Abstract
Advanced welfare states are under pressure to customize services, promptly enough to prevent a cascade of harms. With these goals, the Netherlands in 2015 decentralized social care services to municipalities, and within municipalities to neighborhood teams in continuing contact with clients. The overall results have been disappointing. But the experience of Utrecht, the Netherlands’ fourth-largest city, has been strikingly different. By using hard-to-resolve cases to signal conflicts in rules, obstructive jurisdictional boundaries, and the shortcomings of private service providers, Utrecht is learning to customize and speed delivery of social care through incremental steps. This article explains how Utrecht's success addresses apparently intractable limits to the adaptability of the rule-bound welfare state, such as the problem of low-level discretion or street-level bureaucracy and the division of services into silos, in the process bridging, and perhaps effacing, the gap between the Habermasian life world and the system world of formal rules.
The welfare state has never been so much in need of reform nor seemed so incapable of it. The demands placed on publicly provided welfare are changing and becoming more urgent. Technological development and global trade are dividing the economy and workforces of advanced and developing countries into a modern sector that contributes to and benefits from global innovation and a low-productivity, low-skill, and low-wage sector that does neither. Jobs in the stagnant sector are precarious; adequate housing is hard to come by; debts accumulate; health care, at least in some countries, may be unaffordable or of low quality. This turbulence magnifies the effects and complicates treatment of the psychological problems, family conflicts, and difficulties in school or with the police or drug abuse at the heart of social welfare work. The changing needs of new groups strain standard responses, underscoring the limits of professional expertise and accentuating calls to include beneficiaries, their families, and communities in decision-making and caregiving. Demands for coordination of programs grow too, since families and individuals struggling with one problem likely encounter several, and failure to address problems in one domain easily undercuts progress elsewhere. To be effective today by way of prevention or cure, the social welfare system, in coordination with the schools, the police, or the housing authorities, must provide and periodically adjust integrated, “wraparound” supports fitted to the needs of individuals in “multi-problem” families. 1
But the current welfare state in all its traditional configurations is notorious for two kinds of congenital, rule-bound rigidity that exclude continuing adjustment of individual programs and coordination of complementary services. The first rigidity, rooted in the nature of hierarchy, is familiar as the problem of street-level bureaucracy. The teacher behind the closed classroom door, the police officer on the beat, and the social worker on a family visit, all but unobservable to superiors, interpret the rules made at the top of the organization to say what they think the situation requires, this or that reform notwithstanding. Adding more rules to limit this discretion only increases the likelihood of conflicting instructions, inviting yet more street-level interpretation. The second rigidity, rooted in the division of labor, is familiar under the headings of organizational silos or wicked problems. Social needs crystallize into government programs with budgets and eligibility rules. Struggles among departments and their lobbies over additional resources reinforce the boundaries thus created. Cooperation across jurisdictional lines is discouraged if only because it is unclear at the outset who in the end will pay or benefit. 2 Together the problems of street-level bureaucracy and silos arguably reduce any bureaucracy—and certainly any public administration operating under the manifold constraints of representative democracy—to an incoherent jumble of formal rules rendered serviceable by informal, ad hoc workarounds that address local issues, at least temporarily, but at the price of silence about the large changes they might implicate.
There has been a prolonged effort, starting roughly in the late 1980s, to sidestep these problems by privatizing provision of public services and relying on market mechanisms to ensure that they are adaptive, high-quality, and cost-effective. By relieving public officials of the burden of actually delivering services, these new public management (NPM) or neoliberal reforms aimed to focus the attention of government on what were presumed to be its proper tasks: goal setting—specifying service levels and characteristics—and monitoring the performance of private providers. Competition was to complement the discipline of oversight as the users of services, empowered to choose the provider best suited to their needs, shifted resources to favored firms. But the separation of the conception of service delivery from execution deprived administrators of the information required to devise effective incentives or detect manipulation. Providers met program goals by creaming: supporting client customers who need little or no help to achieve their goals while limiting expenditures on those that do. Competition was no corrective because service users seldom knew when they were ill served or if there was a better alternative. In countries such as Australia, where the reforms were undertaken, at least in part, to enlarge individual choice, and where administrators tried over decades to correct the problems of successive procurement rounds—in short, where the reforms were given a fair chance to succeed—the failures have been as great as where reform grew from budgetary concerns or animus against the poor. In service industry after service industry, concentration increased as firms used the advantages of their size and position to secure client classifications they could cream, limit costly service options for the hard-to-serve, reduce staff qualifications to cut costs, and otherwise sacrifice service quality to profitability. In desperation, the regulators abandoned the hope of increasing the flexibility of welfare programs and imposed standard requirements to limit abuse. 3
The failure of neoliberal reforms has left a moraine of new problems, starting with the demoralization of many social workers, 4 and invites the despairing question of whether there can be an alternative to the rule-bound welfare state we have inherited. More tailoring of welfare services, and coordination of welfare with other domains of the public sector, may be imperative. But how, on the record of recent experience, could this be accomplished?
Yet the insistent demand for effective social welfare and other fundamental public goods that help constitute the polity in meeting common needs makes resignation all but unthinkable even in the face of seeming exhaustion of the standard alternatives of market and state. 5 The growing dualization of the economy and the explosion of inequality that goes with it give rise to persistent, populist protests, often directed against the degradation of public services and cuts in benefits. Responses to the pandemic have underscored, harshly, the scandalous consequences of unequal access to public services and, most generally, the need for these services to develop the capacities for rapid and continuing adjustment increasingly demanded in welfare. Add to these considerations the often unsustainable fiscal burden of many welfare programs, and it is clear that experimentation with new organizational models will continue, cautionary tales of failed reforms and the absence of any consensus on the way forward notwithstanding.
Among current initiatives, the Dutch reforms of 2015 stand out for their boldness and scope and even more for the headway they are making in addressing some of the apparently intractable problems that bedevil not just the provision of welfare but all public services provided by large, bureaucratic organizations. The heart of the reform is a double decentralization. First, responsibility for providing welfare services to children and families, to those at the margins of the labor market, and to elderly and disabled adults is decentralized from the national level to municipalities; then, within municipalities, responsibility for receiving and evaluating new clients and referring them to specialists when necessary is decentralized to a new type of neighborhood-based team. This team is the client's case manager, coordinating and monitoring further interventions. The legal right of citizens who meet specific criteria to corresponding benefits has been in effect replaced by the municipality's obligation to provide care tailored to their individual circumstances (known in Dutch as maatwerk). To meet this obligation, the municipality is given wide autonomy to determine the conditions of access to service, whether service is provided in cash or in kind, and whether service is organized directly by the local administration or contracted out to specialist providers.
The long-term organizational goal of the reform is to shift from the current “relay” model, where limited integration of services is achieved serially, as clients are passed from agency to agency, making their cases to a succession of strangers, to an integrated team model in which the client and neighborhood team, in different configurations, face successive problems together. 6 This shift is presumed to be necessary and, among some reformers, also sufficient to enable early, preventive intervention as opposed to later treatment, and integrated responses suited to needs of individuals or families rather than fragmentary supports reflecting what programs happen to offer more than what clients demand. At its most ambitious, the reformers aimed to mitigate if not overcome the tension between the Habermasian “life world” of needs and moral commitments and the “system world” of formal procedures and institutions that give them substance.
Taken as a whole, the outcome of the reform after eight years in operation is disappointing. There has been neither a dramatic improvement nor a dramatic deterioration in the provision of welfare services, at least so far as this can be detected in data available at the national level. Welfare budgets, reduced when the reform went into effect in anticipation of efficiency gains, must often be supplemented by extraordinary grants from national funds as individual municipalities hit spending limits. Beyond customer satisfaction surveys, which are intrinsically of limited value, there is little information on outcomes, and none on changes in the frequency, duration, and severity of bad spells for individuals and families with particular conditions or other measures of the kind needed to rigorously evaluate the effect of the reforms on the quality of treatment. But neither is there any reason, in view of the widely reported differences in municipal experience, to suspect that in the aggregate treatment quality has improved or deteriorated. Viewed nationally, the outcome of the reform so far is, surprisingly, how little has changed. 7
But the aggregate results conceal more than they reveal. A key purpose of the reform was to encourage innovation in the provision of welfare services at the municipal level. In fact, at least one large municipality—Utrecht—has made remarkable progress in institutionalizing the capacity for flexible adaptiveness, including in contractual relations with private welfare providers. It has done that, moreover, while extending the coverage of service provision and within the agreed budget constraints, itself a notable achievement since many municipalities have incurred significant cost overruns even in the absence of ambitious institutional reforms. Given the stakes and the dispiriting record of failed reform, this advance, invisible in pooled, national results, commands attention.
