Why Research Should Pay Attention to Effects of Marketization of Addiction Treatment Systems.

OBJECTIVE: Researchers generally assume that addiction treatment systems can be viewed as entities and planned with the citizens' best interests in mind. We argue that another steering principle, the market logic, has permeated many Western World treatment systems but is neglected in research. We demonstrate how it may affect system-level planning, service provision, and the service users.
METHOD: We draw on an ongoing Swedish study, with some Nordic references, using several data sources: (1) public statistics on treatment expenditures and purchases; (2) interviews with service users (n = 36) and their service providers (n = 23) on different market features; (3) an observation of a large public procurement process concluding framework agreements based on competitive tendering; (4) interviews with officials involved with steering of the system and procurement (n = 16); (5) a workshop on procurement in the Nordic countries (n = 11 participants); and (6) 77 interviews with professionals, managers, and elected representatives.
RESULTS: We outline seven propositions that call for further research attention: public procurement, as regulated in the European Union, is not suitable for addiction treatment; marketization challenges democracy, equity, needs assessment, and treatment planning; marketization causes new accountability problems and idle monitoring; marketization causes fragmentation and obstructs coordination and continuity of care; marketization causes unification of services and favors big bureaucratically sophisticated providers; treatment professionals' values are downplayed when a mistrust-based market logic replaces a trust- and needs-based logic; and marketization marginalizes treatment professionals and service users by limiting discretion.
CONCLUSIONS: Findings point toward the importance of acknowledging and mitigating market principles in treatment systems to safeguard needs assessments and planning that serve the interests of the service users and the public.

Addiction treatment systems research prioritizes treatment needs/demands estimates, holistic planning, and implementation of coordinated systems (e.g., Ritter et al., 2014). Typically, these approaches perceive the system as an entity governed by policymakers with the citizens’ best interests in mind, striving for a treatment response that matches needs, legislation, and established principles for equity and professional conduct (Klingemann & Storbjörk, 2016).

We aim to show how an alternative steering principle, the market logic as part of and pushed ahead by New Public Management (NPM; Hood, 1991), may affect the planning, service provision, and service users of addiction treatment systems. NPM has permeated service systems of many Western countries (Pollitt & Bouckaert, 2011). “You cannot see, touch, smell or hear the NPM. It is a rhetorical and conceptual construction . . . open to re-interpretation” (Pollitt, 2007, p. 110). NPM argues that business concepts, techniques, and values can improve the public sector’s cost efficiency, quality, and accountability. NPM contains such practices as introducing competition in the public sector, using contracts as the coordinating device, disaggregation into decentralized corporatized units, purchaser–production splits, management by results, performance measurement, and treating service users as customers (Hood, 1991, 1995; Pollitt & Bouckaert, 2011). NPM may thereby alter the systemic field of norms and justifications for action and relationships (Björk, 2016; Raynard, 2016; Reay & Hinings, 2009; Scott, 2008). It accommodates new actors and introduces specific incentives: competition, nondiscrimination of service producers, and profiteering. Competition as the fundamental idea shapes all actors involved (Meagher & Szebehely, 2013).

We base our argument on the Swedish and Nordic addiction treatment systems, which are highly decentralized (except for specialized treatment in Norway) to local political structures, with independent tax funding and responsibility to provide treatment. They have fairly similar social and health care legislation and professional (social work–medicine) anchorage (Edman & Stenius, 2007). Purchaser–provider splits within the public sector and outsourcing production to nonpublic providers were introduced in the late 1980s. Examining national and local market variations and procurement practices gives arch data for analyses.

Marketization has been studied in a Nordic context with regard to eldercare (Meagher & Szebehely, 2013) and youth care (Forkby & Höjer, 2008). Meagher and colleagues (2016, p. 14) argue that Swedish youth care, driven by NPM, unintentionally transformed from a “regionally coordinated, public social service system . . . into a thin, but highly profitable, national spot market in which large corporations have a growing presence.” The incentives and opportunities for all actors changed. Compared with youth and eldercare, addiction treatment is a small sector in terms of financial turnover and political significance. Importantly, the service users have a weak position. However, similar NPM trends are permeating Swedish addiction treatment (Bergmark, 2010; Bergmark & Oscarsson, 1994; Edman, 2016; Oscarsson, 2000; Stenius, 1996, 1999). The trends are visible but weaker in the other Nordic countries (Bjerge, 2012; Nesvåg & Lie, 2010). Still, few studies empirically scrutinize possible implications of marketization (primarily as competitive public procurement, privatization of production, and managerialism) for needs assessment and fulfillment, service supply, and coordination and equity in addiction treatment systems.

