Can Public Healthcare Afford Marketization? Market Principles, Mechanisms, and Effects in Five Health Systems

: Policymakers now have four decades’ experience using marketization to address cost and quality problems in public-sector health services. While much is known about the challenges, it is difficult to draw lessons because there remains no agreed-upon definition of marketization. This article contributes a definition that focuses on the transaction, particularly the effects of funding arrangements on the intensity of competition among providers. Based on prior literature and 106 interviews with practitioners and researchers in five countries, the authors contribute a systematization of 12 concrete market mechanisms enacting three market principles. Furthermore, the authors analyze respondents’ perceptions of healthcare marketization’s effects on costs and quality. While marketization is a multi-faceted, sometimes ambiguous phenomenon requiring further research before definite conclusions can be reached, most statements from our respondents about cost and quality effects were negative.

"T he market" has become a multi-purpose toolbox to address quality and cost problems in public services. In healthcare, policymakers have allowed commercial organizations to provide services while making the public sector more entrepreneurial. However, the consequences of marketization are uncertain, with much conflicting evidence relating to its effects on service cost and quality.
Utilizing 106 qualitative interviews, we conceptualize and empirically evaluate marketization and its effects in health systems. We include the state-dominated National Health Service (NHS) in England, the more privatized French health system, the hybrid systems of Slovenia, Finland, and Greece, and interviews at the European Union level. We identify 12 market mechanisms introduced to promote competition in health systems, grouped under three market principles: • Openness: shifting costs from public to private sources, financializing infrastructure projects, and loosening rules excluding non-traditional providers.
• Competition between public and private sectors: changes to payment systems or purchasing, allowing existing providers to fail, expanding frameworks for performance management and evaluation, increasing patient choice, and competitive tendering. • Management autonomy in the public sector: public hospital autonomization, internal markets, and regulatory decentralization.
Our first contribution is to provide a conceptual starting point for more systematic debate over the desirability of markets in public services, informed by diverse experience in health systems. While prior research has evaluated marketization through examining specific market mechanisms, our conceptualization covers a more comprehensive range of market mechanisms as they appear in five systems.
Many of our respondents' statements about cost and quality were negative. They reported many examples of market mechanisms costing more money than alternatives, and relatively few cost or quality improvements. Numerous other statements revealed unclear or ambiguous consequences. Our second contribution therefore shows, according to practitioners, the problematic effects of the marketization of health systems.
The following section reviews past efforts to assess marketization in health systems and introduces our classification of market mechanisms. Next, we discuss the implementation of market mechanisms and possible explanations for their limitations. Then, we discuss our methods and present findings on perceived cost and quality effects. Finally, we discuss the research's policy and conceptual implications for marketized public services.

Market Principles and Mechanisms
Scholarship on health systems defines marketization in varying ways. Some scholars emphasize privatization, where governments retreat from service provision (Gilbert 2002;Jensen 2011).
Others emphasize competition, as in studies of changing incentive systems in funding, consumption, and allocation (Gingrich 2011;Jacobs 1998). Marketization sometimes denotes change in healthcare organizations, which may be increasingly commercialized (Rylko-Bauer and Farmer 2002) or pushed toward particular models of accountability (Neby 2016) and efficiency (Helderman et al. 2005). Marketization is sometimes located in the transaction, as in Reich's (2014, 1) definition of "market" as "a principle of exchange for profit or gain" and Le Grand's (2003) notion of government as purchasing agent. These definitions vary depending on whether they relate markets to private ownership, competitive incentives, or organizational change; and on the particular market mechanisms under examination. Such a variety of definitions and implicit conceptualizations mean it is difficult to rigorously overview, assess, and compare marketization as a large-scale policy instrument.
Policymakers have employed marketization to solve perceived cost and quality issues in healthcare. Economic crises have been a source of cost constraints. After the 1991 crisis in Finland, for example, policymakers reduced tax subsidies and increased user charges (Häkkinen and Lehto 2005); introduced Diagnosis-Related Group (DRG) funding; permitted municipalities to competitively tender services (Mikkola 2003); and introduced Total Quality Management in public healthcare organizations (Vartiainen 2008). In France, shifts to DRG funding and New Public Management techniques also aimed to reduce waste and improve cost-effectiveness (Minvielle 2006;Umney and Coderre-LaPalme 2017).
