Delegation, deregulation, and business power: a comparative analysis of health insurance in Belgium and France

Abstract

Business power is thought to increase over time when private actors are involved in the provision of public goods and services. This paper argues that this is partially true—and that in certain circumstances, state actors can even swiftly regain control of sectors previously ceded to private interests. When the latter fulfill some public functions on behalf or as delegates of the state, policymakers face ever greater pressures to sustain a relationship flawed by principal-agent problems—allowing business actors to derive appreciable political benefits. However, these conditions do not hold true after deregulation—when state actors retreat from a sector and attempt to direct the newly created market through licensing, norms, and standard setting. We demonstrate that deregulation sets the stage for a more competitive environment, making it harder for private interests to cooperate. This, in turn, can allow policymakers to enhance regulatory capacities and seize opportunities to highlight the shortcomings of private provision. After establishing this argument theoretically, we illustrate its implications through the comparative historical analysis of the health insurance sector in two European countries—Belgium and France. Despite their initial similarities, they experience contrasting developments regarding the welfare state’s dependency on private insurers for the provision of crucial collective goods.

Introduction

In various situations, the political power of business actors does not primarily or originally stem from their direct efforts to influence government decisions. It also does not arise solely from the “privileged position” capital enjoys in capitalist economies. ¹ Instead, it emerges from an initially voluntary decision of policymakers to involve business actors in the provision of essential public goods and services. These notably include public–private partnerships, quasi-markets, or corporatist institutions endowed with important decision-making capacities. While such arrangement might serve widely different purposes, they share an essential commonality—in the sense that the “institutional power” they confer to business actors remains “endogenous to policy design choices” and expands in the long run in large part due to the feedback effects generated by such past policy decisions. ² According to an increasingly important literature in political economy, business power in such a setting could only growth at the expense of state actors, as the latter would gradually lose their capacity to tame private interests.

In this paper, we argue that institutional business power can be subjected to a larger variation over time than usually assumed, and suggest that such variation is correlated to the conditions through which private actors have acquired it in the first place. In short, we contend that the feedback effects that are set in motion by the acquisition of institutional power depends on whether it takes root through delegation or deregulation mechanisms. After establishing this argument theoretically, we illustrate its implications by analyzing how private actors have influenced the long-term developments of two major welfare markets: the health insurance sectors in Belgium and France. Private actors’ institutional power emerged differently in the two countries, and our analysis reveals that it has not been subjected to the same developments. While they have seen their position constantly increasing in Belgium (after important prerogatives were initially delegated to them), policymakers in France, years after the deregulation of the health insurance market, eventually succeeded in claiming lost territory. Contrary to a series of expectations commonly shared in the literature, this suggests that involving private actors in the provision of public goods might have contrasted implications for state–business relations, and, importantly, for policymakers’ capacities to regain control of a sector where possibly important functions have been left to private interests.

Theoretically, the paper provides slight, although we believe important nuances to a series of considerations initially formalized by Busemeyer and Thelen. In a nutshell, their fundamental argument—discussed in more detail below—is that institutional business power, driven by policy feedback and lock-in effects, contributes to an asymmetrical dependence of the state resulting in an increasing power enjoyed by private actors over time. While numerous contributions make this inference, this paper starts with the premise that there are compelling reasons to believe that such an increase is neither necessary nor systematic. Given its reliance on policy design choices, one can reasonably expect different initial conditions to lead to varied outcomes. If this intuition proves true, we should observe cases where institutional power progressively grows, as well as cases where it does not.

It is with verifying and specifying the conditions for this basic proposition that the paper is more directly concerned. After providing a brief introduction to Busemeyer’s and Thelen’s argument, we proceed to identify and briefly review a series of contributions that highlight the diverse implications of institutional power for the enduring relationship between the state and private actors. On this basis, we suggest that business’ fluctuating fortunes might be due to the specific feedback effects activated by different modes of acquisition of institutional power. In doing so, we construct a theoretical argument that aligns consistently with a range of existing accounts in the literature, while also allowing for a broader range of potential outcomes.

Formally, we posit that institutional business power is more likely to increase following delegation, where private actors assume responsibility for providing certain public goods or functions on behalf of the state. Conversely, institutional business power is more susceptible to decrease following deregulation, which entails the state’s partial or complete retreat from a sector, accompanied by the creation or maintenance of specific regulations or oversight mechanisms. We argue that each mode has distinct implications, especially regarding three critical factors. The first pertains to the structure of competition among these actors. Depending on the mode of acquisition of institutional business power, private interests may encounter varying levels of difficulty in coordinating their efforts to prevent both market entry by competitors and state intervention in their activities. The second factor relates to the long-term evolution of state capacities relative to private actors. Lastly, we suggest that policymakers’ initial design choices can significantly influence the political costs and incentives they face when trying to regain control from a sector.

Our comparative historical analysis of health insurance in Belgium and France serves as an illustration of this argument and provides several insights to envision its more general implications. Over the past thirty years, the role of private actors in providing health insurance or occupational welfare benefits to firms and individuals has significantly grown in both cases—yet these two “conservative-corporatist” welfare regimes ³ have followed different paths. In Belgium, private health insurance essentially developed during the 1980s through an extension of the delegation that historical players in the field were already enjoying as part of the statutory healthcare system. During the same period in France, policymakers actively designed a market for private health insurance through a series of deregulations. The share of healthcare coverage supported by private insurers in the two countries increased as a result; but while recent policy developments in France suggest clear (and consequential) attempts from policymakers to take back control from these actors, the power of health insurers has continued to growth in Belgium.