Guided by our empirical data sources and the scarce evaluation literature (including adjacent service areas), we acknowledge advantages with market models and procurement practices (Ritter et al., 2014). The pre-NPM systems certainly had their problems, including paternalism and nonprofessionalism (Edman & Stenius, 2007). Fairness for providers, greater transparency, a growing interest in treatment quality, cost awareness, and demands for estimations of required treatment capacity can be noted as improvements. Some studies support improved process measures, such as reduced waiting times or increased capacity by pay-for-performance models (Hull & Ritter, 2014; McLellan et al., 2008), but there is an overall lack of evidence for effects on outcomes (Hull & Ritter, 2014; Humphreys & McLellan, 2011; Jones et al., 2018; Pedersen et al., 2011; Pollitt & Sorin, 2011). Potential disadvantages for addiction treatment systems are highlighted in our attempt to formulate propositions and identify knowledge gaps calling for thorough research.

Method

Our propositions draw upon the study “Benefits, tensions and inconsistencies in the health and welfare system: The case of New Public Management in Swedish substance abuse treatment” (approved by the Ethical Review Board of Stockholm, EPN 2016/446-31/5; applying informed consent). The ongoing study involved six substudies, each using different data collection methods as summarized below and further elaborated in a technical report (Storbjörk & Stenius, in press). Here, we refer to these substudies as Sources 1–6 to help the reader appreciate how the propositions were derived from different research methods and samples. Sources 1 and 2 are fully elaborated in published reports. Sources 3–6 refer to ongoing work.

Source 1 refers to analyses of official statistics with national/total coverage to chart municipal expenditures and purchases of services (as opposed to in-house production) from different types of providers since 1999, and producers of residential care since 1976 (Storbjörk & Stenius, 2018).

Source 2 involved analyses of paired interviews of service user (n = 36) and staff (n = 23) experiences of different NPM features (procurement, framework agreements/competitive tendering, performance measurement) in the Stockholm area (Storbjörk et al., 2016; Storbjörk & Samuelsson, 2018).

Source 3 refers to an observation of a procurement process in which a procurement agency served as agent for approximately 80 of Sweden’s 290 municipalities and in competitive tendering concluded framework agreements with about 110 treatment providers.

Source 4 consists of preliminary analyses of interviews with national-level officials involved in development, government, and procurement of addiction treatment (n = 16).

Source 5 implied arranging and analyzing the documented outputs of a 2-day workshop of the Nordic Welfare Centre on addiction treatment procurement practices with 11 participants from Denmark, Finland, Norway, and Sweden.

Source 6 comprises preliminary analyses of interviews with 77 professionals, managers, and policymakers in sampled local and regional treatment administrations representing organizations low and high on NPM permeation.

Results

Neglected implications: Seven propositions calling for attention

We start with the observation that public procurement is at odds with addiction treatment systems (Proposition 1), followed by possible implications of NPM on the planning level (2–3), the composition and functioning of the service provision level (4–6), and effects of market models for service users (7). Our claims are interlinked and somewhat inconsistent, as are features of NPM.

1. Public procurement at odds with addiction treatment.

Rooted in imperative European Union (EU) Directives (Directive 2004/18/EC; replaced by Directive 2014/24/EU), public procurement seems at odds with or has not yet found its ultimate form within addiction treatment. EU legislation on public procurement regulates the relations between public purchasers and different providers, aiming to ensure the free movement of goods and services across borders, and the equal treatment of different providers. The legislation will shape who produces services, which services are produced, and how these services are evaluated.

Health and social services are “no ordinary commodities” (Landsorganisationen i Sverige, 2017), and the directives partly acknowledge national interests: Threshold values for obligatory public procurement are higher, and a variety of procurement forms are accepted. Still, several problems remain: the principle of competition (vs. coordination), the complicated purchasing procedures with its bureaucracy and transaction costs, and the difficulty in defining and following up on quality criteria. Procurement practices may thereby cause a problematic “commodification” of care (Bjerge, 2012; Zaremba, 2013). Treatment systems are searching for (a) a procurement model that corresponds with social/health care legislation, obliging the (central/local) government to offer treatment according to population and individual needs; (b) treatment quality definitions and requirements, including service user participation, good client–therapist relations, and treatment continuity; and (c) models ensuring cost efficiency.