Marketization has also been employed to resolve quality concerns, such as waiting times or lack of patient choice, especially in state-dominated systems like the UK (Jacobs 1998;Powell 2015). International institutions have promoted market mechanisms to improve accessibility and efficiency. The European Union's 2011 Directive on Patients' Rights to Crossborder Healthcare, for example, established processes for European residents to receive care in other countries (Greer and Jarman 2012). The European Union has also reduced the time allowed for paying providers and promoted public-private-partnerships to access private-sector capital and expertise (EXPH 2014).
Following Le Grand (2003) and Reich (2014), we view marketization as a property of the transaction between purchaser and provider. We define it as the introduction or intensification of price-or cost-based competition among service providers. We distinguish between market principles (abstract justifications for intensified competition) and market mechanisms (concrete changes to rules and procedures aimed at increasing competition), setting out a more comprehensive conceptualization, which can underpin evaluations of the effects of marketization. We reviewed literature on healthcare marketization to understand the diverse range of market mechanisms used, producing a conceptualization that we tested and expanded through our data gathering (discussed in the methods section below). We grouped mechanisms according to three underlying market principles, each of which seeks to increase competition in a given system. These abstract principles are enacted through twelve specific mechanisms.
The first market principle is openness: healthcare provision should be open to new actors to finance or offer services, even if public and private sectors remain intertwined (Kettl 2015). Quasi-market theorists advocate a "mixed market" of public, for-profit, and non-profit providers (Bartlett and Le Grand 1993). The three mechanisms associated with this principle are: • Cost shifting (known in the US more commonly as 'cost-sharing'): Policymakers can shift costs from public budgets by decreasing tax funding and by increasing out-of-pocket expenditures, such as co-payments or private insurance. As public spending is displaced, business opportunities emerge for insurers to provide insurance packages to employers and individuals, and to offer coverage for private-sector services, as in Finland, where private occupational health and surgical services expanded following increased out-of-pocket spending and employer-based health insurance. • Financialization of infrastructure services: Large-scale infrastructure projects like building hospitals are risky and require borrowing. Through Public Private Partnerships (PPPs) or Public Finance Initiatives (PFIs), governments shift risk onto the private sector and reorganize debt, creating business opportunities in finance, consulting, construction, maintenance, and ancillary services. England has used PFIs and PPPs extensively since the mid-1990s, and France and Greece have experimented with similar arrangements. • Inclusion of non-traditional actors in health provision: Large-scale privatization of hospitals-known as 'material privatization'occurred in several German states and municipalities in the early 1990s. More commonly, policymakers loosen restrictions on public funders purchasing services from private-sector providers, causing more gradual 'functional privatization' (or contracting out). Shifting activity to outpatient settings in (traditionally private) community rather than (traditionally public) in-patient settings is one way the English NHS has increased private-sector involvement.
Together, these mechanisms seek to increase competition by creating a market for insurers to provide more products, a market for financiers to vie for infrastructure contracts, and the ability of nontraditional providers to vie for services.
A second market principle is competition between public and private sectors. According to quasi-markets theorists, competition should make public-sector providers more efficient and responsive (Bartlett and Le Grand 1993). Six mechanisms promote this: • Fixed-price reimbursement rates: Introducing fixed-price reimbursement rates for all providers of given procedures is widespread internationally and usually modeled on the Diagnosis-Related Groups (DRGs) introduced in 1983 in the USA. DRGs fix prices for an entire episode of care, rather than for discrete services and rather than allowing them to fluctuate according to supply and demand, increasing price transparency and incentivizing cost-reduction. DRGs vary in their share of overall funding (much smaller in the UK than France and Finland) and their method of price determination (in Finland, rates are renegotiated annually to reflect hospitals' operating costs, whereas other countries use them to contain costs). • Centralized purchasing: Public-sector health providers can put downward pressure on the prices of services, technology or pharmaceuticals by combining purchasing functions across many organizations. This allows public-sector providers to leverage greater bargaining power, intensifying competitive pressures on the private sector. France, for example, has created associations of hospitals to streamline purchasing, extracting price concessions from suppliers. Greece has used centralized purchasing to reduce drug prices. • Failure regimes: These enable public entities with severe deficits to go bankrupt instead of receiving bailouts. These entities can be broken up and privatized or merged into other public-sector organizations. • Performance management of public-sector organizations: These stimulate competition by benchmarking providers against one another and encouraging reputational or financial rewards for high performers. The best-known examples are in England's NHS, which has imposed extensive performance targets, such as strict targets for waiting times for surgeries and emergency room services or the monthly reporting of over 70 process and quality targets for intensive care services (Bach and Kessler 2012 Together, these mechanisms create frameworks for more direct competition between public and private providers, for patients and for resources, and equalizing the consequences of competitive "failure." The third market principle is public-sector management autonomy. For competition to drive supposed improvements, incumbent public-sector providers must reorient themselves to customers and competitors, becoming more market-facing (Walker et al. 2011). This principle is associated with three mechanisms.