Utrecht's master innovation is to systematically use individual welfare cases in which a current rule or jurisdictional boundary obstructs integrated provision of a tailored support to trigger prompt review of the decision-making process within and across departments. The cumulative effect of these reviews can be the reform, even transformation, of ways of working within the municipal administration in the absence of any precise, initial blueprint for reform.
Utrecht's master innovation contributes as well to addressing the problem that has plagued welfare and other street-level bureaucracies from the start: control of discretion. By enabling and eventually obligating discussion of questionable rules and boundaries, Utrecht ensures that changes that in a traditional bureaucracy would be accomplished by one-time exemptions authorized by a benevolent superior, or by recourse to a slush fund for special circumstances, are instead made by deliberation among the relevant decision makers at various levels; and this deliberation can give rise to formal changes in municipal and, occasionally, national rules and regulations governing welfare benefits. As it becomes routine, this continuing reconsideration of the aptness of rules and jurisdictional boundaries calls into question (without yet providing a full-fledged alternative to) the familiar idea of administration as a hierarchy in which rule making is the sole prerogative of the topmost authorities. 8 More fundamentally, recognition of the likelihood of revision changes the very nature of the rules themselves, making them less like rigid prescriptions and more akin to provisional guidance or rebuttable instructions, presumed valid until there are compelling arguments to the contrary. 9
This same routine and collaborative review of obstructions affords public administrators frequent opportunities to engage external—private or not-for-profit—welfare providers at work, and thus to assess their willingness and capacity to adjust programs to changing and differentiated needs. These engagements reduce the information asymmetry between the external provider and the public counterpart—the asymmetry being the bane of new public management—and make it possible to begin using the contracts as an instrument to advance cooperation rather than for entrenching the self-interested practices of the private party. Just as administrative rules are becoming rebuttable presumptions, contract terms are coming to be seen as setting provisional or indicative goals, to be revised jointly as experience accumulates. 10
The Dutch reforms of welfare and the explicit shift to individualized care ultimately raise large questions about accountability that go beyond Utrecht and cannot be resolved at the municipal level. The welfare state as we know it is committed to universal norms: all citizens, or at least all citizens in roughly similar conditions, are to be treated alike. In emphasizing services, in contrast, the Dutch reform supposes that with respect to many kinds of welfare problems, general similarities, let alone the fact of common citizenship, can be misleading. To be effective, services must take account of the particularities of individual and familial experience that distinguish one beneficiary from another. If equal treatment for all, or all in a similar condition, no longer applies, to what standard can citizens and courts hold public administration in apportioning services adjusted to individual need? This debate has been broached in courts at the national level, with important clarification of the requirements for investigation and well-motivated decision-making that public agencies must assume in providing individualized care.
The body of this essay analyzes the innovations underpinning Utrecht's organizational advances; the conditions under which these emerged; and the possibilities for spreading them directly to other municipalities or, more promisingly, complementing them with national reforms that incentivize and enable municipalities to learn rapidly from each other. To keep the analysis manageable, and because it implicates the widest array of collateral municipal services, we focus primarily on youth care. The analysis draws on the rich documentary sources generated by the municipality and other public bodies, as well as seventeen interviews with protagonists of the local reforms and participant observation of a series of problem-solving meetings in various settings. 11 The rest of the argument is in seven parts. The next part quickly retraces salient developments of the Dutch welfare system leading up to the reform. The second part presents the basic institutional choices that committed Utrecht to bold reforms carried out through incremental steps offering ample possibilities for self-correction. The third part shows the proliferation of case-by-case problem solving as a means of fostering coordination across organizational silo boundaries. The fourth part extends the discussion of organizational reform by showing how these problem-solving mechanisms mitigate the problem of information asymmetry between the municipality and external service providers so that contracting out does not become synonymous with derogation of official responsibility to provide high-quality service adjusted to changing needs. The fifth part focuses on the selection and training of the social welfare workers in Utrecht and the way their participation in interpreting and articulating clients’ needs and in reconsidering rules and organizational boundaries creates a novel professional role that straddles and may help integrate the life world and system world. The sixth part looks at the way Dutch administrative law is clarifying the obligations of reason giving in official determinations of individual needs and benefits and suggests how these procedural reforms might be the germ of a principle of justice for establishing the welfare entitlements of individuals regarded in their particularity rather than as members of broad categories. The conclusion reports important signs of the recognition in the Netherlands of the promise of the Utrecht approach in its diffusion to other municipalities and the national government.
Be forewarned that we have little to say about political or other peculiarities that might explain why Utrecht, rather than any of the many other municipalities that it strongly resembles, developed a strikingly original and promising approach to the decentralizing reform. In part this is simply because the implementation of the new measures took place far below the surface of everyday politics, and there are no salient differences to report. Time may bring to light intramural conflicts that did shape outcomes. But experience suggests skepticism about the possibility of detecting previously hidden, systematic causes even with the benefit of hindsight. Social science is good at identifying causes when cases are alike on all dimensions but for a crucial one, as in the familiar examples of microlevel randomized control trials of the effectiveness of deworming treatments or various education subsidies and Millian critical case comparisons of macrohistorical or macrosociological events such as peasant revolutions. 12 Where, as here, cases differ subtly on many dimensions, exceptional outcomes will result from relatively small variations on common themes, and their precise causes will be hard to trace. Finally, even if some causal preconditions for Utrecht's success could be established, and reformers the world over were impressed by the example, they would no doubt strive to reproduce it at home, all convinced—and some, unpredictably, with reason—that they can adapt the model to the conditions they face or change the conditions to favor the model.
The Road to Reform
By the late twentieth century, the Netherlands had developed an administratively centralized but sectorally balkanized welfare state in which the municipalities, with little taxing power and policy discretion, played only a subordinate part. Successive efforts after World War II to rationalize the historical accretion of local, voluntary, associational, and denominational welfare programs into a unified, national system on the model of the UK Beveridge Plan were blocked by the social partners and the confessional parties. But the postwar insurance and transfer programs covering pensions, unemployment, health, and long-term care were all nationally administered, whatever the mix of public and private actors in their governance. As legislation further extended the scope of the central government to key social services such as housing, welfare work, and assistance to the poor and disabled, national direction displaced the vestiges of local control. 13
Criticism of this centralized system from the left and the right mounted from the 1980s on, in concert with complaints about one-size-fits-all benefits and services, sectoral fragmentation, and, as a result of these, rising costs and lengthening waiting lists, especially in health and long-term care. 14 After the turn of the century, the criticism intensified and, partly under the influence of contemporaneous debates in the United Kingdom, turned in the direction of a new localism and a reconceptualization of the role of professionals in service provision that would shape the reform legislation.