We found that procurement rules and practices differ across the Nordic countries, despite the overarching EU Directives (Source 5). The Swedish Public Procurement Act (LOU 2016; introduced 1992, revised 2007 and 2016) follows the EU directives most closely, adhering to principles of nondiscrimination and equal treatment of all providers. Sweden’s health and welfare sectors, including addiction treatment, increasingly house for-profit enterprises, including venture capital companies (Storbjörk & Stenius, 2018). This has stimulated polarized public debate on profits, and “drainage” of tax money (SOU, 2016).

The Norwegian government, also relying on EU Directives in this case, has favored associations and foundations in addiction treatment by facilitating the exclusion of for-profit providers from competitive tendering (Bogen & Backer Grønningsæter, 2016). Danish surveillance legislation (Lov om socialtilsyn, 2013) enables scrutiny of profit margins and requires financial transparency that, in practice, stops the moving of profits within company groups or abroad. Norway and Denmark have thus taken political and legal efforts to limit competition on the market.

Procurement of services such as addiction treatment can require specific forms (Source 5), recognized in recent procurement law revisions, and more explicitly in Norway, Finland, and Denmark (Hankintalaki, 2016; Anskaffelsesloven, 2016; Udbudslov, 2015) than in Sweden. A negotiation process may be suitable for health/welfare services (Schneider et al., 2016). Problems with procurement may arguably reflect a lack of competence in local administrations. Sources 3–4 and 6 indicate varying local practices of procurements, a slow learning process, and increasing adherence to legislation. It is a major challenge to formulate procurement and contract documents to achieve desirable care and avoid legal disputes, and to reconcile market and health/welfare legislations and logics. One side stresses the rights of the providers, the other the rights of the service users. Longer contract periods, negotiated contracts, and continuous developments and dialogues during the contract period have been implemented in some Swedish municipalities, and more so in Norway (Sources 4–5).

National variations in applying procurement legislation were a key to varying marketization permeation in eldercare (Szebehely & Meagher, 2013). Addiction treatment demonstrates varying market loyalty, and fumbling/experimental searches for good models: The first ambition of research should be to capture variations and pros and cons of different practices.

2. Marketization challenges democracy, equity, planning, and population needs assessments.

Treatment system planning should embrace partnership between authorities, service providers, and users, and should be grounded in local expertise (Ritter et al., 2014). Marketization may compromise such balanced partnership. The politics–market relationship seems to be changing. Earlier policies sought more clearly to mitigate (market) inequalities and secure citizens’ rights, equity, and the common good. Recent policies appear to promote (larger) enterprises that are expanding in Sweden, and to some extent in Finland (Sources 1 and 5) (Anttonen & Meagher, 2013; Esping-Andersen, 1985; SOU, 2016). A strengthened coalition between some policymakers and big companies has been noted in the public debate. That most political parties accept profiteering while most citizens oppose such welfare privatization mirrors this close relation (Lindh, 2015; Nilsson, 2016). Power is redistributed in economized politics (Bergmark & Oscarsson, 1994; Stenius, 1999). Widmalm (2017) suggests a new “enterprise corporatism” of close Government–Capital ties, posing a threat to democracy. The growing power of corporate chains limits the ability of policymakers to direct and regulate the welfare sector (Meagher & Szebehely, 2013), and the role of private providers cannot be to safeguard equity on high-quality services to all citizens (Source 6) (Blomqvist, 2004).

Procurement may increase quality control of the providers submitting tenders but needs assessments and planning are likely to be weakened. Our study suggests that needs assessments are transferred from local to regional/national levels, which compromises local needs. Complex procurement processes are increasingly handed over to procurement specialists and centralized to lower workloads and transaction costs. Norway seeks to favor continuity and comprehensive coverage. Overall needs assessment is problematic in the other Nordic countries when different types of problems and services are handled in numerous procurements (Sources 3–6). Lack of contacts between procurers and social services, and lack of user involvement in procurement are, however, problems acknowledged in Sweden and addressed in the new Finnish procurement legislation (Sources 2–5) (Kaukonen, 2014; Schneider et al., 2016).