• Internal markets: These create an organizational division between public-sector purchasers and providers, with armslength transactions to specify, tailor, and improve health services. The less-marketized alternative in NHS systems was to divide the public budget across regions without (or only informally) specifying the conditions of service provision. Quasi-markets theorists have examined the internal market in Britain's NHS since the early 1990s (Le Grand 2003). • Hospital autonomization: In hospitals converted into government-owned corporations, public managers gain flexibility over spending and restructuring. When public hospitals in England gain Foundation Trust status, for example, they can retain and reinvest surpluses and compete for high-paying patients from outside their area. • Decentralization: Decentralizing regulation aims to increase public-sector bureaucracies' responsiveness to local conditions, helping them to tailor service provision to local needs rather than running potentially unneeded services. This can enhance the public sector's competitiveness (Simonet 2008). France, for example, 'deconcentrated' healthcare planning in 1996 by creating relatively autonomous Regional Health Agencies.
Together, these mechanisms seek to increase the public sector's competitiveness by instituting new, contract-based accountability mechanisms, increasing health sector managers' and administrators' responsiveness to local competitive conditions.
Healthcare marketization is therefore a multi-faceted process of institutional change. Many of its aspects are contested, and their differentiated effects need examining.

Examining the Effects of Market Mechanisms
Health systems appear to have become extremely marketized. Table 1 shows how different market mechanisms have spread into the diverse health systems we examine. Ten out of 12 mechanisms have spread into either four or five countries. Moreover, patient choice and internal markets are not shown for France because these were already longstanding features of its insurance-funded system before the beginning of our study period. One indicator of the successful implementation of market mechanisms is increasing private-sector activity. In health systems, however, the public-private funding mix has been surprisingly stable, with most countries witnessing stability or expansion of direct funding by government (see Figure 1). Where public provision has decreased, as in France and Greece, however, this often reflects public-sector retrenchment under post-2008 austerity budgets rather than private-sector expansion, and elsewhere declines in publicsector hospital beds accompany increases in government spending.
One explanation for limited change concerns the uneven spread of market mechanisms. The three mechanisms spreading to the fewest countries are those most clearly conducive to privatization: patient choice, decreased tax funding, and financialization of infrastructure. Even where mechanisms are introduced, their implementation can be difficult. For example, although all five countries have opened services to non-traditional actors, none has taken the German approach of privatizing whole hospitals. Gradually contracting out particular services is less politically contentious. In Britain, for instance, campaigners have successfully prevented various tendering exercises from leading to privatization, and blocked the restructuring of a financially troubled hospital in Lewisham, South London (Krachler and Greer 2015). Hence, the issue is not only the spread of market mechanisms, but also their implementation.
Another explanation highlights the difficult conditions under which those tasked with implementing marketization often operate. Public administration scholars show the difficulties of balancing competing demands for more efficiency, equity, or democratic voice. Public services are often contracted out at the behest of elected officials under market conditions that public managers know are suboptimal: without large numbers of competitors, managers face "limited information, uncertainty about the future, and the prospect that people or organizations will behave opportunistically in their interactions" (Brown, Potoski, and Van Slyke 2006, 325). Many municipalities facing this balancing act eventually bring services back in-house or mix contracted-out and in-house provision for the same service (Warner and Hefetz 2008). Girth et al. (2012) show that most public-service markets are uncompetitive, especially in rural areas, requiring public-sector managers to expend extra effort making contractors accountable. Bryson, Crosby, and Stone (2015) argue the pursuit of efficiency and effectiveness may blend out other concerns such as citizen engagement, collaboration, and democratic values, and corrode the distinctive values and practices of the nonprofit sector (Eikenberry and Kluver 2004). New Public Management techniques may preclude potentially more effective strategies such as labor-management dialogue (Lindsay et al. 2018;Riccucci, 2011) and inflame tensions with relevant staff unions (Budd 2014;Givan 2016;Greer 2008;Greer, Schulten, and Böhlke 2013;Krachler, Auffenberg, and Wolf 2020). Internationally, public opinion also overwhelmingly supports government provision and funding of healthcare, shown in Figure 2.