Using process-tracing techniques informed by qualitative data coming from diverse sources, we find that these contrasted outcomes are, as anticipated, largely due to the consequences of the different modes through which institutional power initially emerged. In Belgium, the dominance of a handful of private insurers as a consequence of delegation has impeded new competitors from entering the market, allowing insiders to coordinate and exert an ever-increasing influence on state actors reluctant to risk any disruption in the provision of healthcare coverage. In turn, this situation has blurred the distinction between social and private insurance and has shaped popular attitudes toward the public–private mix in Belgium. At the same time, in France, the deregulation of the health insurance market has allowed diverse players with varying business models and market segments to compete. Consequently, French policymakers could assert multiple policy objectives in a politically divided industry while keeping some distance from the market. Several years later, as the delineation of responsibilities between public and private actors became more salient, they found themselves in a better position to impose stricter controls on health insurers by effectively politicizing the limitations of private provision. Taken together, these findings suggest that the acquisition of institutional business power does not necessarily preclude the possibility of a (potentially significant) return of the state in formerly privatized or marketized areas. More importantly though, they also indicate that the structure of the resulting win-sets is likely to align to the design choices by which institutional power initially emerged.

The rest of the paper is organized as follows. A first section introduces the basic arguments of Busemeyer and Thelen and provides a concise discussion of several contributions on institutional business power. We develop on this basis a series of theoretical propositions to account for the variation of institutional business power over time, that we test in a second section through the comparative analysis of the health insurance sector in Belgium and France. A concluding section discusses the broader implications of our findings.

Institutional dependencies, state capacities, and business power in advanced capitalism

Tacking stock of the variation of institutional business power over time

The literature on business and politics conventionally assumes that business actors’ ability to influence the policy process arises from two analytically distinct sources. One is usually labeled as “instrumental.” It manifests in business actors’ “ability to staff governments with supporters and to exert direct influence on government decision-makers through campaign contributions and lobbying efforts” (Hacker and Pierson, 2002). The other is termed “structural,” as it derives from business actors’ “privileged position” in contemporary capitalism. Investments that are necessary for the well-functioning of a capitalist economy are indeed largely at the discretion of private capital holders. In such context, policymakers are strongly incentivized to maintain private sector profitability and protect business activities given their broad impact and repercussions on the political economy – something which is often achieved without the necessity for business actors to directly exert pressures on policymakers (Culpepper, 2015). Most contemporary perspectives treat structural and instrumental business power as independent variables rather than constant. More significantly, they recognize the hybrid nature of structural and instrumental sources of business power, noting that they often complement each other (see Babic et al., 2022 for an extended discussion).

In the context of these debates, Busemeyer and Thelen have recently highlighted the existence of a third source of business power. It emerges when business actors are involved in the provision of collective goods either on behalf of or as substitutes for the state. This third source—labeled “institutional” business power—is formally defined as the “power flowing from the entrenched position of business actors in the provision of essential public functions or services”. In such a context, institutional power does not flow directly “from a firm’s position in a market economy” (unlike structural power) or primarily as a result of its lobbying efforts or direct pressures on government officials (unlike instrumental power). Instead, it primarily arises from initially active policy decisions that “either invite or allow private interests to play a central role in the provisions of goods on which society crucially depends”. ⁴

In contemporary capitalist societies, occurrences of institutional business power are both numerous and extremely varied. For example, corporatist institutions that are typically found in Europe’s coordinated market economies constitute a prominent instance of institutional power—as business associations, along with state actors and trade unions, are here granted a crucial role in economic governance. ⁵ However, institutional business power is not exclusive to a particular type or “variety” of capitalism. In many countries, market reforms in public services, such as education, health, and elderly care, have often resulted in diverse combinations of public and private interventions. This, in turn, may have significantly enhanced the influence of business actors like insurance companies, private care providers, and large corporations. ⁶ Arguably, the trends toward liberalization and marketization observed in virtually all advanced capitalist societies have made this source of business power ever more significant in the recent decades.

As it highlights the enduring political significance of feedback effects inherent in policy design choices, the notion of institutional power evidently echoes familiar interpretations of historical institutionalist scholarship, which emphasizes the impact of “diverse institutional arrangements that facilitate the realization (or not) of longer-term economic interests”. ⁷ In this context, the term can be viewed as characterizing a particular state–business relationship configuration. ⁸ What is indeed significant in the examples mentioned above is that business power does not primarily fluctuate over time due to the mechanisms usually emphasized by structuralist interpretations of business power (in short, the natural operations of the capitalist economy) or by instrumentalist conceptions such as private firms’ political activism. Of course, the three main sources of business power can empirically complement and enhance one another. Instrumental, structural, and institutional power alike can make business actors more successful in their attempts to direct the behavior of policymakers, obtain favorable decisions, prevent certain ideas or issues from entering the political economy, or shape the concepts, standards, and measurements used by government organizations. ⁹ Yet, this third source remains distinctive as it primarily arises as a consequence of an initial voluntary decision on the part of state actors to share public responsibilities. As a result, and due to this peculiarity, it is likely to enhance the varied forms of influence of business actors in policy-making through distinctive mechanisms, as discussed further throughout the paper.

While the appearances of institutional business power may thus substantially differ, Busemeyer and Thelen argue that such power is generally acquired through three generic modes, illustrated in their contribution by a series of case studies drawn from the education sector. The first mode—delegation—designates situations where private actors directly fulfill some public functions on behalf of the state, possibly in close collaboration with state actors—as exemplified by the case of vocational training in Germany. Deregulation is a second mode through which the state establishes a market and allocates rights or public resources to private actors (such as through licensing, marketing approvals, subsidies, or tax exemptions), contingent upon their adherence to predefined rules and regulations. This mode, which is therefore distinct from full-fledged privatization, is illustrated by Sweden’s “controlled school market”. ¹⁰ Both delegation and deregulation thus encompass firms providing crucial public goods and services. They differ in that delegation involves firms acting directly as delegates of the state (like the training firms in Germany) while in deregulation, firms compete in a market with standards and rules set by state actors. Lastly, private actors can also acquire institutional power through accretion, namely “by taking the initiative to move into a policy arena previously dominated by the state, or into an emerging arena where the role of the state is still limited to gradually assume a central intermediating role” ¹¹ —a case exemplified by the rise of venture philanthropists in US education.