The left-leaning social care professionals and journalists advocating this new localism argued that the breach between the life world of citizens and the institutional logic of the system world had been exacerbated by NPM's curbs on the discretion by which professionals adapted rules to meet clients’ needs. Indeed, to guard against criticism, many professionals had themselves become rule followers, further distancing themselves from the life world of their clients. As a corrective, the localist reformers called for professionals to support citizens in mobilizing their own resources (eigen kracht) and those of their social networks (burgerkracht) to solve individual and collective problems in the neighborhood—bypassing the state with its rules. 15
These ideas gained wide currency as they flowed together with the notion of a localist, voluntary, and devolved or decentralized “big society,” developed by the British Conservatives under David Cameron. In the variant circulating in the Netherlands, one pillar of the big society was an alliance of frontline, social welfare professionals with citizens and communities in autonomous, co-owned associations to renew the provision of public services. 16 The kindred reform proposals were synthesized in the call for a fundamental reform of the Dutch welfare state toward a “participation society” in the 2013 King's Speech, drafted by Liberal prime minister Mark Rutte, and endorsed by his Labour coalition partners, on the eve of the enactment of the laws decentralizing welfare. 17 Dutch Liberals saw the participation society as a vehicle for achieving the dream of a smaller state, while Labour embraced the concept as a renovation of local solidarity bringing government closer to citizens. 18
This call culminated in 2015 in three linked reforms mandating a radical decentralization of responsibility to and within municipalities on the principle of “one family, one plan, one case manager.” Many stars aligned to favor their passage. The burgerkracht reform vision, now a component of the “participation society,” was embraced not only by a broad cohort of social care professionals fed up with bureaucratic rules and NPM protocols, but also by an influential cross-party group of politicians at both the national and local level. 19 At the same time, policymakers and professionals concluded that the existing system of social care was simply unsustainable. An official evaluation of the previous Youth Care Act found, absurdly, that it often took longer and cost more to get a diagnosis for an individual condition than for a treatment that, at best, addressed only an isolated part of the client's problems. 20 There were immediate, short-term considerations as well. At the national level, the coalition parties and government ministries embraced the decentralizations as the most promising way of achieving substantial budget savings on social care without drastically cutting services at the height of a major financial crisis. Municipalities, for their part, were prepared to accept sweeping cuts of 15 to 40 percent in the budgets transferred from the central government in order to realize their long-held aspirations for greater autonomy in a crucial policy field. 21
Despite the long gestation period of the reforms, little was done to facilitate rapid learning from the partial successes and failures inevitable in their implementation. Plans for national monitoring of ongoing results in cooperation with the municipalities never got off the ground, beyond a small Transition Bureau. Nor did the municipalities themselves make special provisions for pooling their experience.
One reason for this inattention to the foreseeable difficulties of implementation was municipalities’ resistance to intrusive national monitoring, out of fear their hard-won policy autonomy would be restricted before it had been exercised. 22 National policymakers, moreover, appear to have been misled by past experience to overestimate their capacity to steer local implementation of the reforms at a distance through administrative guidance and parliamentary motions, as well as the ultimate sanction of placing municipalities with serious budget overruns in receivership. 23 In retrospect, national policymakers also plainly underestimated the strains that such far-reaching decentralizations would place on planning bureaus, inspectorates, and research institutes developed for review of administration in a unitary state. Ruefully the Council for Public Governance (Raad Openbaar Bestuur, ROB) has recently recommended that “no further decentralizations be implemented without first thinking through how the administrative level responsible for the new task will be provided with the necessary knowledge.” 24
Eight years after the transition to the new regime, there is growing concern that the great decentralization is failing its transformative goals, especially in the field of youth care. Interim evaluations of the Youth Act and reports of national inspectorates and local courts of auditors find that most municipalities have made little progress toward institutionalizing an integrated, demedicalized approach to youth care based on a support plan developed with the child and family. Collaboration of the neighborhood teams responsible for generalist basic youth care with specialist care providers and other key local partners such as schools and general practitioners remains underdeveloped. 25 A number of large cities, such as Amsterdam and The Hague, have massive social welfare budget deficits, triggered by efforts to reduce waiting times for services through referrals to costly specialist care. The national government placed several other municipalities, including pioneers in the initial phase of the reform such as Eindhoven and Leeuwarden, in temporary receivership to address budget overruns. In Rotterdam, waiting lists grew so long that the national Inspectorate for Health and Youth Care reported a general “blockage in the system.” 26
Within this bleak national landscape, however, at least one large municipality—Utrecht—has been making good progress toward the reform's original objectives as outlined above. Utrecht's success in advancing the reform's goals is attested not only by external and internal evaluations but also by informed observers attentive to local variations, including several of the fathers of the decentralizations themselves. 27 Strikingly, too, Utrecht has managed to transform its youth care system without running up significant budget deficits, 28 while at the same time providing help to more than 15 percent of the under-18 population, the highest proportion in the country. A good measure of the ability to customize youth care is the ratio of outplacements (to institutions) to ambulatory (nonresidential) care. Utrecht's score on this measure was the lowest by a large margin of any major city in the Netherlands. 29
In the next section, we analyze the key steps in Utrecht's progressive reorganization of the youth care system and its relations with other municipal services focusing on the innovative mechanisms underpinning this ongoing institutional transformation.
The Utrecht “Model”: Making Routines and Rules Routinely Revisable
Utrecht is a socially and politically unremarkable city of some 360,000 residents in the center of the Netherlands. It is the smallest of what the Dutch refer to as the G-4 municipalities, after Amsterdam (873,000), Rotterdam (651,000), and The Hague (546,000). On measures of social stress such as the proportion of low-income households, children in poverty, early school leavers, and share of immigrants in the total population, Utrecht scores somewhat below its peers, though well above the national average for these indicators. More telling, Utrecht has substantial areas of concentrated and severe socioeconomic deprivation, with large immigrant populations, notably Overvecht and Kanaleneilanden, that are fully comparable to like neighborhoods in the other big cities. Indeed, Utrecht is the most socially segregated of the G-4. 30
Utrecht does not stand out politically either. In the last decade the city council has been governed by a succession of center-left/center-right coalitions. This has been the pattern in Amsterdam as well, while Rotterdam and The Hague have seen the repeated surge and ebb of far-right anti-immigrant parties that mark national politics. 31 Neither is there anything distinctive about the level of education or degree of experience of the social welfare workers most directly implicated in Utrecht's decentralizing reforms, nor, for that matter, in the municipality's early announcements of its reform ambitions.
In sum there was and is nothing in such coarse indicators to suggest that Utrecht would respond to the common problems of decentralized youth care provision in a successful, innovative way. So far as we know, on the eve of reform no canvas or informal list of municipalities of any size well-positioned to put decentralization to good use included Utrecht. We, for example, started our research in Rotterdam, as its reform proposals looked on paper to be as promising as any. We only shifted attention to Utrecht as a succession of innovations there, most conspicuously among the G-4, has touched off a self-reinforcing and self-correcting process of decentralization.
The first of these innovations was institutional processes for developing and testing new approaches toward the broad goals of the reform (from roughly 2011 through 2014). Utrecht recognized the dangers of a “big bang” transition and explored key institutional choices in intense, collaborative, small-scale experiments with key stakeholders: 32 a process of months-long brainstorming and revision in the manner of charettes familiar from the initial stages of architectural or industrial design. 33 In these neighborhood-level proeftuinen (test beds or experimental plots in the sense that such terms are used in agricultural research), Utrecht developed a vision of the decentralized youth care system in which frontline care would be delivered by integrated teams, with members drawn from different social work disciplines but working as generalists with youth and family clients. In Utrecht these are called neighborhood youth and family teams (Buurtteams Jeugd & Gezin). 34
As would become regular practice, the two neighborhoods for the initial proeftuinen were chosen for their different socioeconomic characteristics and distinctive challenges: in Overvecht immigrant families predominate and the schools play a leading role in youth care; in Ondiep residents are predominately native Dutch families, and general practitioners and other health care professionals are key to youth care. The teams themselves deliberately chose not to define their procedures too precisely at the outset, in order “to discover in practice what works and what not,” with “as few rules and as much professional freedom as possible ... in order to do what is needed (doen wat nodig is).” 35
The team members worked in pairs to combine different types of expertise in meeting each family's needs and to learn as much as possible from one another. Joint case review (casuïstiekbespreking) within and between the teams was regularly used to solve individual problems, identify good and bad practices, and develop policy advice for the new youth care system. Organizationally, such joint case reviews with the municipality and external partners were also used to establish provisional boundaries between generalist and specialist care and to develop cooperation with adjacent domains. In matters of child and public safety, for instance, “the whole Utrecht safety chain, including the police, neighborhood managers, child protection, and the neighborhood teams were brought around a single table to discuss their respective roles on the basis of concrete cases.” 36
Several foundational design choices related to the character and selection of the providers of the new services could not, however, be tested in proeftuinen, though they were informed by extensive prior experience: Did Utrecht want to employ the neighborhood teams directly (as for example in Rotterdam)? Hand them over to a consortium of established providers (as in The Hague)? Create a joint venture between the city and external care providers (as in Amsterdam)? And if Utrecht did not want to provide the service itself, how would it go about choosing or building the external organization that would?