Some studies have shown generally negative cost and quality effects of healthcare marketization, which may help contextualize uneven implementation. Rice and Unruh (2016) discussed the poorer cost and quality performance of the more marketized US healthcare system compared to other OECD countries, sometimes attributed to higher administration costs (Himmelstein et al. 2014), overall higher prices (Oberlander and White 2009), and the US's unique disconnect between higher governmental spending and lower avoidable mortality (Heijink, Koolman, and Westert 2013). Nemec and Kolisnichenko (2006) found the introduction of health insurance in Central and Eastern European countries caused higher costs for patients and no increases in life expectancy, other than Slovenia, the least marketized system in the region. More marketized systems might have poor incentives including creaming (choosing to treat low-cost or lucrative patients above others based on expected revenue rather than need), upcoding (providers charging insurance for more lucrative procedures), and oversupply of services (such as unnecessary tests). Managing these is itself costly due to transaction costs on the public-sector side and private-sector administrative bloat. Some policymakers we interviewed similarly argued that market mechanisms were too costly to implement.
Research on the effects of specific market mechanisms is also ambiguous. In evaluating British internal market reforms, West (1997) and Le Grand, Mays, and Dixon (1998) found few efficiency improvements. While studies find significant cost and quality benefits from shifting care into outpatient and primary care sectors (Klein, Laugesen, and Liu 2013;Xing, Goehring, and Mancuso 2015), other research has found small patient satisfaction improvements (Stokes et al. 2015) or ambiguous effects (Brown et al. 2012). DRGs have been associated with significant reductions in average lengths of stays (Albreht et al. 2009; Moreno-Serra and Wagstaff 2010) as well as with expenditure increases due to increased activity (O'Reilly et al. 2012), no significant efficiency gains in Germany (Herwartz and Strumann 2014) and Finland (Mikkola 2003), and an exacerbation of Greece's public health crisis (Kondilis et al. 2013). Regarding competition for patients, Gaynor, Moreno-Serra, and Propper (2013) associated increased competition with lower mortality rates (without cost increases) in the UK while in the US, large teaching hospitals providing specialized services used their market power to extract higher prices (White, Reschovsky, and Bond 2014). Finally, in the UK, lean management techniques have had ambiguous effects on waiting time (Radnor, Holweg, and Waring 2012) and public-private-partnerships were often associated with cost increases (Roehrich, Lewis, and George 2014).
This review of the literature reveals much ambiguity in evaluating the effects of marketization. This ambiguity may reflect diverse definitions and a focus on different individual mechanisms. Below, we present a more comprehensive conceptualization of three market principles and twelve market mechanisms, which provides a means for examining perceptions of cost and quality outcomes in comparative perspective.

Methods
To develop a comprehensive conceptualization of the spread and effects of market mechanisms, we chose an inductive, iterative case study approach (Corbin and Strauss 1998). We employed a 'diverse case' selection strategy, seeking to capture a phenomenon's maximum range of variation and increase the representativeness of case findings (Seawright and Gerring 2008). Moreover, to theorize the elements and effects of healthcare marketization, we aimed to identify the commonalities (rather than divergences) across this diversity.
Hence, we selected country cases capturing variation in the most common types of health systems (NHS systems, social health insurance systems, and hybrid systems). Past literature has compared markets in health systems referring to the public-private mix in funding and provision (ECFIN 2016). England is a paradigmatic case of a weak-market, state-dominated NHS but has gone further than other British regions in implementing market mechanisms (Greer, 2004). France is a strong-market system, with more private provision, funded by social insurance more than taxes (ECFIN 2016). We sought further diversity by including the hybrid systems of Greece, Finland, and Slovenia that combine governmentdominated systems with social insurance (Böhm et al. 2013). A further benefit of these countries is to cover major European regions (Southern European, Nordic, Western European, and Eastern European countries). Table 2 summarizes these public-private mixes. Moreover, we included European-level respondents to understand key policymakers' perspectives on healthcare marketization, and the spread of mechanisms across Europe.