Institutional business power thus comes in variety, yet all these manifestations share a commonality: private actors acquire a significant ability to influence policymaking in specific areas by providing essential public and collective goods that society relies upon. Partly as a result, Busemeyer and Thelen argue that these modes activate broadly similar feedback effects. Once formally involved in the provision of public goods and services, private actors would become key participants in the policymaking process within a given sector. In addition to discentivizing policymakers to maintain their own capacities to deliver the goods and services in question (typically reinforced over time by classical path dependent dynamics), this would enable private actors to effectively shape future policy developments. As a result, policymakers “loathe to risk any disruption in essential public services” will face “ever-stronger incentives to attend to the interests of business in order to maintain private actors’ commitment to keep up their side of the bargain”. ¹² In turn, the stabilization of any public–private arrangement over time is also very much likely to result in feedback effects at the mass public level—with the general public becoming increasingly accustomed to the new division of labor between public and private actors. In sum, institutional business power, once established, tilts state dependence asymmetrically, increasingly favoring private interests over time.

These general arguments are actually consistent with the findings of a large body of scholarship in political economy. Public sector privatization and procurement contracts, involving enduring participation by private actors, have been linked to increased business power relative to state actors. ¹³ Similarly, the delegation of welfare service provision to private companies has been shown to create control issues for state actors, ultimately benefiting these firms. ¹⁴ Recent studies even suggest that once established, institutional power may account for a greater share of business actors’ influence than their efforts to directly instrument policymakers’ decisions or due to their structural position in the economy. ¹⁵ However, in most cases, the different sources of business power are interconnected, leading to “infrastructural entanglement” that durably hinders the capacities of state actors. ¹⁶

Overall, many contributions thus support the idea that institutional power increasingly benefits business interests over time. However, a smaller but significant body of research presents different outcomes. These variations are often seen in case studies where institutional business power gradually decreases, occasionally resulting in full de-privatization or renationalization of economic sectors. Populist executives in Hungary ¹⁷ and Poland ¹⁸ have for instance undergone a wave of re-nationalization of foreign-owned domestic banks in recent years. While their strategies have relied on domestic private actors, the literature has also documented cases where the involvement of public actors was less equivocal. Australia and New Zealand, both pioneers of rail deregulation, have recently seen government take back ownership and operations in the sector, reversing privatization. ¹⁹ Similar cases of reverse privatization were also reported in other areas and varieties of capitalism, such as in the field of waste collection in Germany. ²⁰ Consequently, empirical support for a continuous increase in institutional business power coexists with suggestions of more temporal variation than typically assumed.

The varying implications of delegation and deregulation on institutional business power

That institutional business power can vary more than usually expected is, to a certain extent, unsurprising. Both instrumental and structural power are also subject to potentially significant variation. ²¹ But if institutional power might also result in more or less ability of business actors to get their way in the policy process, what factors are likely to cause such variance?

At first glance, it seems that sectoral differences as well as broad institutional characteristics of capitalist economies are of a relatively limited explanatory purchase here. That institutional business power might decrease or increase over time was indeed observed in widely different sectors, and such variation was shown to be possible within dissimilar political economies. Taken together, the studies mentioned in the previous section however suggest that the mode of acquisition of institutional business power could play a significant role. Notably, these contributions highlight a crucial distinction in the developmental trajectories of institutional business power, depending on whether it emerges through delegation or deregulation mechanisms. Moreover, the findings—whether indicating an increase or decrease of institutional business power over time—converge within cases belonging to the same mechanism, but diverge markedly from those where institutional business power arises through alternative means.

More specifically, the propensity that institutional business power grows over time seems to be higher where it was acquired through delegation, and more likely or susceptible to decrease where it was acquired after deregulation. For instance, the recent literature on state–finance relations, that focuses on the implications of monetary policy’s reliance on dense private finance infrastructures, typically belongs to the first category. In this case, shadow money and shadow banking have become the central governance infrastructure for central banks due to an initial decision to delegate crucial functions to private actors operating in financial markets, leading to their empowerment. ²² In contrast, greater competition in the rail sector was promoted in Australia and New Zealand through a process of deregulation. ²³ Twenty years later, it is a process of de-privatization that has been observed in these liberal market economies—as well as in others (like the UK) that previously undertaken similar initial moves. Numerous contributions report similar trends—namely toward an increase of business power for those where institutional power was acquired through delegation (as notably in the corporate welfare literature ²⁴ )—and to a decrease of business power (possibly paired with government’s reclaiming full control of the sector) for cases of deregulation (as in some of the cases studied by Gingrich ²⁵ and in the area of the provision of local public services in Germany). ²⁶ Therefore, there are compelling reasons to believe that institutional business power varies depending on the initial means through which private actors have acquired it. And indeed, a closer examination of the feedback effects associated with each mode reveals significant differences. ²⁷

Consider delegation first. In this model, the state and private actors collaborate in providing certain public goods and services. A notable example is vocational training in Germany, as mentioned by Busemeyer and Thelen. This system combines theoretical education in vocational schools with practical training in firms. Like bureaucratic actors, firms in this context have a broad legal mandate outlined in the delegation of authority through legislation, which effectively creates a “controlled agency”. ²⁸ Given these initial conditions, delegation is very likely to activate the feedback effects theorized by Busemeyer and Thelen through three main mechanisms.

The first involves private actors, as delegation typically requires active participation from a stable group of firms, explicitly mentioned in the delegation contract. ²⁹ This often restricts market entry and creates higher barriers for potential newcomers. As a result, it reinforces state actors’ dependence on a specific group of private companies that play a more decisive role and can coordinate more effectively, thereby strengthening their influence in the realm of “quiet politics” of sectoral policy debates. ³⁰ Crucially and as Busemeyer and Thelen rightly point out, these actors however remain motivated to participate as long as it is in their interest, thus potentially increasing the exit threat and strengthening institutional power.

The second mechanism pertains to state capacities as induced by the governance structure. Indeed, delegation involves a division of activities between the state and private actors, often leading to the dismantling of the state’s independent delivery system and eroding its capacities in this domain. This aligns with an argument of Busemeyer and Thelen, particularly relevant in the case of vocational training in Germany, where the state has gradually lost its capacity to establish alternative public training facilities, especially on a large scale. The cost of doing so has in fact increased as business actors have become more entrenched in the sector. ³¹

These first two mechanisms also interact significantly with a third mechanism related to the sanctioning process accompanying delegation. In such setting, policy efforts are directly tied to political representatives, who are in turn accountable for the performance of their private agents—just as they are for any actors in the bureaucracy. In other words, policymakers remain directly responsible for the provision of public goods and services under delegation, even if the providers are private actors. This is likely to trigger feedback effects at the mass public level because the continued presence of a means for directly holding electoral representatives accountable is likely to enhance acceptance of this division of labor between public and private actors.

