quality of care. Following the ICT system’s introduction, providers had to report their activities and meeting documents in real time. These reforms all have newly added and reinforced detailed guidelines to enhance the quality of care.

The strongly centralized regulation caused another problem for service providers. According to them, cases of fraud have paradoxically increased, partly due to centralized monitoring and regulation. Two examples of this can be seen in the focus group interviews. First, if individual providers attempt to introduce an innovative management idea, they may easily breach the rules. In an interview, one manager of an LTC facility stated that she had to pay $4000 because she introduced a temporary-stay scheme so that service users could become accustomed to their care home before residing there permanently (Interviewee A, August 31, 2017). Second, service providers can be fined up to 15% of the total cost if a service user’s co-payment is late, and it is very difficult for them to submit evidence that users delayed payments. One service provider reported, “If I cannot provide evidence such as text messages, e-mails, or phone calls, I will be fined. It is often very difficult to collect evidence in daily practice. It also demands substantial effort and causes stress” (Interviewee B, September 5, 2017).

Regarding the new accounting regulations, one provider mentioned, “New rules are added all the time, but it is very hard to follow all of those changes . . . so I decided to selectively ignore them because I cannot do everything” (Interviewee C, September 1, 2017). Another manager asked,

Why can we not comply with the accounting requirements? To embezzle public money? No! We do not have an accountant . . . I attended the training program five times this year throughout summer and fall but it was too difficult. We really make every effort to keep up with those requirements, but they are nearly impossible to follow. (Interviewee D, August 31, 2017)

A final problem is that the tightening of inspections by the central government using the IT system has not reduced the implementation gap. A manager of a privately owned service center recognized the problem, stating, “Even if you provide very bad service, you can still receive a good evaluation grade as long as you are an expert at documentation” (Interviewee E, September 12, 2017). Some managers have complained that the state does not allocate sufficient resources to complete the ever-increasing administrative tasks and comply with inspections. One manager said, “Now there are increasing rules such as using specified products [and] paying more to care workers . . . but they do not give us the resources to implement them” (Interviewee F, September 1, 2017). These problems seem to have resulted in the widespread distrust of central bureaucrats: “Bluntly speaking, there seems to be a very stubborn pig-headed . . . very stupid and bone-headed old man sitting at the top. Someone who never listens to anyone else” (Interviewee G, September 1, 2017). It has been reported that some providers closed their facilities just before a major inspection by the government and reopened them afterward. The closure rate of residential care homes with fewer than 10 elderly residents exceeded 20% in 2014.[61]

As explained in the previous section, local governments have some autonomy to regulate the care market. However, local governments do not have much resources or incentives to regulate the market. According to focus group interviews with public officials, a number of social care services have been introduced without recruiting additional public officials or granting significant financial resources to them from the central government. One public official mentioned, “In the case of facility management, the guidelines are detailed, but it is physically impossible to inspect every facility. We just go to major facilities and formally write checklists” (Interviewee H, October 11, 2018). Another interviewee said, “The people who need moderate care are under the responsibility of the municipalities and the NHIC has been tried to develop referring system to them but it did not work” (Interviewee I, August 31, 2018). In the end, Interviewee J and other public local government officials commonly insisted, “We do not have enough personnel to administrate these programs . . . I just want all the social care services to be taken by the central government” (Interviewee J, October 2, 2018).

Evidently, for-profit providers’ domination of the care market has created many problems, leading to stronger government intervention. In this situation, the central government cannot positively respond to demands from providers to increase financial support. However, stronger government regulation without sufficient financial subsidies has made providers even more evasive and less responsible. In this respect, the vicious circle between centralization and marketization without expanding public resources and infrastructure has contributed to low quality of care. Furthermore, in spite of the strong centralization trend, according to governmental research,[58] regional inequality has become significant. Home care services are abundant in urban areas where many workers live, but residential care homes—which require high initial set-up costs—are seriously lacking. The situation is the exact reverse in rural areas.

In principle, service users have the discretion to choose their type of service and provider. However, as noted, the centralized evaluation system does not work properly and cannot eliminate information asymmetry in terms of the quality of care between service providers and recipients. In addition, depending on where the recipient lives, the choice of provider may be highly limited. This makes it difficult for consumers to “shop around.” Therefore, Korea’s LTCI system does not take advantage of the strengths of marketization.