Before data gathering began, the first author conducted a review of literature on marketization and healthcare in case countries and the European Union across the disciplines of health policy, sociology, political science, and health economics to identify as wide as possible a range of market mechanisms. This review forms the basis for the set of principles and mechanisms outlined in the preceding sections. We covered marketization reforms from the early 1980s, generally regarded as the beginning of the reorientation of health policy toward marketization (Gingrich 2011). This review served as the basis for a general interview questionnaire (and later, a first coding framework), which informed semi-structured interviews with hospital managers; policymakers (mostly staff in ministries, regulatory agencies, municipalities, and the European Commission); industry bodies (including employer associations and industry associations), health campaigners and trade unionists, and frontline professionals (mostly physicians) between 2012 and 2016 ( Table 3).
The aims were to test the existing framework regarding the spread of market mechanisms and gather expert perspectives on their effects. In reviewing the literature, we had found little systematic analysis of effects, with most studies focusing on describing particular mechanisms. We used theoretical concerns to select respondents (Corbin and Strauss 1998), seeking those with at least five years' experience in a senior position of regional or national importance, and to cover a diversity of normative perspectives toward healthcare marketization: market proponents (such as privatesector managers), market opponents (such as public-sector trade unionists and managers, and public healthcare campaigners), and actors with normative views that could not be anticipated a priori (such as commissioners, policymakers, physicians and researchers). The diversity of perspectives also has the benefit of allowing us to triangulate different perspectives and cross-check statements against one another, increasing the internal validity of our findings (Patton 1999).
A team of researchers conducted and transcribed most interviews in respondents' native languages and summarized them in English. In Slovenia, Finland, and Greece most interviews were carried out in the local language by a native speaker, and a small number were conducted in English when the principal investigator visited (most of these fluent English speakers were senior physicians, policymakers and administrators). In Brussels, where the working languages are French, German, and English, interviews were in English or German; interviews in France were all in French. Language barriers did not exclude participants in the research sites.
The interviews were 'semi-structured', with a common set of themes but also allowing country researchers and respondents to guide the interview as needed. This allowed for local contextualization; the tailoring of questions to respondents' experience and current position; and, the co-design of our data through respondents, permitting them to bring up unanticipated topics they deem important (Silverman 2001). One drawback of this approach is that respondents' co-design means that certain questions may not be fully answered. Hence the volume of data on the effects of individual market mechanisms is uneven (see Table 5) and the largest volume of responses comes from England and Finland (see Table 4). Within each category of effect, however, the responses have a substantial amount of counter-bias (a respondent's perception of an effect opposed to our a priori assumptions based on position; for instance where a private-sector manager identifies negative effects of marketization, or where a public-sector trade unionist identifies benefits), indicating internal validity. This was particularly so for negative effects on cost-containment where 50 percent of responses run counter to respondents' bias (see Table 4). Moreover, a significant percentage of statements also comes from neutral actors, especially for negative quality and unclear effects. The variation in the volume of statements from each country does not reflect overall numbers of interviews. Rather, it highlights the more contentious and wide-ranging packages of market reforms introduced in these countries which prompted more numerous distinct comments on a wider range of separate mechanisms. In French interviews, for instance, much respondent-led discussion involved in-depth focus on particular mechanisms upon which policy had relied particularly heavily (such as DRGs), resulting in rich data but with fewer distinct statements.
Our initial priority was identifying the main forms of marketization in each country and the policies underpinning them, checking them against our literature review. Our interview questions focused on (1) whether there was increased competition between providers or privatization; (2) how increased competition was related to particular health reforms; (3) whether changes affected health system functioning, patients, or workers. As research progressed, we identified two new market mechanisms (centralized purchasing and failure regime) and explored effects on cost and quality in each country. We then analyzed the commonalities of our findings across the countries. In 2017, to ensure the validity of our research, we filled in knowledge gaps systematically by asking country experts about categories for which we lacked information (especially mechanisms). We subsequently also reviewed more studies in prominent US-based journals to ensure sufficient breadth of our reviewed literature.