Politicians, professionals, and service users may experience curtailed discretion in determining needs and required services based on equity and population needs. The risk is that those with a strong voice in the market—for-profit enterprises, bureaucrats with expertise in contract law and finances, and the most resourceful citizens—decide what, where, and how treatment is provided (Source 2) (Vamstad & Stenius, 2015). Mistrust of the public sector as purchaser and mistrust between providers are increased when procurement laws encourage lodging formal complaints over procedures. Cooperation is undermined.

3. Marketization causes new accountability problems and does not solve monitoring problems.

The question of accountability for the quality and accessibility of care is complex (McLellan et al., 2007). It becomes even more troublesome with NPM (Pierre & Peters, 2017) when administrations responsible for treatment provision are separated from care producers. Accountability is regulated partly by legislation and inspection, and partly by detailed contracts.

The degree of openness in tenders and pricelists (i.e., trade secrets) differs. Swedish competition rules also stress what a producer promises to do in a tender, whereas a provider’s previous negative performances are paid less or no attention. Some local procurement practices ask bidders for references, whereas others argue that subjective judgments and competition laws are difficult to reconcile (Sources 3, 5–6).

NPM is intertwined with a growing auditing bureaucracy (Power, 1997). Performance monitoring can improve quality and control of services and is important regarding accountability (Ritter et al., 2014). Purchasers also declare that procurement demands have led to better quality (and perhaps increased costs) (Sources 3–4, 6), yet quality may also be undermined by profit-making incentives and cost-cuttings (Source 6) (Meagher & Szebehely, 2013). NPM’s tendency to equate quality and accountability with documentation and monitoring of quantitative outputs (Sources 3, 6) may also lead providers to favor statistics of measurable interventions over treatment outcomes (Moore & Fraser, 2013). Contract conformity follow-up is a challenge in all Nordic countries. Of note, no one systematically monitors treatment outcomes (Sources 2–6). Transaction costs increase. Each provider has numerous contracts with numerous purchasers—each of them required to follow up on their contracts.

A mistrust-based contract logic may also cause problems. A contract regime, drawing on competition and mistrust, stresses detailed regulations, instruments of accountability, and routine controls, whereas in a trust-based regime, performance assessments only apply when there is reason to assume unsatisfactory accomplishments (Pierre & Peters, 2017). Treatment professionals are primarily guided by socialized norms and professional standards. Highly detailed contracts and performance criteria may diminish their willingness or possibility to put in that little extra effort—more than the contract obliges (Pierre & Peters, 2017) (Source 6). It was notable how a performance-based payment model pushed professionals toward encounters below their own and the patients’ hopes and expectations. Still, based on their professional beliefs, the staff sought to challenge such incentives (Source 2).

In sum, NPM causes tension between bureaucratic accountability and professional values. Paradoxically, performance measurement systems and contractual arrangements, meant to safeguard performances, may promote gaming the system (Burton & van den Broek, 2009; Evetts, 2009; Lu & Ma, 2006; Pierre & Peters, 2017). As increased monitoring does not solve accountability issues, research should study treatment quality dimensions across jurisdictions in and between countries, with and without outsourcing, or in one region before and after introducing competition and profit incentives. Do professionals in a quasi-market respond to the payer, the profit-maker, or the service user (Source 2) (Gingrich, 2011)?

4. Marketization causes fragmentation and obstructs coordination and continuity of care.

NPM reforms have been found to prevent fundamental preconditions for effective addiction treatment. NPM creates a fragmented system with linear transfer of responsibility (Moore & Fraser, 2013; Nesvåg & Lie, 2010). Calls for tender may aim to promote coordination and continuity (Source 3), but tenders are evaluated separately: bidders compete and are granted individual contracts. A provider may be awarded a contract for assessment but lose the treatment contract to another provider (Source 3). The competition logic weakens interconnections of the system and can be conceptualized as the antithesis of coordination (Bergmark, 2010; Klingemann & Storbjörk, 2016). Post-NPM literature stresses reintegration and needs-based holism (Dunleavy et al., 2006). Finland has attempted to integrate services for improved continuity and lowered costs, but the efforts turned into an immense reform proposal (Finnish Government, 2018) entailing marketization, freedom of choice, and large procurements. This conflicts with the integration goal (Source 5). Continuity, not only for the service user but also for the providers, is also counteracted by short-term contracts (often 2 plus potentially 2 more years) (Sources 5–6). Further, while Norway seeks to counteract such problems by applying contracts with longterm notice (Source 5), long-term fixed contracts can impede response to changing demands and hinder new providers from entering the field (Sources 5–6) (Ritter et al., 2014).