Our discussion of delegation suggests that the mechanisms identified by Busemeyer and Thelen are likely effective in this mode. Arguably, the same does not applies to deregulation. In such cases, state actors more directly relinquish the provision of public goods and services to private companies. However, policymakers commit themselves to ensure the enforcement of rules and general principles, such as compliance with the standards, norms, and other governmental regulations as seen in the case of Sweden studied by Busemeyer and Thelen. This is often accomplished by appointing a specialized regulatory body.

While we observe in this model the interaction of the same mechanisms as for the case of delegation, they are likely to play out differently. First, deregulation usually does not create the same conditions for private actors, as it is expected to enable all providers who can adhere to predefined rules to participate. Beside cases of capture, deregulation is thus significantly less directed toward specific firms and by design, it cannot create disproportionate advantages for some actors at the expense of others without immediately conflicting with its initial objectives, namely the establishment of a more competitive environment. ³² As a result, policymakers often confront a more diversified array of firms, and the existence of lower entry barriers than in the case of delegation prevent insiders to forge robust and enduring coalitions. While it generates a comparable necessity for state actors to maintain their ongoing involvement in the provision of public goods or services, deregulation offers them more room to leverage divisions among market participants. In turn, firms depend on state actors for marketing authorization, subsidies, or tax exemptions, thus lowering their exit threat— as acknowledged by Busemeyer and Thelen themselves in their case study on Sweden.

The governance structure on which deregulation is typically based—that closely corresponds to Bertelli’s “independent agency” model—has also crucial implications for the second and third mechanisms. In theory, deregulation does not impact state capacities in the same way as delegation because it does not entail a complete division of labor between the state and private actors. Instead, the state must maintain robust capacities to define unilaterally ³³ the rules, standards, and principles that firms must adhere. While it may also reduce the state’s administrative and organizational capacities for delivering goods and services, the ongoing adherence of participating actors to these regulatory requirements triggers positive feedback effects at the firm level. Indeed, this typically results in modifications to their core business models, strategies, and operational approaches, creating the complementarities that may facilitate, in certain circumstances, a closer integration of these firms to the state.

While not necessarily the most likely scenario, this outcome is more credible under deregulation than under delegation, primarily due to the third mechanism—the sanctioning mechanism. In deregulation, political actors are held accountable for the performance of the regulatory body overseeing the market, rather than the performance of individual firms. This apparently minor difference with delegation has important practical consequences. First, political actors have incentives to manipulate market rules or the regulatory agency, drawing from the government’s stock of political capital gained from a supportive electorate for efficient regulation. ³⁴ Policymakers can also more easily shift public blame for a contested market to the regulator or to another agent. ³⁵ More broadly, they can more readily denounce the failure of private provision and thus, present de-privatization as a credible alternative to the current market-based arrangements as they are not sharing responsibility for the provision of public goods and services. Any failure of the system, in this context, might be realistically presented by policymakers as a failure of the market itself.

Overall, these differences suggest that the feedback effects that deregulation is likely to activate may differ from those associated with delegation—with direct implications for the variation of institutional business power. To be sure, we are not claiming here that the distinct feedback effects linked with each specific mechanism will invariably result in an increase (under delegation) or decrease (under deregulation) of institutional business power. ³⁶ Rather, we suggest that these two modes do not create the same initial conditions for state and business actors; and that without additional intervening factors, such conditions should logically lead to different outcomes. The next section provides an empirical illustration in support of this argument.

Delegation, deregulation, and institutional power

A comparative historical analysis of the health insurance sector in Belgium and France

A review of the literature on institutional business power suggests that it can be subjected to greater variation over time than usually assumed—and theoretically, we have argued that how such power is acquired could explain a significant share of the variation in question. Table 1 provides a synthesis of our expectations with respect to the effect of the mechanisms discussed in the previous session.

Table 1. Summary of the main mechanisms

In this section, we undertake a comparative historical analysis of the developments of the (private) health insurance sector in Belgium and France to better substantivize and illustrate this argument. We focus on healthcare in two continental European countries as an instructive case study, as it constitutes a long-standing and particularly well-developed area of governmental intervention where public provision accounts (and this by far) for the largest share of total expenditures. Yet, the growth of the public fraction in health care spending (which has continuously expanded since the postwar era) has come to an end during the 1980s in most, if not all European countries and has not improved ever since. ³⁷ In an overall context where demographic factors are putting strong pressures on healthcare budgets (combined with additional demands in the realm of policymaking for limiting the increase of public expenditures), various governments have tried to organize (if not to develop) private health insurance markets. This trend has been more particularly pronounced in “conservative-corporatist” welfare regimes, ³⁸ that have seen the emergence of voluntary, typically “complementary” private health insurance schemes—thus reimbursing goods and services that are not or no longer (fully) covered under the statutory health care system. The rise of private insurance in these cases is largely explained by the fact that in these countries, preexisting social insurance schemes (mostly funded upon employer and employee contributions) already offer the initial conditions and institutional complementarities for private insurance to prosper. In turn, sharing healthcare coverage with private actors was typically implemented through various public–private arrangements, with state actors holding important capacities to control markets closely intertwined with social insurance schemes.

We observed these developments in both Belgium and France, that we selected due to the broad similarities between their healthcare systems. However, private health insurance emerged differently in the two countries, leading to distinct long-term implications. This serves as the second motivation behind our choice of these cases.