In line with its broader approach to the transformation, which emphasized the need for ongoing “co-development” with residents, professionals, and societal partners, Utrecht did not want to provide youth care directly. 37 But at the same time the municipality was determined that the neighborhood teams be fully independent of the existing care providers to avoid the constraints of established organizational routines and habits, as well as conflicts of interest in the referral of clients to specialist care. 38 To reconcile these requirements, the municipality found an elegant compromise combining the advantages of a new organization with a continuing commitment to the principles and practices that had informed the design of the teams. Proposals from incumbent care providers for the creation of Youth and Family Teams were rejected. Instead, the municipality contracted with a wholly new organization, soon to be known as Lokalis, assembled by an experienced manager from a local specialist care provider. The Lokalis proposal was judged to be more promising than competitors’ both because it committed more fully to the cocreation of services with the municipality and because it committed the new organization to take over the members of the existing teams working within the proeftuinen and recruit new employees from the existing local service providers. 39 The choice of Lokalis thus built on recent experience while safeguarding the space for joint innovation created in the run-up to decentralization. 40
A further foundational choice concerned the financing of the neighborhood teams and specialist care providers. The neighborhood teams would be funded not through an hourly rate or fee per client, but through population-based costing: a fixed amount per year linked to the number of children, adjusted for the socioeconomic characteristics and needs of the various neighborhoods. The aim here, and through a similar system of fixed, annual payments to specialist providers, was to encourage the teams to meet their clients’ needs without the distraction of calculating billable hours per case, while at the same time stimulating them to “de-escalate” interventions where possible. 41
Case-by-Case Problem Solving and Institutional Transformation
In the years following the initial proeftuinen and design by charette, Utrecht continued innovating with the creation of multidisciplinary, cross-functional roundtables to tackle the problems within the problems that emerged as decentralization of social care proceeded. The initial aim of these roundtables, which variously combined neighborhood team workers, officials from different municipal services, local medical professionals, and specialist care providers, was to find solutions adapted to individual client needs in cases deadlocked by existing rules and routines. But in time these roundtables came to be seen as a mechanism for using the resolution of difficult, individual cases to detect and correct broader problems in frontline practice, municipal organization, and even national welfare programs: casuistics (casuïstiek), originally developed within Dutch social work as a means of training and a system of quality control, thus also became a method of reevaluating current routines, rules, and jurisdictional boundaries.
More often than not this problem-solving process resulted in the (further) decentralization of relevant parts of the municipal administration and specialist care services to the neighborhood level to work together with the youth and adult teams in an integrated, multidisciplinary network—a kind of neighborhood-based welfare state (or at least as much of one as necessary to approximate comprehensive social care)—that could both solve local problems and suggest ways of applying the solutions citywide. This further decentralization reshaped in turn the relationship of the neighborhood teams to specialist care providers and of municipal domains such as work and benefits and housing to one another.
Redrawing the Boundaries between Basic and Specialist Care
In the years following the decentralization, as demand for youth care surged and referrals to specialist care rose rapidly, specialist care providers ran into capacity limits and budget ceilings, forcing the municipality to choose between raising a provider's budget, redirecting the client to another inside provider, or bringing in an outsider.
To facilitate and learn from these critical decisions, Utrecht established a municipal-level multidisciplinary body, the Appropriate Alternatives Committee (Commissie Passend Alternatief, or CPA), to review cases that had hit budgetary limits or program boundaries and determine whether solutions could be found or created with the city's current providers. The committee, meeting roughly once every two weeks, consisted of experienced frontline practitioners, a neighborhood team member, a child welfare specialist, and a pediatrician or general practitioner, chaired by a representative of the municipality. Often the cases involved, for example, proposals for inpatient institutional care, the most disruptive for kids and families, and also the most expensive, in response to which the CPA proposed intensive ambulatory support as a cheaper and more effective alternative. 42
But in 2017 the CPA's workload increased by almost tenfold, to 560 cases, from 65 in 2016, and it became clear that the body had become a forum for devising ingenious ways to relax the budgetary constraints on the inside providers. The problem, the municipality concluded, was that the CPA intervened at the end of the care chain, too late to affect the decisions that produced a mismatch between the demand for customized services and the supply. Intervention had to come much earlier in the decision-making process: the neighborhood teams and the inside providers working with them had to deliver simpler forms of customized help themselves, so that clients could be helped before their problems escalated to the point that institutional outplacements or other complex specialist care became necessary. 43 Utrecht's solution, arrived at in steps through experimental pilots, was first to establish decentralized roundtables where frontline teams and specialist providers jointly develop (potentially generalizable) strategies for stuck cases, and then to create in the neighborhoods multidisciplinary teams of specialists to advise the Youth and Family Teams and provide lighter forms of specialist care themselves.
The first step was the formation in 2017 of neighborhood-based Customization Roundtables (maatwerktafels) to enhance the capacity of the teams and the inside specialist care providers to find custom solutions without referring cases to the CPA. 44 These roundtables meet about once every week and a half, depending on the demand. The case manager invites the relevant specialist care providers, other youth care professionals, and social network members, and usually (for part of the meeting) the child and parent(s) as well. A representative of the municipality chairs the meeting and liaises with the municipal youth administration and other municipal departments when issues of jurisdiction or collaboration arise. 45
A second step, taken later in 2017 through the launch of the “Extr@Teams” pilot, was to decentralize not just case management but the also the provision of key aspects of specialist care to the neighborhood. Together with Lokalis and a group of specialist care providers, the municipality created integrated teams for “neighborhood-oriented specialist youth care,” with attention as in the original proeftuinen to distinctive local problems, based on profiles developed in consultation with the buurtteams and family doctors and historical patterns of referrals to specialist care. In a prosperous semisuburban area with many higher-educated parents, for example, the team included specialists in complex divorces and learning problems, while in a mainly autochthonous working-class district the focus was on broader family problems, including mental health, truancy, and domestic violence. To improve the delivery of these specialized services, but also to guard against the possibility that some groups were excluded from care or underserved—say, because the initial characterizations of differing local needs reflected embedded expectations and inarticulate prejudice rather than the true distribution of demand—the scope of the Extr@Teams’ work and the adequacy of the resources deployed were regularly reviewed in discussions with clients, family doctors, child protection workers, and other professionals from each neighborhood. Because of their thorough investigations of individual cases and ability to customize treatments, the Extr@Teams reported less reliance on DSM (Diagnostic and Statistical Manual) indications, and greater willingness to question requests for standard indications such as AD(H)D, than conventional specialist care providers. In both neighborhoods, referrals to specialist care declined substantially in 2018, especially for residential outplacements, not only compared to the two previous years but also compared to others where no Extr@Teams were active. The Extr@Teams’ comprehensive view of local cases also helped identify patterns, such as a wave of related referrals from a specific school, that had previously gone unnoticed because of the fragmentation of specialist care. In this way they contributed directly to the other major area in which Utrecht uses roundtables, pilots, and joint case review to revise rules and routines: the reconsideration of the relationship between social care and other public or quasi-public domains such as work and benefits and housing—de-siloization. 46
Revising Rules and Routines across Public Domains
From an early stage, external observers of the Utrecht model drew attention to the constraints on the transformation of social care imposed by bureaucratic routines in adjacent domains at municipal and national levels. Thus, for example, the 2014 report by the local Court of Auditors on Utrecht's transition plans warned that “many families’ problems have to do with income and debt issues. The municipal department of Work and Incomes [Werk en Inkomen, W&I] works with strict protocols, and there is little room to take the family context into account in decision making.” 47 The 2016 visitation committee, while praising the collaboration of the city and the neighborhood teams, observed that “other municipal domains . . . still work in a much more traditional way . . . with many specialized regulations.” 48
A national initiative, the Inclusive City Deal (agreed in March 2016 by Utrecht and four other municipalities with the Ministries of Internal Affairs, Health, and Social Affairs), provided the impetus to tackle the relationship between the transformation of social care and other public domains. 49 Its goal was to explore the scope for alternative integrated forms of social care better adapted to actual family and community conditions. The participating cities identified the bureaucratic rules and procedures or “system elements” that stand in the way of such an integrated approach through joint review of 100 concrete cases (20 per city). The ministries considered removing statutory and regulatory obstacles to integrated provision of social care identified by the project, focusing predictably on standardized administrative procedures that produce self-defeating outcomes, as when a family faces eviction because municipal creditors garnish its welfare benefits to recover what is owed them. 50