With shifting power relations comes fragmentation affecting overall planning (see #2). Municipalities can gain strength and may put pressure on treatment producers through joint and centralized procurements. But it is also evident that they lose power to decide who will serve as their producer. It is up to the bidders to submit tenders. Purchasers seek to balance their requirements to assure that reputable units will qualify and rogue providers are barred. To do this, the purchasers use specific formulations in the complex procurement documents (Sources 3–4, 6). In this process, decision-makers may lose focus on the most important matters. Public procurement may thus counteract a good mix of services and cause a mismatch between population needs, service demands, and available capacity. Whether this happens and how it can be counteracted should be further studied, as well as how coordination is possible in a system governed by rivalry.

5. Marketization causes regimentation and favors big bureaucratically sophisticated providers.

Stenius (2011) and West (2011) highlight the mix of services for all groups of substance users in the centrally or locally planned addiction treatment systems of the 1980s. Producer type diversity increased in the 1990s (Source 1) in accordance with bourgeois goals. Yet, marketization seems to encourage provider regimentation (Kaukonen, 2014). For-profit enterprises now dominate provision of purchased addiction care in Sweden, at the expense of NGOs (Source 1; see Norwegian alternative, #1). The tendency toward corporate acquisition of smaller enterprises is troublesome, and cartels can manipulate purchasers. As small, often not-for-profit, units have traditionally had an important innovative role in Nordic addiction treatment (Sources 1, 3–5), the new EU directive promotes participation of smaller enterprises. The administrative load, detailed monitoring, and necessity for vast contract law competence in advanced procurements (Sources 3, 6) favors large bureaucratically sophisticated organizations. Detailed procurements may require bidders to demonstrate everything from clean criminal records to provision of organic food (Source 3). Framework agreements do not promise any customers and presuppose financial margins to outlast empty beds. This disfavors small providers and must be considered in pricing if prices are fixed during the contract term—Year 1 appears expensive. In addition, lock-in effects may counteract innovations during contract periods (Kaukonen, 2014; Sources 3–4, 6).

Purchasers’ demands for measurable products may also produce streamlining and uniformity of treatment content and packaging, and hamper consumer choice and needs fulfillment (Bjerge, 2012; Stenius, 2011). Producers may be forced into the categories decided by the procurement documents and service users into package deals of, for example, five sessions per price unit (Sources 2–3) (Vamstad & Stenius, 2015). One development area could focus on how proper needs assessments can counteract potential loss of diversity driven by marketization.

6. Treatment professionals’ values are downplayed when a mistrust-based market logic replaces a trust- and needsbased logic.

As noted in #3, marketization has been claimed to coincide with a shift from trust to contracts and mistrust (Pierre & Peters, 2017; Sulkunen, 2007). Over the last 30 years, Sweden appears to have turned from trust-based to mistrust-based relations in public administration (Montin, 2016), and also toward detailed steering of addiction treatment staff (Creutzer, 2014; Statskontoret, 2016). The bureaucratic control–professional autonomy balance shifts (Christensen & Lægreid, 2011; Gingrich, 2011). Some speak of de-professionalized social and medical professions, prompted by NPM (Bjerge, 2012; Christensen & Lægreid, 2011; Evetts, 2009; Power, 1997), whereas others claim that such control may cause employees to “shirk” at work (Pierre & Peters, 2017).

Procurement principles may undermine the importance of social workers’ judgment and previous experiences of and accrued collaboration with different providers, if they are forced to adhere to rankings in framework agreements. National treatment guidelines and steering documents are incorporated into more advanced procurement documents, which strengthens this “superstructure” (Sources 2–3, 6).