Healthcare systems in Belgium and France are usually classified as as “Etatist” social insurance systems where the state holds the regulatory power but grants privileges for the financing and provision of health services to quasi-public entities. ³⁹ In both countries, employer and employee contributions account for a large share of health expenditures for statutory healthcare coverage, which is organized around a national, social health insurance scheme. Social insurance per se is managed through sickness funds with their own health facilities in France (caisses d’assurance maladie), and by five national alliances (landsbonden) of mutual benefit societies (mutualiteiten or ziekenfondsen) in Belgium that play a similar role—plus two additional (much smaller) sickness funds, including one specifically restricted to the employees of the Belgian railroad company. This overall architecture is a legacy of Belgium’s consociationalist model ⁴⁰ ; however, mutual benefit societies no longer mirror the internal—ideological, social, and political—divisions upon which they were initially built. Affiliation with a mutual benefit society is indeed both free and compulsory in Belgium, and social contributions are identical for all funds (as well as the benefits they offer).

These two countries have also seen a significant development of private health insurance over the past forty years. While it has existed before (most notably in France), it is from the 1980s onwards that the sector more substantively developed through a series of policy interventions that have delineated the perimeter of the public–private mix in healthcare for the decades to come. Private insurers have seen their institutional power formally recognized and expanding as a result—although through different mechanisms, namely after a tacit (then more explicitly formalized) delegation in Belgium; and through deregulation followed by the design of a private health insurance market by policymakers in France. These initial choices constitute the starting point of our empirical analysis. We trace the processes that have unfolded until the most significant recent policy developments, “placing the study of power in time”, and paying equal attention to episodes of non-decisions and open contestation. ⁴¹

Methodologically, our study follows a process-tracing approach that we use to illustrate how our theoretical assumptions “work in the real world”. ⁴² Specifically, we seek to evaluate whether the causal mechanisms described in the previous section are explaining our outcome of interest. ⁴³ In this regard, we specifically focus on the differential combination of the three mechanisms induced by each governance structure. This includes how competition among business actors, changes in state capacities, and the political incentives of policymakers affect the variation of business power over time. In sum, and after describing the initial event (the acquisition of institutional business power by private health insurers), we assess whether the expected outcome occurred due to the previously identified mechanisms. ⁴⁴

To arrive at our conclusions, we combined various primary and secondary data sources. The main sequences and policy developments have been first identified through the consultation of the (primarily historical) literature about the health insurance sector in the two countries, especially for the period that runs from the second half of the 1970s to the mid-1990s. In addition to these academic sources, we also collected official documents providing exhaustive syntheses of healthcare reforms conducted in France and Belgium over the past 50 years. ⁴⁵ Once the main critical reforms for the development of private insurance in the two countries were identified, we then collected a large number of policy documents, press articles, and public reports about the situation of private health insurance during each of these reforms, sometimes complemented by more targeted archival research. ⁴⁶ A similar strategy was followed for the more recent sequences (namely after 2010)—although our analyses were here informed by a larger number of primary sources due to the greater availability of public and private documents and data about private health insurance, its coverage, and the motivations behind the reforms affecting its regulation. We also collected second-hand data about the perception of the role of mutual benefit societies in Belgium and France in the general population, relying mostly on the repeated comparative surveys conducted in the two countries by the Solidaris Institute (Belgium).

For both of these sequences we also base our inferences on the reanalysis of approximately 60 interviews with key policymakers and private actors conducted during four separate interview campaigns. As it is often the case with reanalysis, the primary goal of these interviews was relatively independent from the questions that motivate the present inquiry. ⁴⁷ Yet the richness of this material provides crucial information for the understanding of the developments discussed below. The first two interview campaigns, conducted between 2018 and 2019, involved private actors and policymakers in the two countries, offering vital insights into the analyzed policy developments. The two subsequent interview campaigns (conducted between 2021 and 2023) primarily focused on private actors, providing information about the health insurance sector’s industrial organization, the shared and differing positions of these firms in the policy process, and the factors influencing their representatives’ stances. ⁴⁸

Accounting for private health insurers’ varying institutional power over time

Two distinct modes of acquisition of institutional business power

This section traces the emergence and consolidation of institutional business power. We assess the preliminary effects of the three identified mechanisms, beginning with the case of Belgium and then examining the French case.

In Belgium, private health insurers initially acquired institutional power through a form of delegation. These developments occurred as a result of pressures that are hardly specific to this case. During the 1990s, the two coalition governments, led by Prime Minister Jean-Luc Dehaene of the Christian-Democratic Flemish Party (Christen-Democratisch en Vlaams), introduced stringent measures to curtail public spending, in large part to meet the EU’s Maastricht convergence criteria. ⁴⁹ These measures were implemented alongside significant healthcare system reforms, typical of other conservative-corporatist welfare regimes, aimed at enhancing federal-level regulation, reducing healthcare costs, and minimizing the impact of social contributions on the country’s competitiveness. ⁵⁰ In 1995, key legislation was enacted to enhance the organization of the system, ⁵¹ followed by a series of more or less ambitious initiatives to contain health expenditures through macroeconomic policy. ⁵²

In this context, mutual benefit societies, which already provided healthcare coverage to the majority of the population as sickness funds, began offering members complementary healthcare benefits through private insurance. These benefits typically covered services and expenses that were no longer or only partly covered by the statutory healthcare system. ⁵³ Private health insurance in Belgium thus initially emerged as a result of an initiative from private actors themselves. It was however rapidly acknowledged and tacitly encouraged by policymakers, who saw it as a way to make less directly perceptible their austerity measures. Private health insurers’ institutional power thus developed through an initially informal, yet increasingly consequential, model of delegation.

In the early stages of the scheme, developing private insurance in exchange of reduced capacities seemed neutral to political actors as it did not lead initially to a loss of control. When the government agency responsible for the (public) health and disability insurance— the Rijksinstituut voor Ziekte en Invaliditeitsverzekering—determines the range of benefits covered and the level of spending, it de facto delineates the contours of the complementary benefits and services offered by mutual benefit societies. Through their allocation decisions, policymakers can thus constantly shape the development of both the demand—with the rise of private insurance being almost perfectly correlated with the reduction of statutory coverage—but also the supply of private health insurers. More importantly, this model provides them with appreciable political benefits. Since a 1994 law, the annual growth of health expenditures in Belgium is indeed limited to 1.5% on real growth (above inflation). To compensate for the chronic funding shortfalls that have emerged, they can rely on private insurance to generate additional revenues for funding technologies, services, and benefits. ⁵⁴ Crucially, the dual role of mutual benefit societies (part of the statutory healthcare system and providers of private insurance) obscures the actual sources of these cuts and the resulting higher costs for households.