Of the participating municipalities, only Utrecht established a multidisciplinary roundtable to help the neighborhood teams deliver customized solutions across multiple social domains through joint case review. 51 In February 2017, Utrecht launched Ondiep Ontregelt (Ondiep Disrupts) in one of the original proeftuin neighborhoods as its local experiment within the Inclusive City Deal. 52 The core idea was to tackle cases that had gotten stuck because of legal and regulatory restrictions across different domains, such as benefits, debts, housing, and education. Such stuck cases could be brought by local social workers to a weekly City Deal meeting, where a multidisciplinary group developed customized solutions, making exceptions to established rules and organizational routines where necessary. The City Deal meetings consisted initially of representatives from the municipal W&I department and the two neighborhood team organizations, but was broadened to include the major local housing corporations and health insurer. Alongside the Ondiep City Deal meetings, which took place in the neighborhood itself, the municipality established a parallel citywide meeting, to which social workers from other districts could also refer stuck cases. In addition, Utrecht created a bimonthly “Learning from the City Deal” meeting, with top officials from the neighborhood team organizations and the departments of Social Support (MO) and W&I, where the findings of the weekly roundtables dealing with individual cases were reviewed to identify and agree on structural changes in policy and services needed to overcome the problems encountered. 53 When the Inclusive City Deal initiative formally concluded in late 2017, 54 Utrecht extended the approach to other neighborhoods with similar problems, and continued to convene a metalevel review body to revise rules and policies. 55
One of the chief innovations of the City Deal meetings was reform of debt assistance. In one of the original proeftuinen, the neighborhood team worked with a local housing corporation on a debt early warning system. In more recent pilot projects, this early warning system has been expanded to other neighborhoods and to other major creditors such as the health insurers and utility companies. Participating creditors alert neighborhood teams to potentially unmanageable debt, following which the teams discuss remedial measures with clients before the problem becomes critical. In neighborhoods with high levels of debt problems, W&I officials review individual cases with neighborhood team members and the household itself on a weekly basis. There is also a “red button,” instituted by the national City Deal, that neighborhood team members can activate through the municipality to seek solutions from the Unemployment Insurance Agency, the Tax Office, or the Central Judicial Collection Agency. 56 Similar forms of collaboration have led to an agreement between the municipality, the semipublic housing corporations, and care providers (including the neighborhood teams), with the participation of tenants’ and clients’ organizations, to provide suitable living space with appropriately adapted support services for specific vulnerable groups, including the homeless, young adults in crisis, refugees, victims of domestic violence, and those with mental health problems, especially autism. 57
The actual process by which rules and routines are adjusted depends on the initial organization of decision-making and the distribution of authority and interests; and the path from recognition of perversities to effective revisions is typically circuitous. For example, a pilot project on prevention of youth homelessness attached to the City Deal convened representatives of W&I, the neighborhood teams, specialist care providers, and youth workers, in biweekly meetings to review stuck cases. A promising solution, developed in collaboration with a local housing corporation, was to provide youth in assisted living facilities with apartments in buildings scheduled for demolition in the coming years, coupled with ambulatory support services, in preparation for transition to independent living in standard housing. But the rent for these apartments substantially exceeded the maximum subsidy allowed for persons under twenty-three under the national housing rules. As the proposed solution was better for the young people and cost-saving, the municipality first bridged the gap from its own funds, using its prerogative to grant “special social assistance” (bijzondere bijstand) in urgent cases. But as, by law, such special assistance can only be awarded to individuals, but not groups of similarly placed claimants, the municipality changed its own rules and raised the local housing subsidy for young people to the adult level. Complaints from Utrecht and other municipalities about the perverse effects of the national rules on youth homelessness then moved the Minister of Housing to reduce the eligibility age for receipt of the full housing subsidy from twenty-three to twenty-one. 58
These innovations—design in proeftuinen or by charette and casuistics in maatwerktafels to correct obstructive rules and jurisdictional boundaries—share a commitment to fallibilism: the recognition of the practical impossibility of devising, even through exploration and experimentation, a fully workable plan of institutional reform ex ante, and the consequential need to continuously reevaluate and revise initial designs. In rejecting “big bang” solutions that aim at comprehensive improvement of failed institutions, fallibilism is not limited to tinkering that leaves the foundations of the existing order intact. On the contrary, the municipality of Utrecht does not hesitate to decisively reallocate resources or decision-making authority among organizations or groups of professionals when case-by-case problem solving indicates the need. But transformation, when it occurs, is the result of a cumulative learning process, and the bold actions that are taken are themselves subject to incremental correction.
Learning by Contracting
Contracting for services from private providers has been the hope and despair of welfare states for decades: the hope, because the freedom to collaborate with outsiders enmeshed in the current experience of client families, and to change partners when expectations are disappointed, promises to free the state from the rule-bound bureaucracies of its own creation; the despair, because renegotiating service contracts at long intervals, and unable to observe the behavior of its partners directly, the state repeatedly agrees to terms that allow the service provider to do what is to its own advantage, not the clients’—to “cream” by concentrating on the easiest-to-service cases. 59
An unintended but now welcome side effect of Utrecht's commitment to minimizing the organizational and professional obstacles to the provision of customized services is a change in the working relations of the municipality and its outside partners that makes the latter's performance regularly and directly observable, reducing the information asymmetry that typically puts the public at a disadvantage in renegotiations. These changes clear the way for the eventual introduction of some form of contracting under uncertainty of the type already in use among private actors who, like Utrecht and its service providers, can only specify through the process of collaboration itself what each needs to do.
The key to the new relationship, and the mutual transparency it affords, are the frequent roundtables used, as we just saw, to find customized solutions to particular, stalled cases and identify rules, routines, and forms of cooperation across organizational boundaries that should be changed to better facilitate such outcomes in the future. These are working meetings. Representatives of various teams and departments within the municipality—sometimes from the neighborhood, sometimes from the center, or both—discuss the details of refractory cases and weigh alternatives with representatives of the relevant service providers, among others. On these occasions the municipality can observe directly what the different providers understand by “customized” services; how they respond to the suggestion of possible innovations, or if they take the initiative by proposing innovations themselves; and whether they regularly deliver on their commitments.
In one meeting of the CPA we observed, for example, the participants questioned in several cases whether the treatments requested by the neighborhood team worker—for depression alongside autism, for family psychoeducation, and for help with language development—should not be supplied by one of the specialist providers as part of its normal offerings. In a Customization Roundtable we attended, the participants asked why none of the specialist providers offered a combination of autonomous living space with daily therapeutic guidance for teenagers with psychological problems transitioning toward self-sufficiency. 60
Where the problems identified at the roundtable involve an individual service provider, as in the first example, the chair passes on the information to the municipal account holder (the formal contract partner), who takes up the issue with the service provider in an informal meeting or site visit, as well as in the next, regular quarterly review. Where the problems involve cooperation among a group of specialist providers, as in the second example, the roundtable chair convenes a separate meeting of the parties to discuss how to resolve the issue, with the possibility of referring matters to their respective account holders if cooperation does not improve. Where a broader pattern of problems is identified through the roundtables, for example regarding the relationship between specialist and general care providers (neighborhood teams, family doctors), the municipality organizes meetings on the problem area with the contractors to discuss how best to address the underlying issues. Next steps can be a joint review of anonymized cases in order to clarify which kinds of issues can be handled, for example, by the neighborhood teams and which need to be referred to specialists. The provisional solutions identified through this process are then elaborated through proeftuinen and pilots, before being generalized across the city, as with the Extr@Teams and neighborhood-based specialist care. 61
Alongside the roundtables, the municipality has a second major tool for redressing the information asymmetry with external contractors: a comprehensive overview of current movements of clients among service providers, derived from the administrative metadata associated with referrals to specialist care. From these data the municipality can spot, for example, churning of clients between specialist care providers, and it can take up the issue with them in ad hoc site visits or regularly scheduled review meetings with the account holder. The same tool can also be used to challenge the (frequent) claim of contractors with long waiting lists that the delays are due to the particular complexity of their caseload. More powerfully still, information from the roundtables can be combined with the metadata to serve as a check on the plausibility of contractors’ assertions generally. Thus, as the municipal youth care controller told us, the Customization Roundtables provided them with a “random sample” of cases; and the rich case information often makes it possible to interrogate the contractors’ claims, for example, regarding children with disabilities. “With that information we say to the providers, your own experts tell us that 30 percent of children with disabilities could have been helped in a different way [through the neighborhood teams], but we see that they’re still over there [with you].” 62
The effectiveness of these mechanisms for symmetricizing information is reinforced by the fact that the account holders for the external contractors are mostly municipal policy officers rather than specialized financial controllers or professional accountants. This means that the account holders focus in their discussions with specialist care providers on achieving the policy goals of the Utrecht youth care system and have sufficient substantive expertise to evaluate the contractors’ arguments. In this way the policymakers themselves develop a more holistic perspective and a better understanding of practical implementation problems. 63
The cumulative effect of the continuing revision of the working relation between the parties through these mechanisms is to change their understanding of the very nature of the agreement between them, if not its formal terms. The contract becomes in effect a framework or platform for the discussion of what each currently expects from the other, given their shared and jointly revised understanding of what is needed and possible.