Some providers and bidders manage to provide honeyed images of their organization and activities that do not match the treatment actually provided (Source 2) (Alvesson, 1990). This stresses the importance of the social workers/professionals as a mitigating party, and the significance of continuous follow-up of contracted services. Several logics compete with different views on quality and how it can be captured: Social workers may stress respectful encounters and relational aspects, whereas procurers must rely on checklists of safety aspects and certificates.

We suggest more theoretical and empirical studies on the compatibility of these logics and on how incompatibility issues may be managed to serve the interest of the service users.

7. Marketization may marginalize service users.

Turning to the service user level and to how marketization may affect needs fulfillment and unmet demands, much is already found in previous propositions, such as unification that limits choices.

Evidence is inconclusive, but Jones et al. (2018) found that payment by results led to poorer access and treatment completion. Privatization and performance measurement may disfavor the less “profitable” with the most severe problems by causing customer selection such as cream skimming and more care for the healthy and wealthy (Hansen Löfstrand, 2012; Hartman, 2011; Järvinen, 2002; Kaukonen & Stenius, 2005; Moore & Fraser, 2013; RiR, 2014; SOU, 2016).

NPM suggests stronger customer orientation, but whether it offers substance users more involvement, choice, and patient-centered care is debatable. NPM may obstruct user involvement through the absence of procured services for certain groups, restrictions imposed if desired services are not contracted or top-ranked, or performance-management system incentives (Sources 2–3, 6) (Storbjörk & Samuelsson, 2018). There is an interesting discrepancy between procurers’ voices stressing individual needs over competition/LOU, and the accounts by service users and professionals of limiting effects in daily practice as imposed by LOU (Sources 3–5, 6).

Further, choice mechanisms are rare in addiction treatment (Source 6) (Schneider et al., 2016; Stenius, 2015). A system in which the service user chooses the most appropriate treatment will, claim the advocates, improve effectiveness. A Swedish study showed that the least resourced service users, including addiction services, experienced the least possibilities in choice systems and were the least satisfied with the services (Vamstad & Stenius, 2015)—a social gradient also in terms of choice (Blomqvist, 2004). Moreover, if treatment fails in a system with far-reaching choice, treatment providers may be exonerated: the responsibility can be unfairly imposed on the service users by, for instance, excluding them from further services (Scourfield, 2007). Users become individual consumers that shall choose (and may make the wrong choice). They will not form political pressure groups (Stenius, 2015).

Purchasers may carefully design contracts to prevent cherry-picking. But private providers cannot be forced to take on the responsibility for the well-being of the population. Ritter et al. (2014) argue that purchase practices affect addiction treatment outcomes. As such conclusions are inconclusive, it is crucial to establish how outcome effects are associated with market mechanisms, especially how vulnerable groups are handled in an increasingly marketized society.

Discussion

We have charted how marketization—managerialism, tangible privatization of treatment provision, strengthened for-profit enterprises, and procurement practices—may counteract core treatment systems principles, such as treatment according to needs, continuity, and user participation (Kaukonen, 2014). Population-based needs assessments become scattered and transferred to procurers. Rather than coordination, competition implies fragmentation, mistrust, and reduced possibilities of planning and control. Procurement may promote high-quality services, but long-term contracts with fixed fees may cause lock-in effects and suppress investments and innovations. Market forces may not match treatment demands if profit interests outweigh other principles, such as providing a full range of services to all citizens regardless of consumer strength. Linking monitoring and performance systems to payments in such contractual purchaser–provider relations may obstruct needs assessments and fulfillment when financial incentives arise for what is fed into the systems.

The pros and cons of marketization are inconclusive. Our findings imply the importance of acknowledging and mitigating market principles in treatment systems. Politicians need to make sure that the principles laid out in health and welfare legislation are not jeopardized by procurement laws and market logics. Needs assessments and planning activities shall serve the interests of the service users and the public. Market dimensions can no longer be ignored when conceptualizing and studying addiction treatment systems.

Limitations

Forthcoming publications will provide more conclusions from this ongoing study. NPM was introduced as a costefficient management model, and as Hood and Dixon (2015) show, cost analyses of administrative reforms are extremely difficult to pursue. Our study does not include such cost analyses. However, total costs for addiction treatment have not been reduced in the last decades (Storbjörk & Stenius, 2018). A further question to study would be, in line with Hood and Dixon’s analyses, whether NPM has increased the administrative costs in this sector.