1. Introduction and rationale
In this paper, we argue that a certain level of indeterminacy in policy design (indeterminacy of policy objectives and contents, of formal planning of the change management process, of the expected give-and-take balance for the main stakeholders) may be a factor facilitating rather than hindering system-level change, provided the management of change is based on actors' agency that, combining with context features, leads to the triggering of concatenations of social mechanisms affecting the inner dynamics of the change process. The key message of our study is that system-level change can happen also without being directly produced by the design of the policy; indeed, some vagueness in the policy design may open up opportunities for actors to interact and effect change at later stages of the policy process. The causal textures of the dynamics of the policy process may be read through the theoretical lens of the analysis of the activation of concatenations of social mechanisms. This study aims at shedding light on such concatenations of social mechanisms and their interplay with indeterminacy in policy design, by drawing on the analysis of a case of systemic change in chronic disease management in the Italian region of Lombardy occurred over 2016–2017, when a radically novel governance of chronic disease for the 10 million population of the region was put in place (the health care system of this region was later tested to its limits by the COVID-19 pandemic outburst which reached dramatic intensity here, though this event is analytically unrelated to the episode of system-level change recounted in this contribution).
In the case observed, policymakers' focus has been the legal enactment of the reform bill – mostly consisting of relatively broad-scope and vague prescriptions about institutional and service innovations to tackle the growing epidemiological shift from acute to chronic needs – coupled with some extensive political story telling (‘selling the reform’) and the setting up of venues for continuous and adaptive interactions with stakeholders. The starting point of our analysis is the observation that the prescriptions of the reform were quite vague and potentially conducive to limited change, or to outcomes profoundly different from the ones observed, and there is evidence of policy-makers dithering in their stances on both the ends of the reform and the means to achieve it: quite plainly, the design of the reform was akin to a recipe for limited and at most fragmentary change. Still, substantive systemic change in the governance of the welfare for long-term illness actually occurred, and intended and unintended processes determined such level of change. We claim that such processes may be studied by means of the conceptual tools of the social mechanisms, a form of mid-level theorizing for the explanation of social phenomena, whose usage for explaining specifically policy and administrative change has been advocated by leading scholars in the field of public policy and management (see recently two collective efforts led by Howlett, Ramesh, and Capano: Capano et al., Reference Capano, Howlett, Ramesh and Virani2019 and Reference Capano, Howlett and Rameshforthcoming).
In this paper, we advance this research agenda aimed to explore the explanatory power of social mechanisms in public policy and administration specifically when applied to innovation in the governance of the welfare and to systemic change, which we consider as a specific type of change. We aim to contribute to the analysis of social mechanisms as a broad research agenda carried out collectively by a range of prominent scholars in sociology in general and public policy more specifically (like Barzelay, Gallego, Hedstrom or Swedberg – see Hedstrom and Swedberg, Reference Hedstrom and Swedberg1998; Barzelay, Reference Barzelay2003; Barzelay and Campbell, Reference Barzelay and Campbell2003; Barzelay and Gallego, Reference Barzelay and Gallego2006, Reference Barzelay and Gallego2010) by making a contribution specifically to bettering our understanding of what specific concatenations of social mechanisms can explain system-level change under conditions of indeterminacy of the policy design. We define system-level change as change occurring in both the rules and the organization of the delivery of a public policy.