To grasp how such inchoate agreements can function, it is helpful to look at the very similar, but more developed, contracts in the private sector between parties, at their frontier of knowledge, who undertake a joint innovation whose feasibility and form can be determined only in the course of their collaboration. Such contracts for innovation are commonplace in domains as diverse as biotechnology, IT, and advanced manufacturing. As in Utrecht, change in the form of contracting has been driven in these sectors by an increase in uncertainty, understood generally as the inability to anticipate problems in realizing some end, let alone the solutions to those problems, in advance of actually working toward the goal.
Under uncertainty, the parties are by definition unable to specify their respective obligations, as they would in a standard contract. Instead, they agree on broad goals and a regime for exploring the most promising approaches and regularly evaluating the prospects of success. The regime, fixed in the contract for innovation, provides for periodic, joint reviews of progress toward interim targets or milestones; procedures for deciding whether, and with what exact aim, to proceed, or not; and mechanisms for resolving disagreements. By exchanging this information, the parties clarify the shared goal and improve their assessments of one another's capacities and reliability. Mutual reliance increases as collaboration progresses, and the cost of switching partners rises accordingly: for as the project advances, it is extremely unlikely that it will be possible to replace a current partner by another with equivalent capacity and experience. This mutual reliance deters opportunistic defection and can eventually trigger norms of reciprocity. Trust is as much the result of the process of collaboration as its precondition, just as the precise aims of cooperation are the outcome, not the starting point, of joint efforts. 64
While it is of course unclear how much of the formal mechanisms of contracting for innovation Utrecht (and eventually other municipalities) will adopt, there is no doubt that the core procedures of ongoing revision of the de facto agreement in light of frequent, joint efforts to solve concrete problems are well established. In the latest development of the Utrecht model, the municipality has sought to deepen and intensify its continuous collaboration with external contractors in the transformation of youth care by consolidating the seventy-odd specialist providers inherited from the pre-2015 system into two purpose-built organizations responsible for delivering neighborhood-based specialist care in different areas of the city. In selecting these new service providers, the municipality organized an elaborate process of “dialogic tendering,” 65 involving an intensive series of interviews and role-playing exercises with policymakers and a variety of stakeholders (including the neighborhood teams and a group of youth clients) in order to assess not only the applicants’ commitment to the vision and leading principles of the Utrecht model, but also their capacity to apply them in different practical situations. 66
Frontline Professionals between the Life World and System World
Utrecht's innovations in the design of welfare organizations build on and transform the self-understanding and skills of frontline professionals in the buurtteams and other municipal departments. The active participants in proeftuinen and pilots are trained to think of welfare rules as provisional: subject to revision in light of a convincing showing that they do not serve the larger, lawful purposes to which they are committed. By taking part in such rule revision, the leading frontline professionals step back from the welfare workers’ traditional role of caregiver, representing the substantive interests of the life world against the formalism of bureaucratic process. Instead they become active participants in both, perhaps blurring by this novel combination of responsibilities the boundary between them.
The Netherlands entered the reforms with a comparatively well-trained workforce in social services. In a recent international survey of skill levels in employment services, for example, 84 percent of Dutch workers had earned the equivalent of a four-year degree at a German Fachschule, as against 45 percent in the United Kingdom and 35 percent in Australia. 67 But despite these superior formal attainments, Dutch social workers are not prepared for the kind of interdisciplinary cooperation demanded by the Utrecht reforms. 68 Attempts to address such problems by reconfiguring professional training have been repeatedly broached but only fitfully implemented. 69 Utrecht has therefore had to devise its own methods to attract and form the workforce it needs to provide customized services. For instance, Lokalis takes a page from the practice of apprenticeship and asks applicants to actually demonstrate a skill indispensable to an eventual job—here, casuistics: early in the selection process candidates must successfully engage in close, collaborative discussion of a case to proceed further. 70
For training in how to do maatwerk, Utrecht draws on the Institute for Public Values (lnstituut voor Publieke Waarden, IPW). The IPW was founded in 2010 by two brothers prominent in the movement for decentralization; one conducted a field study of the self-defeating tendencies of welfare bureaucracy as experienced by clients. The IPW focuses on the roughly 50 percent of all those drawing on social services and supports with three or more problems, implicating five to ten different programs: the multiproblem families who, in a splintered service landscape, are disproportionately likely to encounter rule conflicts and contradictory advice (as when a doctor insists on a rest and a participation coach a redoubled job search) and whose cases, frequently stuck, most depend on customized solutions to advance. 71
To unblock such cases, the IPW has developed a “breakthrough” (doorbraak) method by which frontline workers directly address legal and organizational constraints in light of continuing immersion in the clients’ circumstances and challenges. The starting point is a thorough discussion with the clients of long-term goals and immediate blockages, which typically reveals that while the blockages are only remotely related to the underlying or root causes of the clients’ difficulties, progress is impossible unless they are overcome. In a case from the IPW's course materials, for example, a teenager recently released from detention has already secured a place in a training program likely to lead to a steady job. But he cannot support himself in the months before training begins, and by law he is only eligible for unemployment insurance once he has spent a month, from the day of application, searching for a job. Unless this blockage is overcome, the chances of recidivism, already substantial, approach certainty. 72
Identification of the impediment then triggers a search for legal openings through which to introduce contextual arguments persuasive enough to break the impasse. In the case of the recently released teenager, for example, a law governing unemployment assistance itself provides that aid can be granted to a person with “no right to assistance” if, “in consideration of all circumstances,” “very compelling reasons necessitate this.” Similarly, in another case from the materials, a mother with a young daughter cannot secure a place in public housing because of her debt problems, and she cannot get help restructuring her debts because she does not have a fixed address; a municipal mandate making it illegal for mothers with children to sleep on the street authorized resolution of the Catch-22. 73
Once the legal plausibility of a customized solution is established, the additional costs of the purpose-built measures are weighed against the gains to society and the client as a final check that no interest, public or private, has been disproportionately burdened or benefited with respect to the other. 74
In Utrecht, the IPW taught the doorbraakmethode to the interdisciplinary teams that took part in the initial pilot of the City Deal, Ondiep Ontregelt, and its offshoots in other, neighborhood-based tables—now renamed, in recognition of the currency of the term, doorbraaktafels.
75
The language with which Utrecht describes its approach to customization is becoming indistinguishable from the IPW's. A recent public announcement asks, How do we put the life world, rather than our system world at the center? Where do the laws possibly clash? . . . We as professionals prefer to look with one another at what room the law offers. The room whereby, together with residents and partners, we can reach solutions for those with problems in multiple life domains.
76
Though the IPW operates nationally, Utrecht's use of its methods is, as the foregoing suggests, atypical. Elsewhere the method is employed mainly to resolve individual cases, as the IPW itself intends. Reports of the experience of thirteen other municipalities that participated with Utrecht in a pilot program for homeless youth under the tutelage of the IPW focused overwhelmingly on exemplary case outcomes. Utrecht, in contrast, generalizes the method into a principle of institutional design, continually creating, as we have seen, settings in which breakthroughs in individual cases prompt reconsideration not just of rules but also of the routines and organizations in which they are embedded. 77
Although it is far too soon to estimate the effects of the diffusion of the doorbraak perspective on the self-conception of social-service professionals, use of the method will foreseeably reshape their understanding of the relation between social norms guiding discretionary action and the legal provisions and bureaucratic structures constraining it. In simultaneously helping to articulate clients’ needs, and reconfiguring the rules and structures that give them substance, social welfare professionals doing maatwerk are no longer traditional street-level bureaucrats, continuously choosing between (self-protectively) enforcing alien rules or vindicating the morality of caregiving against the law, in the discretionary spaces opened by the gaps and conflicts in the rules. 78 Instead the frontline workers and their colleagues might be said to be infusing the formal procedures of law with the substance of moral commitments even as their articulation of the moral commitments becomes more cognizant and respectful of legal normativity. The IPW points to the novelty of the situation: evoking Ronald Dworkin's metaphor of discretion as judgment exercised in the hole of the doughnut formed of rules and regulations, the institute writes that “to deliver maatwerk professionals have to stand with one leg in the doughnut hole and walk with the other on the dough.” 79 Less figuratively, and in the jargon of the reform movement itself, we venture that the practice of customization, concretized in the doorbraakmethode as propaedeutic for case analysis and principle of institutional design, is the avatar of a world, still almost unimaginably remote, in which the very distinction between the substantive morality of the life world and the processual formality of the system world is effaced.