Thus, the discretion mix for LTC services differs greatly from the welfare mix for service production. Since the LTCI’s enactment, many studies have revealed that the dominant role of private service providers, particularly for-profit service providers, is a key feature on the provisional side of the welfare mix. Indeed, publicly managed care facilities are uncommon. However, the discretion mix framework reveals that the LTCI’s governance is characterized by the dominating presence of the central government and the increasingly weaker role of private providers. With local governments and street-level bureaucrats providing a negligible bridging role, the 2 very different mixes generate significant tension and many problematic issues, and it seems that stronger regulations have been unable to diminish the implementation gap.

Conclusion and Implications for Comparative Studies

This research begins by finding that existing concepts are useful but not fully satisfactory for understanding the reforming process. Thus, this article has attempted to elucidate systematically the different levels of discretion that various policy actors have in their implementation of LTC services by introducing the concept of the discretion mix. Through empirical analysis, we have shown that the discretion mix for the Korean LTC care system has changed, and have outlined the consequences of the new discretion mix. The efficient expansion strategy pursues rapid marketization and an increasingly strict centralized regulatory framework; however, the gap between policy intentions and actual implementation has continued to grow.

As argued earlier, just as with the welfare mix, the discretion mix is the product of historical institutionalization. For instance, the dominant role of the central government in Korea compared with the parliament or local governments can be easily understood by considering the legacy of the authoritarian developmental state. Thus, this concept could be used fruitfully by analyzing the cross-sectional differences of the LTC services and their transformations. In this section, we explored the possibility of a comparative study here by depicting the discretion mix in German and Swedish social care services.

Germany introduced LTCI as a comprehensive benefits scheme for the frail elderly in 1995. The eligibility criteria were applied nationwide with a high level of specificity as stipulated in law (SGX XI §§14–15).[62] The Medical Review Board (Medizinischer Dienst der Krankenversicherung, or MDK) performs an assessment according to the law. There is almost no case management or social assessment that considers service users’ self-expressed needs.[6,63] Regarding the benefit type, LTCI offers discretion to its service users. Beneficiaries can freely choose between cash and in-kind benefits. The SGB XI prescribes that insurance funds make a contract with service providers and that insurance funds must accept every provider who fulfills the defined preconditions regardless of the market situation.[64] The selection of providers depends on the contracts with their own funds and service users must choose service providers on their own. The federal states are responsible for ensuring the LTC infrastructure to guarantee an adequate level of offers, and Länder and the local governments are supposed to have general responsibility for the service system. However, without specific roles compared with the MDK, their roles are limited. Through introducing the LTCI, the German case seems to embrace a centralized approach to supporting the elderly.[9]

In Sweden, the roles of the law and the central government in social care are less strong than in Germany or Korea. The Social Services Act stipulates that if someone cannot provide for their own needs, that person is entitled to assistance from the municipalities board of social welfare (§6). At the central level, the Ministry of Health and Social Affairs (Socialdepartementet) is responsible for developments in areas such as health care, social insurance, and social issues. However, the central government engages in the policy process in a limited range, such as by setting a maximum monthly fee and monitoring how the LTC systems work in Sweden.[1] As municipalities have substantial autonomy to prioritize political issues and make their own organizational arrangements, social care managers at municipalities have substantial discretion in determining service eligibility.[5] The Swedish care system also offers more discretion to service users regarding their choice of service providers. However, local authorities have more power to engage in users’ choices in residential care. Users can choose opt-out cash benefits instead of in-kind benefits. Still, the selection of providers depends on contracts with local authorities.

Both Korea and Germany are known for being under insurance schemes, but there is a crucial difference in the roles of their parliaments and central governments. Sweden gives more autonomy to local governments and street-level bureaucrats in care decisions. Discussing their historical backgrounds and the different kinds of discretionary problems in Germany and Sweden would go beyond the scope of this research, but the concept of the discretion mix could offer an understanding of LTC development from a different perspective.

Further research could investigate whether a certain type of discretion mix combined with a particular welfare mix or a different type of marketization produces a higher or lower quality of LTC services and why a particular care regime has adopted and developed a specific type of discretion mix. By complementing existing concepts such as the welfare mix, subsidiarization, and co-production, we hope that the concept of the discretion mix can contribute to an understanding of the transformation and quality of LTC services.