Our data comprises 106 semi-structured interviews which were supplemented with (1) written sources such as research articles and studies (starting from the European Observatory on Health Systems and Policies), news articles, and policy papers (many identified with our respondents' help, including from prominent think tanks like the King's Fund, public auditing institutions like the French Cour des Comptes, or interest representation groups like private hospital federations or activist organizations); and (2) publicly available statistics. The 'triangulation' of diverse sources is important where respondents' differing perspectives and the diversity of market practices lead to contrasting statements about the same phenomenon.
Reconciling and cross-checking the validity of these differences was part of the iterative process of conducting interviews and analyzing memos, documents, and transcripts using MaxQDA (Patton 1999).
To explore effects, we coded respondents' individual, distinct (i.e., not double-counted) statements related to costs and quality for a specific mechanism as either "positive," "negative," or "unclear", and collated these codes (Table 4). A positive effect on cost meant a cost reduction or containment of cost growth; a negative effect on cost meant cost increases, at national or regional health systems levels. For example, we coded this passage as a positive cost effect: "[through DRGs] we've been seeing some surgery is more expensive than in the other districts, then we've been able to negotiate about the price and what can be done to reduce the price" (Finnish Municipal Purchasing Manager). A positive effect on quality meant quality improvements, and a negative effect meant a decrease in service quality. For example, we coded the following passage as a negative quality effect of DRGs, which highlights the problem of incentivizing public hospitals to shorten patient stays: "Not all patients are equal. If someone is alone, who will return to their home with nobody to welcome them? That poses social problems, and the time of stay will extend because of these social difficulties" (French Physician and Researcher). "Unclear" meant respondents were unsure or said no information was available. For example: "Interviewer: 'Have you contracted out services such as cleaning?'. Slovene Public-Sector Hospital Manager: 'No, not at this time.
Since, I am not sure, I was not sure and I still am not sure that outsourcing services is of any benefit'."

Perceived Cost and Quality Effects of Marketization
Next, we apply our framework to respondents' views of the cost and quality effects of market mechanisms in healthcare (Table 5). The most frequently discussed mechanisms were competitive tendering (114 statements), inclusion of non-traditional providers (53), and fixed-price reimbursement (43). Respondents reported almost twice as many negative effects than positive ones (58.5 percent versus 31.0 percent of responses). Respondents were more likely to see positive outcomes for cost containment (75 statements) than quality (43), but many more statements linked these market mechanisms with higher costs (123) and reduced quality (100).

Marketization in General
We used the code "marketization in general" for statements relating to policy packages comprising multiple market mechanisms. The most talked-about examples were England's 2012 Health and Social Care Act, Finland's proposed (and later shelved) "SOTE reform", and a series of reforms in France starting in 1995. Respondents made five positive quality-related statements about marketization in general; respondents in France and Finland suggested that reduced restrictions to providers entering the market may improve health service access. Eleven concerned cost containment, as when Finnish respondents argued that outsourcing purchasing to the private sector had reduced administrative costs.
Negative comments over marketization in general were more numerous, with 12 concerning quality and 38 cost. Respondents, especially in England and Finland, argued that the private sector provided less comprehensive care, prioritized certain tasks, or cherrypicked low-cost or lucrative patients, had worse working conditions, and had less interest in serving rural areas. Most negative comments relating to cost effects came from England, where many participants regarded the public sector as reinvesting surpluses in care while private-sector provision removed resources as profits. Many English and some Finnish respondents noted high administration costs of restructuring healthcare systems through marketization.

Openness
The openness principle reveals a mixed picture. About 53 of 90 statements concerned the mechanism of including non-traditional providers, where positive comments (28) outnumbered negatives (22). Most positive quality-related statements came from England and included private innovations like giving primary care physicians more telephone access to specialists and introducing care coordination roles. Positive cost-related statements were also concentrated in England, and concerned increasing private-sector investment in information technology, and reducing infrastructure and maintenance costs through increased home visits by community providers. Some cost savings were at the expense of workers or patients, through avoiding collective bargaining and increased out-of-pocket payments.
Other "openness" mechanisms were discussed more negatively.