All of these developments, however, proved rapidly consequential in terms of the rise of health insurers’ capacities to shape health policy making and its outcomes. The total amount of premiums paid for private health insurance rose from five billion Belgian Francs in 1987 to 12 billion at the end of 1995, and has continued to grow ever since. ⁵⁵ The benefits from the growth of private health insurance were concentrated among a very small number of actors, as only mutual benefit societies already involved in the statutory healthcare system were involved by policymakers. These actors also gained increased negotiating power with healthcare providers as part of their private insurance activities, without state oversight. Since all successive governments until the late 2000s considered social and private insurance as the same activity (partly to obscure the actual effects of privatization), the general public became gradually accustomed to these actors’ expanding roles and the growing benefits they offered, without clear state intervention on the private segment of their activities.

During the same period, similar (and commonly interpreted) pressures have also resulted in a significant rise of private health insurance in France. The existence of the sector predates these more recent policy developments—as non-profit health insurers such as mutual benefit societies or provident institutions have also existed in France long before the formation of its welfare state. ⁵⁶ Contrary to the case of Belgium however, these actors never formally participated in the statutory healthcare system built around employer and employee representatives. While formally recognized by a 1947 law, ⁵⁷ the private health insurance sector (where non-profit health insurers and notably, mutual benefit societies were enjoying a dominant position) has remained highly limited in the context of the significant development of the healthcare system. From the mid-1980s onwards, policymakers concerned with rising healthcare costs however actively designed a private health insurance market as a solution to the (anticipated) stagnation and possibly, cutbacks in healthcare budgets. These initial choices triggered a dual movement of privatization of health expenditures, combined with an increasing regulation of the emerging market—thus constituting a typical case of deregulation in the sense of Busemeyer and Thelen.

Most of these developments occurred after the Socialist Party’s (Parti Socialiste) accession to power in the early 1980s. After an initial “social-democratic moment”, the party progressively embraced a more moderate economic policy orientation and committed itself to a rigorous management of social expenditures ⁵⁸ —and thus anticipated a gradual reduction of the benefits covered under the statutory healthcare system. This context was perceived by policymakers as highly conducive for the rise of private health insurance, that they sought to preemptively regulate. After some initial debates in the early 1980s (then paused by the right’s return in office in 1986) a law was finally enacted in 1989—with the explicit aim of creating the conditions for free and fair competition on the emerging private health insurance market.

Before that, several options were considered and debated both in the government and within the Socialist Party itself. ⁵⁹ Initially and during the early 1980s, Pierre Bérégovoy (then Ministry of Social Affairs) pushed for the creation of a private monopoly for complementary health insurance—and for delegating such monopoly to mutual benefit societies that were perceived, as non-profit insurers with a historical record of promoting access to care, as sharing similar values and principles to those of the Social security system. ⁶⁰ Several senior civil servants close to the Socialist Party stood on the other side of the debate, and notably François Mercereau—the head of the Social Security Directorate ⁶¹ . They presented mutual benefit societies as unable to direct such a monopoly given their fragmentation (almost two thousand mutual benefit societies, often very small, were existing at that time). Additionally, the proximity of some mutuals to the Communist Party further complicated (this time politically) their ability to direct such a monopoly. The benefits of free and fair competition, notably with more efficient and larger for-profit insurance companies, were also emphasized at a time were the government was divided over the course of its economic policy. ⁶² Lastly—and crucially—the prospect of greater European economic integration, that was gaining in prominence in the government’s agenda, was also presented by Jacques Delors (then Ministry of the Economy and Finance) as an additional obstacle to the creation of a monopoly. ⁶³

In 1989, the market option finally prevailed—and the so-called Loi Evin harmonized the rules and regulations applicable to private health insurers. ⁶⁴ The law however makes a series of compromises that reflect some positions expressed in the debates. While the idea of delegating the management of private health insurance to mutual benefit societies was abandoned, the initial rules and regulations of the market greatly advantage these actors. A number of mutualist principles—such as the prohibition of risk selection and an overall high degree of protection of policyholders—serve as the basis for the regulation of health insurance contracts. These principles (and their regulation by a newly created committee) will thus lead mutual benefit societies to dominate the market for the two decades to come. ⁶⁵

Consequently, private health insurers effectively gained institutional power, here as an anticipation of future reductions in healthcare budgets—that were effectively reduced during the next following years by successive governments, with private expenditures now covering around 14% of healthcare costs in France. The differences are however palpable with the case of Belgium. While the initial regulatory principles governing the health insurance markets favor mutual benefit societies, they also establish the conditions for open competition, allowing any actor willing to comply with these principles to enter the market. It would be wrong to assume that state actors retain the capacity to provide the goods and services now formally offered to private insurers. However, they maintain bureaucratic knowledge about delivery through the creation of a specialized regulatory body to oversee the market—the Control Commission of Mutuals and Provident Institutions (Commission de contrôle des mutuelles et des institutions de prévoyance). The same developments also led to a more explicit division of labor between the state and private actors.

The long-term implications of delegation and deregulation on business power

The early developments of institutional power among private insurers activated similar mechanisms, albeit with differing interactions in the two cases. In this section, we assess the long-term feedback effects of these early choices, demonstrating an increase in private insurers’ power in Belgium due to delegation and a decrease in France resulting from deregulation.

In Belgium, mutual benefit societies’ institutional power expanded gradually due to their increased capacities relative to state actors, growing popular support, and their ability to prevent new entrants from challenging their dominant position—all aligning to the expected consequences of delegation.