In particular, we challenge the notion of the policy designers' purposeful social action in interaction with deliberate policy design as the decisive factors, and emphasize the functions performed by the triggering of social mechanisms in affecting the outcomes of a change process of the public governance in a given area of public services. Under certain context conditions, the triggering of concatenations of social mechanisms may perform key functions in the process of change, to a significant extent irrespective of and beyond the deliberations of policy designers, whereas other social actors' agency may perform a relevant role in the change process. As regards the way in which we operationalize the notion, we use the notion of ‘policy indeterminacy’ to refer to a design of the contents of the policy (the reform package) characterized by, first, vagueness of the prescriptions contained such that radically different outcomes could ensue from these contents whereas the claim can be upheld that prescriptions have been met, and, second, such that the reform package refers to the adoption of provisions to be taken subsequently to substantiate and detail the exact contents of the reform. It may be queried the extent to which policy indeterminacy can be operationalized as a gradation, or to what extent it is a binary condition. Our, at least tentative, answer to this pertinent question is that policy indeterminacy, although it can indeed be seen as a gradation, ultimately can be treated as a binary condition for the purposes of delineating the contours of the domain of applicability of the findings of our study. Accordingly, we would consider most public policies, and even more so in sectors where technical knowledge and expertise is a key factor as is the case of the health care sector, to be generally characterized by ‘determinacy’ of the policy design, and yet there are instances in which the level of indeterminacy is so relevant to enable classifying it as ‘high’ on policy indeterminacy – and this condition may open us a set of process dynamics which differ from the (more common) circumstances under which this is not the case. This is what we investigate in this study (it may be noticed for clarity that this conceptualization of the term ‘indeterminacy’ is altogether different from definitions in economics to refer to non-calculability or non-determination of the value of a variable).
Our research question concerns how can system-level change to the welfare system under conditions of indeterminacy of the policy design be explained, and our tentative answer is that the analysis of the triggering of certain concatenations of social mechanisms as specified in this paper and shedding light on how developmental patterns of change may unfold (Van de Ven and Poole, Reference Van de Ven and Poole1990) provide a valuable repertoire of conceptual tools to this purpose. Our empirical question is explaining the change process that occurred in chronic disease management in the case of the Italian region of Lombardy. Finally, we envisage our broader contribution to lie in contributing to research agendas aimed at explaining policy and administrative change (as a subset of broader processes of social change), and in producing a form of utilizable knowledge for effective change management in welfare governance and public governance at large.
This paper is structured in seven sections: following the present introduction to the argument, section 2 reviews selectively the literature on social mechanisms and, where directly pertinent to the purposes of this contribution, the literature on the policy process; section 3 discusses methods; section 4 then provides the case study evidence; section 5 examines the dynamics that led to the kind of stakeholders engagement that has enabled systemic change; the role of a certain concatenation of social mechanisms for explaining change is discussed in section 6 and the final section wraps up on the findings.
2. Theory: the analysis of social mechanisms as conceptual tools for explaining change
We adopt the analysis of social mechanisms as the centrepiece of the analytical apparatus. Social mechanisms can be defined as conceptual tools for the analysis of complex change processes. The analysis of social mechanisms is an approach to reveal the multiple causes of change through context-sensitive accounts (Pettigrew, Reference Pettigrew1990; Pollitt, Reference Pollitt2013; Ongaro, Reference Ongaro and Pollitt2013b) and specifying the social ‘cogs and wheels’ of investigated phenomena (Elster, Reference Elster1993: 3), an approach with a long history in the social sciences (Merton, Reference Merton1968) and application largely in the field of sociology and only more cursory and intermittent in applied social sciences like organization studies and public management, possibly also due to a relative disconnect with the mainstream frames adopted in these applied fields (exceptions include Barzelay, Reference Barzelay2003; Barzelay and Gallego, Reference Barzelay and Gallego2006, Reference Barzelay and Gallego2010; Capano et al., Reference Capano, Howlett, Ramesh and Virani2019; Melloni et al., Reference Melloni, Pesce and Vasilescu2016).