In insurance-funded systems (France, Slovenia and Finland),   3  2  6  1  0  1  13  France  2  2  6  3  0  3  16  Greece  0  7  1  3  7  0  18  Slovenia  4  3  1  1  4  0  13  England  7  13  2  13  0  0  35  Total  23  27  16  22  11  7  106 cost-shifting onto service users and private insurers was viewed negatively (nine negative quality-related statements versus zero positive; three negative cost-related statements versus one positive). Respondents saw these shifts as reducing access because some patients could no longer afford care. Aside from the consequences for individuals, hospitals would have to treat patients' conditions that they could have prevented earlier more cheaply.
Respondents in England viewed financialization as problematic, chiefly the long-term costs of PFI/PPP construction projects. There was one positive quality-related statement versus two negative and one positive cost-related statement versus 15 negative. Respondents stated that financialized infrastructure projects entailed higher longterm costs than public-sector-financed construction due to high return rates for private equity firms. Such costs could exacerbate hospitals' debt burdens, especially combined with DRGs, which paid for procedures without accounting for higher borrowing costs. EU-level respondents confirmed the UK had the most financialized infrastructure projects, while respondents in France and Greece viewed similar projects in their countries as failed experiments.

Competition between Public and Private Sectors
The most commonly discussed mechanisms for increasing publicprivate competition-for positive and negative comments-were fixed-price reimbursement (DRG) systems and competitive tendering. These mechanisms comprised 157 of 188 statements about this principle. A widespread view, expressed in 55 statements, was that this principle increased costs, especially in administration.
DRGs were associated with two positive quality-related statements (versus 21 negative ones), and six positive cost-related statements (versus 10 negative ones). While some respondents argued DRG systems could reduce costs by reducing stay lengths, others cited the same effect as a negative for quality. Most negative comments came from France, reflecting how the French system moved rapidly toward DRG-based payments under the Sarkozy Presidency.
Respondents argued DRG funding ignored the greater complexity of public hospital workloads, for instance, where they treated people with more vulnerable living arrangements or high-acuity conditions. Others argued it accelerated frontline hospital work and pushed hospitals to prioritize lucrative services, undermining quality.
The negative cost effects of DRGs highlighted perverse incentives. Respondents from France, England, Slovenia, and Finland reported increased administration costs to ensure providers code all services in the most profitable category; elevated levels of provision for profitable codes (such as C-sections); and potentially fraudulent behaviors (such as coding services that were not rendered). Some respondents doubted the reliability of cost data used to calculate prices.
Competitive tendering was associated with positive quality effects in 17 statements. Some respondents (in England and Slovenia) argued it could reduce pressure on public waiting lists or improve management practices (England and Finland). Some argued privatesector contractors could introduce new information technology, improving service delivery. In 22 statements, respondents in England stated competitive tendering improved cost containment, mostly because new market entrants were seen as more efficient, for example, by using automation to reduce reliance on professional pathologists; or, by maximizing the running time of radiological machines.
Statements about competitive tendering were more often negative regarding both quality (28 in England, Finland and Slovenia) and cost (36 statements, mostly in England). Common criticisms concerned the specialization of private-sector organizations in lowacuity, low-complexity services, such as elective surgeries, compared to complex, high-acuity services like emergency, gynecology, or intensive care. Respondents also felt criteria for contract awards prioritized price over quality, and that competitive tendering could split services apart, undermining coordination. Respondents also viewed competitive tendering as lengthy, costly, and risky. Providers faced the risk that funders might withdraw tenders, have unclear expectations, or make decisions slowly. Moreover, competitive tendering created expenses where providers bid below their operating costs and abandoned contracts early. In small countries, like Slovenia and Finland, with few contractors able to submit bids, the scope for reducing prices through competition is limited. Some respondents reported instances of insourcing ancillary services after outsourcing raised unexpected costs.
For other market mechanisms relating to this principle, the mix of negative and positive statements was more even. For instance, Finnish respondents argued that measures to increase patient choice, such as vouchers, had expanded service access, while expressing concern that vouchers could lead service users to select weaker providers. Centralized purchasing was on one occasion cited as achieving economies of scale in England. Failure regimes in England and Finland were cited three times, producing economies of scale (with potential positive cost and quality effects) but also neglecting local provision and needs.