Customized Service Provision and Equal Treatment before the Law
Utrecht's success in adjusting rules, professional roles, contracting practices, and jurisdictional boundaries that obstruct individualized welfare raises new and fundamental problems. As municipalities differentiate services, taking the individuality of beneficiaries’ circumstances explicitly into account, the current reform undermines the constitutional guarantee of equal treatment to those presumed to be in the same circumstances of need.
Most directly the new and still evolving responsibilities of decision-making in public administration compel reconsideration of administrative law. Instead of applying rules, fixed principally by legislation, to determine eligibility for standard benefits, officials, as we have seen, must ascertain the conditions of individuals and families in need of assistance, and sometimes reinterpret the conditions of eligibility as they recombine elements of existing programs, perhaps with innovative supplements, to respond to the needs they find. What assures citizens that the discretion inherent in such case-by-case determination of entitlements works to their benefit without exposing them to abuse in moments of high vulnerability? Can a welfare regime that treats citizens as individuals also remain faithful to the commitment of constitutional democracy to treat them as equal before the law?
These and related questions have recently appeared on the horizon of public discussion in the form of widely noted court decisions reviewing decisions of the social welfare administration. In this section we draw on two, complementary lines of judgments to look ahead to the possible articulation of alternative principles of justice in a welfare state attentive to individual condition. The first set of judgments concern judicial requirements for administrative decision-making and its justification. In reviewing recent appeals against municipal determination of welfare benefits, the Dutch administrative review court (Centrale Raad van Beroep, CRvB) is defining the scope of the requirement for customized service provision and the key elements of the decision-making process by which a social welfare agency can reach, accountably and effectively, a customized determination of benefits. By making a searching investigation of individual circumstance and deliberative evaluation of the findings a test of the persuasiveness of decision-making in the individualizing welfare state, the CRvB in effect gives citizens a right to customized consideration of their claims and allows judges to verify that administrators have proceeded accordingly in particular cases. The new, judicially imposed process requirements draw on and reinforce the doorbraakmethode's purposeful approach to interpreting and revising rules, which is likewise based on careful empirical investigation and deliberative review. Together they can become the foundation of a right, anchored both in administrative practice and in law, preserving equal treatment before the law in an age of deliberate differentiation.
The second set of cases concerns not the process of decision-making but rather the standard for review of the substance of the final outcome. 80 The Council of State (Raad van State, RvS), the highest administrative court in the Netherlands, responding to widespread outrage at the draconian penalties imposed by the tax authorities in disputes over day-care subsidies, 81 have overturned their own precedents and found that essentially all administrative decisions are subject to a final, all-things-considered judicial test of their proportionality: a determination that the public purpose pursued by government policy does not disproportionately burden affected interests. 82 Like the final step in the doorbraakmethode, this kind of comprehensive assessment is aimed at avoiding partial and perverse outcomes, where single-minded pursuit of one goal frustrates achievement of others. Reading this set of cases together with the first, citizens are protected against abuses of discretion by both an entitlement to a decision-making process that demonstrably considers their individuality and a final check that, process safeguards notwithstanding, decision makers have not given undue weight to some interests—typically those of the state—at the cost of others. Together these safeguards hold out the promise of a highly contextual form of decision-making by both administrators and courts—required to do justice to individual condition—which is nonetheless disciplined enough by requirements of fact finding and reason giving to conform to the rule of law.
Many of the cases regulating the process of decision-making and insisting it have a deliberative character arose as municipalities reauthorized existing services, such as help to impaired persons with home cleaning. Utrecht, for one, allocated 80 percent of existing clients a standard service of 1.5 hours per week, while the rest received customized help, 83 touching off massive protests and a flood of complaints (871 formal objections between January 2015 and October 2016). 84
The discontent in Utrecht led to two decisions by the CRvB requiring the municipality to conduct an “independent” and “objective” investigation into the individual circumstances of each client to reach a service allocation enabling autonomous living. Indirect consultation, with the service provider or a council representing welfare recipients, was judged insufficient, as was an investigation focused primarily on financial considerations. 85
Other decisions by the CRvB have further specified the investigation of individual condition by insisting that experts be consulted by the municipality when necessary; that their expertise meet professional standards; and, crucially, that experts address and assess the client's needs for service provision, even when the legal claim as presented could be decided without taking those needs into account. 86
Together these cases define what citizens may expect as a matter of right when advancing claims for customized services. As tersely summarized in a guidance letter recently circulated by the Dutch Association of Municipalities (Vereniging van Nederlandse Gemeenten, VNG) to its member municipalities, “Every citizen can apply for help ... and is entitled to a careful procedure.” Although an applicant's financial strength may, together with other capacities for self-help, be taken into account in deciding benefits, “a careful access procedure and the provision of individual customization are and will remain the central principles.” Above all the “CRvB has also emphasized time and again the importance of conducting a careful investigation, supported eventually by expert advice and well-substantiated justification in the reporting.” The quality of these investigations “requires permanent attention from municipalities.” 87
Additional commentary by the RvS emphasizes that the “new vision of care creates a heavy burden of reason giving in concrete cases.” While rules will remain “indispensable in the new relationships,” they must, to serve the purpose of customization, be “approached and applied differently.” “The whole course of decision-making determines its quality; the investigation into the concrete, personal circumstances, the provision of information to the person concerned,” are relevant to assessing the legitimacy of the outcome. This concern for the chain of decision-making as a whole reflects, for some observers, a general tendency in administrative law to subject to judicial review not “only the final decision but also the process towards it.” As a practical matter, quality control of the decision-making process within administration “will have to focus more on the professionalism and quality of . . . judgment,” as gauged by peer review and the other instruments normally used for this purpose. 88
The RvS's insistence of a final, judicial review of administrative interest balancing was a direct outgrowth of public dismay at the tax authority's apparently willful indifference to the well-being of recipients of day-care subsidies who failed to comply with the letter of the law, if only because private placement services, acting on their behalf, had defrauded both them and the state. Dutch administrative law provides as a general rule that decisions are subject to judicial review of their proportionality. But there is an important exception: when a law explicitly prescribes commitment to one goal to the exclusion of others, it is exempted from proportionality review in deference to the legislature's authoritative weighing of interests. As the law governing day-care subsidies made the beneficiary's unfailing compliance with the obligation to pay her share of the costs of care a condition of support, decisions enforcing this requirement were exempted from proportionality review regardless of the consequences; and the administrative courts accordingly did nothing as the tax authorities, often with the intent of discouraging or punishing fraud, forced beneficiaries at the brink of indigency to repay thousands of euros in support for failure to pay tens or hundreds of euros in delinquent fees. Indeed a prominent decision of the RvS in this period unequivocally reaffirmed that the decisions were not subject to proportionality review. 89
Outrage at the manifest inequity of the tax authority's insistence on enforcing the letter of the law without regard to the consequences forced reconsideration. Reexamining the provision in the law on administrative procedure governing exceptions to the requirement of proportionality review, the RvS found that it had erred. Read properly, the law did indeed require respect for the legislature's preferred weighing of competing goals. But this respect, the court is understood to have found, does not preclude review—and judicial correction—of outcomes. On the contrary: proportionality review is necessary even when the lawgiver has decided preferences to avoid “harrowing” or “distressing” (schrijnende) outcomes. 90 Thus courts reviewing final administrative decisions, no less than frontline workers applying the doorbraakmethode to cases stuck in a logjam of rules, are called on to ensure that conflicts are not resolved to the cost of the vulnerable and against the overall intent of the law. In this sense we might imagine courts and social workers as implicit partners in integrating the life world of actual experience and the system world of formal rules and procedures.