Mutual benefit societies indeed began to play an increasing role in the provision of additional public goods and services, a development that policymakers largely failed to anticipate. They faced increasing difficulties in directing the growing offerings by these societies due to their dependence on them within the statutory healthcare system and their growing political influence in absorbing and mediating cost-containment measures. During the 1990s, private insurance expanded within the existing delegation to include an ever-greater array of benefits and services such as medical transportation, home care, equipment leases in case of disability, or reimbursement of homeopathic medicines—with significant variations in terms of the benefits offered across providers. ⁶⁶ Mutual benefit societies also grown in size, sometimes through creating satellite entities dedicated to the provision of specific benefits. This was notably the case of the francophone entities part of the alliance of Christian mutual benefit society (Landsbond van Christelijke Mutualiteiten), which in the early 2000s was covering almost half of the Belgian population. ⁶⁷ They created a whole complementary insurance service (Solimut) providing co-payments, supplements, drugs, and medical devices to their members. ⁶⁸ This service was offered on the basis of modest, family-based monthly contributions and without age or medical restrictions. Mutual benefit societies also used private health insurance to provide additional benefits to their members, often unrelated to those covered under the statutory healthcare system—including childbirth benefits or financial support to join a sports club.

Through these kind of services, mutual benefit societies thus further reinforced their status of key providers of social care among the general population, as confirmed by multiple survey evidences. Before the Covid-19 pandemic, a representative sample of the Belgian and French populations was asked to react to a series of propositions about their perception of mutual benefit societies. Approximately 73% of respondents in Belgium agreed that these actors were actively working to improve their quality of life, whereas only 59% of respondents in France expressed the same sentiment. ⁶⁹ Thus, such developments were never really subjected to an intense contestation in the political realm. According to the data gathered by Martin and Benoît (2022), around 90% of the Belgian population was formally covered by one of mutual benefit societies’ complementary health insurance schemes at the end of the 2000s. ⁷⁰

Yet these developments resulted in increased business power in relation to both state actors and potential competitors, as evidenced by the largely unsuccessful attempts of for-profit insurance companies to challenge the rules of the delegation contract that bind mutual benefit societies to state actors. Since the mid-1990s, insurance companies are indeed trying to expand in health insurance. This is the case, for instance, of DKV (that belongs to the German Munich Re), which offers contracts covering hospital expenses, dental care, and various benefits in the area of occupational welfare. While not being a highly profitable market for large and transnational insurance companies, healthcare however constitutes a powerful tool for them to attract new customers—as well as way to expand on short-term, less risky activities in the context of stricter European prudential rules and regulations. Yet, such development was largely impeded in Belgium due to mutual benefit societies’ de facto monopoly, a situation that will trigger an important mobilization of the insurance industry. Through their representative body (Assuralia), insurers filed a complaint with the European Commission explicitly denouncing the dominant position of mutual benefit societies, ⁷¹ which they claimed contradicted the EU Insurance directives of 1992. ⁷² Two years later, the Commission issued an opinion condemning Belgium. In 2010, such opinion was confirmed by the European Court of Justice. ⁷³

These events opened a window of opportunity for policymakers to reduce their exposure to the dominance of mutual benefit societies in the healthcare system. Yet and even under the conjoint pressures of the European Commission and a unified insurance industry, the privileged position enjoyed by mutual benefit societies proved resilient. As a response to the Commission, the government held negotiation involving insurers, mutual benefit societies, their representatives, and the Ministry of Health headed by Laurette Onkelinx (Socialist Party). ⁷⁴ The government sided with mutual benefit societies for reasons that are hard to disentangle—mostly pointing toward the feedback effects triggered by these actors’ role in the Belgian system. Their strong ties enabled representatives to swiftly establish a common position, presenting a united front. Major federal associations, particularly the Socialist and Christian mutuals, activated longstanding relationships with corresponding parties and unions to secure broader popular support for the status quo—presented as a defense of the Social security system itself. On the government side, deregulating the private insurance sector would have meant exposing the costs of stagnating health expenditures in social insurance and, more fundamentally, officially endorsing the privatization of the system. Consequently, the result of the process was largely in favor of mutual benefit societies. After several back-and-forth with the European Commission, ⁷⁵ they obtained that complementary health insurance become legally binding, and that their de facto monopoly on this segment be recognized as an extension of the statutory coverage and thus, insulated from EU requirements in terms of free and fair competition. ⁷⁶ To comply with European law, a small segment of complementary health insurance was defined and opened to competition with insurance companies—a segment that mostly involves hospital and dental care, which account for only 5% of health expenditures. ⁷⁷ The law of 26 April 2010 confirmed this overall architecture, ⁷⁸ while a Royal Decree in 2011 added to the final text a number of extra (and not negotiated) provisions to the benefit of mutual benefit societies, despite vocal complaints from insurance companies. ⁷⁹

As expected, the development of private insurers’ institutional power in Belgium is thus very consistent with the predictions and findings of a large segment of the literature, and most notably those of Busemeyer and Thelen. In this case, institutional power increased due to their greater and prolonged involvement in policymaking and implementation. The model of shared responsibilities underpinning the public–private partnership ultimately led to greater capacities for private actors in relation to state actors, accompanied by a gradual acceptance by the general public of the role of private providers in this area. However, instead of being a generic feature of institutional business power, the developments observed in the French case suggest that different initial design choices can yield different outcomes.

As for Belgium, private health insurance has come to occupy an ever-greater important place in the French healthcare systems over the course of the 1990s and, more importantly so, in the 2000s. As of 2020, it accounts for €38,3 billions in premiums. ⁸⁰ For a number of benefits (like hearing aids, dental and optical care), its share even largely surpasses that of the statutory healthcare system. However, the fragmentation of the industry induced by a more competitive environment than in the Belgian case will lead to recurrent collective action problems. At the same time, state actors enhance their regulatory capacities, imposing additional requirements on private insurers to qualify for state subsidies and tax exemptions. The increased political salience of the public–private mix in the French healthcare system (sometimes instrumented by policymakers themselves) eventually led to several recent attempts by the state to regain control from private actors.

Greater competition in the French health insurance market began shortly after its creation. During the 1990s, the selective advantages that mutual benefit societies are enjoying were gradually removed (notably, though not only in application of the European rules and regulations mentioned in the Belgian case). As a result, insurance companies dramatically expand ⁸¹ —with their market share increasing by around 61% between the beginning of the 2000s and the end of the 2010s. In the same time, provident institutions (the third category of health insurers found in France) maintain their own market share through expanding in occupational welfare and corporate health insurance plans. Different actors with different interests thus compete in an increasingly disintegrated market.