Public-Sector Management Autonomy
Concerning public-sector management autonomy, 21 of 22 statements concerned either hospital autonomization or decentralization of regulation, with the third mechanism (internal markets) being cited rarely. For these two mechanisms, negative statements predominated. All statements on hospital autonomization were from England. None were positive. However, five negative statements related it to pressures to achieve Foundation Trust status and thus, to a recent scandal over understaffing, poor care, and patient deaths at Mid-Staffordshire. For decentralization, there were six negative quality-related statements and one positive. Respondents in England and Finland argued new regional institutions had increased inequalities in care, fragmented services and produced a less holistic view of care.

Discussion and Conclusion
After four decades of healthcare marketization experiments and debates on the desirability of markets in healthcare, the literature has no consensus about the effects. In public administration and health policy, there remain both advocates and critics. Our findings position us on the critical side of this debate: our research suggests that policymakers and other stakeholders often regard the application of market principles to healthcare with deep skepticism. Although our research has limitations, the conceptualization in this article can serve as the basis for future research on the effects of healthcare marketization.
In the article, we have pursued two objectives. First, we have aimed to inject greater conceptual clarity into our understanding of the varied policy tools involved in marketizing health systems. Previously, literature on market mechanisms in healthcare systems has been limited by a lack of conceptual comprehensiveness and precision. Studies often focus on different aspects of marketization or interpret the term in differing ways, reducing the scope for more ambitious comparative evaluation of its effects. By synthesizing a set of three main market principles and 12 market mechanisms, we provide a framework to facilitate valid and rigorous comparative study in future. Even where our empirical data has limitations requiring further research, this conceptual work provides a valuable contribution.
Second, we have empirically evaluated these principles and mechanisms using testimony from 106 key informant interviews in five countries. Our data reveal some perceived benefits of marketization and some cases where its effects remain ambiguous, but overall we find a generally negative view of its effects on both quality and cost. Our empirical contribution is therefore to provide a macro-level view that balances positive and negative claims gathered from expert interviews. The framework has also enabled us to delve into the varied effects of different specific mechanisms in a detailed way. We highlight the financialization of hospital governance, DRG systems, and competitive tendering as the most commonly discussed problematic mechanisms for quality and cost. By contrast, the principle of bringing nontraditional providers into healthcare systems can have benefits in some circumstances. However, the mechanisms used to pursue this aim often raise significant cost and quality problems of their own. These findings have clear policy implications. Our assessment of healthcare marketization suggests that market principles in health systems are causing important problems and that policymakers should turn their attention away from marketization as a response to problems of cost and quality. A detailed and concrete alternative prescription is, however, beyond the scope of this article.
In light of our respondents' positive assessments as well as a substantial amount of unclear statements, however, our research also highlights new directions requiring further research into healthcare marketization. Our study is exploratory and its several limitations mean caution remains warranted. The main aim of our qualitative methodology was to provide a conceptualization of what healthcare marketization is and how it functions, including its mechanisms and effects. While this is suitable for theory generation and our exploration of questions of cost and quality effects, it limits the certainty of our empirical assessment. Future research could therefore use our conceptualization as a basis for quantitative research. This may include the generation of survey data to run regressions, or analyzing administrative data to employ econometric methods (such as combining a difference-in-difference analysis with propensity score matching; see for example Stokes et al. 2016 who tested case management interventions in this way). This may allow analysis of the specific conditions under which certain mechanisms generate certain effects, as well as a more nuanced examination of effects which can check for interactions and trade-offs between cost and quality. In our research, our respondents' discussion of effects separated cost and quality concerns and did not enable systematic identification of interaction effects.
Methodologically, these quantitative approaches could also mitigate the challenges inherent in our emphasis on experts' perceptions of effects, which may be more vulnerable to individuals' own perceptions or biases. However, we also stress that the validity of our findings is bolstered by the substantial amounts of counterbias statements we found. Quantitative tools could also generate information evenly across respondents, whereas our data features some unevenness, with a larger amount of effects data coming from England and Finland. While this unevenness is a limitation, it is partly explicable by the relatively greater contestation of market mechanisms in these countries.
Moreover, centralized purchasing, internal markets, and performance management of public-sector organizations deserve more attention. We found potentially important consequences associated with these mechanisms, but our respondent-guided approach meant other mechanisms received greater focus. Finally, our research's focus was cross-national to comprehensively capture the characteristics and elements of healthcare marketization itself. Future research could apply our conceptualization and test our findings on effects at different analytical levels, such as the organizational level or for a specific country, region or locality, to give a more granular account of variation in the effects of market mechanisms.