In setting requirements for fact finding, reason giving, and judicial review in individualized social welfare decisions, Dutch law is both encouraging the development of the new welfare state and controlling its exercise of discretion in accord with the rule of law. By directing attention away from legal formalities and to assessment of the quality of decision-making—what in the United States is called process or “hard look” review, to distinguish it from review of conformity to formal, procedural requirements or a minimal check that the outcome is not inconsistent with the evidence considered 91 —Dutch law puts continuing pressure on officials at all levels, from those ascertaining facts to those making and revising rules, to base decisions on an informed understanding of individual circumstance. This safeguards citizens against unwarranted official discretion, in sharp contrast with US practice, where concern with the choice of decision-making procedure, and with formality generally, often dominates judicial review of administrative acts, reducing considerations of substance to an afterthought in judicial opinions and academic commentary. 92
This insistence on a thorough and reliable decision-making process in the allocation of welfare benefits suggests finally a way of thinking about an ideal of justice when like treatment in like conditions will no longer do. Pushed to the hilt, the jurisprudence of the CRvB and RvS entitling applicants for welfare benefits to “a careful procedure” recognizes a right of each citizen to call when need be on a responsive administration, willing and capable of understanding what makes claims distinctive. This right to a common process fully respecting their individuality is what citizens share when membership in broad categories defined by general rules no longer captures what they have in common.
Conclusion
By way of conclusion, we look to the gathering indications that the Dutch, belatedly attentive to the successes of their bold, decentralizing reforms, are embracing elements of the Utrecht model.
Surging municipal budget deficits and long waiting lists for specialized care have heightened interest in local experience. A joint steering group convened by the VNG and the government called for institutionalized exchange of experience among municipalities. 93 Pressure for such a framework increased in 2021 as negotiations over the budget overruns deadlocked and the VNG exercised—for the first time—its right to independent arbitration of differences with the government. The arbiters’ report largely concurred with the municipalities’ demands for additional financial support, 94 while noting pointedly that “there is room for improvement by learning better from the best-performing municipalities.” 95
Meanwhile, elements of the Utrecht model have been diffusing to other municipalities and to national organizations. Nijmegen, a midsize city in the southeast of the Netherlands, after costly missteps, has embraced Utrecht's approach to organizing generalist youth and family care in integrated neighborhood teams. 96 Amsterdam, buffeted by the highest welfare budget deficits in the country, is likewise emulating Utrecht, creating neighborhood hubs of specialist providers to work closely with the frontline teams and making the teams themselves the responsibility of an independent foundation, on the Lokalis model. 97 Within the framework of the national City Deal, Tilburg and other smaller municipalities in the “Heart of Brabant” region have established a multistakeholder “doorbraak team” to apply maatwerk in hard cases and a higher-level work table to identify systemic problems and possible solutions. 98
Driven by the fallout from the day-care subsidy scandal, national ministries and administrative bodies are embracing maatwerk and joint case review to meet legal requirements and the expectations of public opinion. The Customization Program for Multiproblem Households (Programma Maatwerk Multiprobleemhuishoudens, PMM), for instance, created by a remarkably encompassing consortium of ministries and administrative agencies, has established a “national escalation team” (Landelijke Escalatie Team, LET) to which stuck cases involving national administrative bodies can be referred for resolution. While Utrecht is particularly active in referring questions and cases to the PMM, twenty other municipalities also make use of its services, and a further forty-one are preparing to do so. 99 Case-by-case reform of rules and routines guided by the doorbraakmethode is also being embraced by administrative agencies beyond the circle of those participating directly in the decentralizing reforms. For example, the Unemployment Insurance Agency (Uitvoeringsinstituut Werknemersverzekering, UWV), notorious for its rigid implementation of bureaucratic rules, has created municipal and national fora where multidisciplinary teams of managers, jurists, policy workers, and implementation specialists first pursue individual solutions to complex problems and then seek to generalize them more broadly. 100
If the Utrecht model continues to withstand scrutiny, this is just the beginning. In a deeply unsettled age, in which even the most successful nations acknowledge the need to learn from others as their own institutions falter, such continued local success with vital problems will command global attention. The Dutch decentralization was the result of pressures common to the welfare states in all the advanced countries: the need for prompt, preventive provision of services more customized to individual need, coordinated across the range of problems that often co-occur in vulnerable families, and at affordable cost. The promise of Utrecht is that continuing, collaborative evaluation by families and public and private decision makers of needs, laws, and organizational boundaries can in providing those services also begin to reintegrate the life world of our experience and hopes with the world we build of formal systems that today frustrate them.
Footnotes
Appendix: Authors’ Interviews,Meetings,and Participant Observations
Seminar on “Learning by Monitoring and Joint Case Review in Youth Care” [Leren door monitoren en gemeenschappelijke casusbeoordelingcasuïstiekgesprekken binnen de jeudghulp], co-organized with the VNG Support Team for the Youth Care Landscape (Ondersteuningsteam Zorglandschap Jeugd), Amersfoort, June 7, 2018.
Research meeting with officials of Gemeente Utrecht and Lokalis, December 3, 2018.
Participant observation, Appropriate Alternatives Committee (CPA), Gemeente Utrecht, March 7 and 21, 2019.
Interview with Wieke Westgeest, policy advisor, Maatschappelijke Ontwikkeling/Societal Development and Coordinator City Deal meetings, and Joop van der Zee, advisor, betekenisvol sturen/“Meaningful Steering,” Gemeente Utrecht, March 22, 2019.
Discussions and correspondence with Lisa Huibers-van Tetering, policy advisor and coordinator CPA and Customization Roundtables, and Wieke Westgeest, Gemeente Utrecht, March–December 2019.
Interview with Rianne Ruiter and Kim ter Beke, team managers, Lokalis, March 25, 2019.
Participant observation, City Deal meeting, June 6, 2019.
Interview with Rianne Ruiter, Lokalis, July 5, 2019.
Interview with Ruud Ilbrink, family social worker, Lokalis, July 5, 2019.
Interview with Floor Roks, team manager, Samen Werken aan Veiligheid (SAVE), September 3, 2019.
Interview with Wytse de Jong, senior policy advisor and account holder for Lokalis, Gemeente Utrecht, September 4, 2019.
Interview with Jos Linskens, team leader, Department of Work and Incomes, Gemeente Utrecht, September 4, 2019.
Participant observation, work visit of the Lokalis supervisory board to Extr@Team Leidsche Rijn, September 13, 2019.
Participant observation, Customization Roundtables, September 21 and 26, 2019.
Participant observation, Learning from the City Deal meeting, September 24, 2019.
Interview with Jeanine ten Haaf, program manager, Lokalis, October 10, 2019.
Meeting with Riane Ruiter, Lokalis, and Marlies Kennis, youth program manager, Gemeente Utrecht, December 12, 2019.
Interview with Pieternel Boerenboom, trainer and senior policy advisor, Gemeente Utrecht, February 20, 2020.
Interviews with Souhail Chaghouani, business controller, Maatschappelijke Ontwikkeling (Social Development), Gemeente Utrecht, February 20 and May 5, 2020.
Interview with Marian Dobbe-Kluijtmans, director, KOOS Utrecht (specialist youth care contractor), April 17, 2020.
Interview with Marenne van Kempen, director, Lokalis, April 20, 2020.
Interview with Albert-Jan Kruiter, co-founder and co-director, Institute for Public Values (IPW), March 14, 2022.
Interview with Jessica van der Toorn, project manager, City Deal Eenvoudig Maatwerk, and Bregje Spaans, project leader, Pilot Approach to Youth Homelessness, Gemeente Utrecht, July 15, 2022.
Acknowledgments
First and foremost, we thank our interview partners for their generous and stimulating cooperation, as well as the leadership of Lokalis and the City of Utrecht for facilitating our research. Earlier versions of this article were presented at the University of Amsterdam, Columbia Law School, and an international online seminar. We are grateful to the participants for helpful comments and suggestions; to Mark Considine, Sigrid Quack, Stefan Berger, Onno de Zwart, Kim Putters, and other researchers at the Sociaal en Cultureel Planbureau; and to the editorial board of Politics & Society.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research has been supported by the Radboud Excellence Initiative Scheme, granted to Charles Sabel by the Radboud University.