This led to an overall stagnation, if not a relative decrease, in the power of private insurers in the realm of policymaking, as they repeatedly failed to agree on a common strategic orientation. Consequently, private insurers acquiesced to most of the additional regulatory burdens imposed by policymakers over the past twenty years. They easily accepted to participate in CMU-C (“Couverture Universelle Maladie Complémentaire”), a scheme introduced by the left-wing coalition government in 1999. ⁸² By complementing the universal healthcare coverage plan introduced the same year, this device enables private health insurers to contribute to increased access for various population categories that were previously excluded from the national health insurance system, including precarious workers and migrants. In 2004, an ambitious reform led by Philippe Douste-Blazy (the then ministry of social affairs of the right-wing coalition government) creates an additional device in which health insurers are involved—the Aide à la complémentaire santé, a financial aid to purchase private insurance. ⁸³ The same law created a single representative body for all categories of private health insurers—that remained internally fragmented given the divergent interests and preferences of its members. Overall, the two mechanisms that led to an increase in health insurers’ power in Belgium dynamically interact in the French case to result in a decrease in institutional business power: facing numerous and diverse challengers, firms could not resist policymakers’ demands without risking potential advantages for their competitors in future decisions. Consequently, state control and regulatory capacities improved without strong resistance from the industry.

As anticipated, a model based on a more independent agency led to a clearer separation between public and private actions. When the public–private mix in healthcare gained greater political significance, the resulting configuration resembled a situation where governments rely on their political authority to address market failures, ⁸⁴ in contrast to the blurred landscape observed in Belgium. This is notably revealed in the policy developments that arise from the growing threats, sometimes initiated by public actors themselves, of the state potentially assuming partial or complete control of the sector. This may involve expanding the scope of statutory coverage or integrating private insurers into the Social Security. Over the course of the 2010s, several influential administrative reports directly explored this option, often arguing that it was both feasible and desirable. ⁸⁵ In the same time, senior civil servants (often with a longstanding experience in the social security administration), through op-ed pages in national newspapers ⁸⁶ and articles in specialized journals, ⁸⁷ promoted the same views. This created a favorable ground for policymakers’ more ambitious interventions in the health insurance market. In 2019, the social security finance law delineated a range of healthcare benefits for which the share reimbursed by private health insurance was high (mostly in hearing aids, dental and optical care) and imposed a range of fixed rates so that individuals no longer have to support out-of-pockets expenditures. ⁸⁸ At the same time, President Emmanuel Macron explicitly warned health insurers, urging them not to transfer the cost of these reforms onto their prices. ⁸⁹ At the end of his first term in office, his ministry of health (Olivier Véran) demanded an additional report to the highest advisory body of the national health insurance system, and this to explore new “modes of articulation” with private health insurance—including the possibility of an increase of statutory coverage. ⁹⁰

Although these initiatives were partly slowed down by the Covid-19 pandemic and its aftermaths, they largely triggered a largely publicized debate among experts where the possibility of a state’s takeover of private health insurance gained wider visibility and support. The consequences of the greater politicization of France’s mixed system of health insurance are perceptible—with private health insurers confronting state actors gradually regaining control over their activities. The most recent period has seen an additional extension of the services and benefits subjected to greater state intervention to limit out-of-pocket expenditures, and for which private insurers face tighter pressures over their benefits by the executive instrumenting the threat of a wholesale return of the state in the sector. ⁹¹

Conclusion

A rich literature in political economy suggests that when state actors either invite or allow private actors to play a central role in the provision of collective goods, they “foster asymmetric dependencies of the state on the continued contribution of business actors in ways that, over time, tilt the public–private balance increasingly in favor of business interests” (Busemeyer and Thelen, 2020). This paper has tried to establish two series of arguments that provide slight, although we believe important nuances to this general depiction. First, and after a review of policy developments in widely different areas and varieties of capitalism, we have suggested that institutional business power might be subjected to a more important variation over time than usually assumed, ranging from the expansion of business actors (as predicted by most of the literature) to state actors regaining control of the sector. Secondly, we have argued that the initial policy design choices through which institutional power was acquired could explain such diverging outcomes. The comparative historical analysis of the fluctuating fortunes of private health insurers in Belgium and France has provided an empirical illustration of this twofold argument. Although the respective capacities of state and private actors differed ex-ante in the two countries, we observed a significant trend toward a relative increase in private insurers’ institutional power in the first case and a relative decrease in the second. Both were empirically linked to the distinct and specific consequences of delegation and deregulation, respectively.

Taken together, our findings have we think several implications for current strand of research in political science and political economy. They first and foremost suggest that students of the interactions between state and private actors in public–private settings should pay a greater attention to the institutional configurations within which such power emerges and expands. Our case study has indeed revealed, in that respect, the important implications these initial design choices could have not only on policymakers’ capacity to regain control of a sector (the usual focus of the literature), but also and crucially in terms of their (political) incentives for doing so. These same initial conditions are also likely to shape both the type of private actors involved, but also the structure of competition among them—and thus their ability to form durable and unified coalitions to successfully pressure policymakers in the course of future decisions.

In a similar vein, the different feedback effects that we respectively linked to delegation and deregulation suggest that the extent with which state actors remain involved in a public–private arrangement may have some counter-intuitive implications. The state, through delegation, maintains an active presence alongside private actors instead of governing the market at a certain distance as in the context of deregulation. While the first configuration appears to be more conducive to a potential return of the state in the longer term, our theoretical analysis and the comparison of health insurance in Belgium and France have shown that countering private interests is particularly challenging after delegation. Thus, “agency capture” by “rent-seeking interest groups”, which is well-documented in the literature on corporatism, seems in effect hard to prevent. ⁹² While granting private actors with more leeway, deregulation, by contrast, offers greater possibilities to state actors to take back control from private interests, especially when the later are more fragmented and the former, backed by broader political coalitions or manage to raise the salience of market-based arrangements.