Social mechanisms can be defined as ‘unobserved analytical constructs that provide hypothetical links between observable events’ (Hedstrom and Swedberg, Reference Hedstrom and Swedberg1998: 7). Elster (Reference Elster1989, Reference Elster, Hedstrom and Swedberg1998) and Stinchcombe (Reference Stinchcombe1991) interpret social mechanisms as building blocks for an advocated middle-range theorizing which can enable a revitalization of the study of social phenomena by getting beyond covariation between variables or events, and rather aiming at investigating the causal texture of social phenomena (rooted in a social ontology encompassing both individual agency and social structure as explanations of social processes, see Ongaro,Reference Ongaro2020, chapter 4). This approach has strong roots in sociology, as illustrated by Boudon (Reference Boudon1991, revisiting the study of Merton, 1968: 43–44 in particular). Scholars in the field of public governance and public policy and management argue for the renewed significance of the analysis of social mechanisms for the study of institutional, policy and administrative change (Capano et al., Reference Capano, Howlett, Ramesh and Virani2019; Ongaro, Reference Ongaro, Capano, Howlett, Ramesh and Virani2019). It is generally a concatenation of mechanisms that enables the explanation of change processes (Gambetta, Reference Gambetta, Hedstrom and Swedberg1998: 105), as ‘[E]xplanations of most concrete social events or states require resort to several elementary mechanisms; one is not enough. Sometimes, these mechanisms counteract one another, and sometimes they work together’ (Hedstrom and Swedberg, Reference Hedstrom and Swedberg1998: 21). Social processes (McAdam et al., Reference McAdam, Tarrow and Tilly2001: 24) are in this perspective seen as concatenations of such mechanisms.
In particular, the analysis carried out in this paper uncovered the triggering of certain social mechanisms for explaining the process of systemic change observed in the Lombardy case of chronic disease management, and notably a set of social mechanisms able to explain how key stakeholders got engaged and sustained momentum of the change process. In particular, we hypothesize the triggering of the mechanisms of: actor certification, attribution of opportunity and threats, appropriation of mobilizing structures and band wagon effect, and we argue that the combination of these mechanisms may explain the unfolding of the process of systemic change that we observed, and that may be puzzling when factoring in the vagueness of the policy design.
The mechanism of actor certification refers to the validation of actors, of their performances and their claims by external authorities (McAdam et al., Reference McAdam, Tarrow and Tilly2001: 121). The mechanism of attribution of opportunity and threat can be defined as an activating mechanism responsible for the mobilization of previously inert social groups which involves (a) invention or importation and (b) diffusion of a shared definition concerning alterations in the likely consequences of possible actions undertaken by some ‘political’ actors (McAdam et al., Reference McAdam, Tarrow and Tilly2001: 43 and 95). The appropriation of mobilizing structures refers to social spaces put at the service of interpretations of situations and objectives that may be employed to mobilize actors towards certain courses of action (McAdam et al., Reference McAdam, Tarrow and Tilly2001: 102). Band wagon effect is the mechanism whereby it becomes more convenient to join (or at least be seen to join) a given social process (notably, the process of reform) rather than being side-lined and hence inhibited to enjoy the benefits if the outcome of the process turns out to be favourable.
In the public policy and public management literature, the most prominent applications of the analysis of social mechanisms have been focused on the ‘policy of public management change’, and related fields like ‘public administration innovation’ and ‘decentralization and devolution in the public sector’, which has been investigated in a series of studies (see the special issue edited by Barzelay and Gallego, Reference Barzelay and Gallego2010, and the related theoretical groundwork in Barzelay, Reference Barzelay2003 and Barzelay and Gallego, Reference Barzelay and Gallego2010: on the policy dynamics of innovation in public administration, Mele, Reference Mele2010; on decentralization, Asquer, Reference Asquer2012 and Ongaro, Reference Ongaro2006). In a related study, Barzelay and Campbell have investigated the dynamics of strategy change in a large public organization (namely the US Air Force and its envisioning process). What seems to be lacking in the extant literature in public policy and administration is the study of system-level change, and applications of the analysis of social mechanisms to the field of welfare administration, a gap this contribution aims to fill.
3. Research design and methods
The research design is a single longitudinal case study and the selected case is the systemic change occurred in the Region of Lombardy in Italy to implement the new chronic care model, universalist in thrust (the health care model is applied potentially to any patient in chronic condition resident in the region, and for any type of illness: an estimate of 3,350,000 target patients out of 10 million residents). The period of observation is from January 2016 to December 2017.
We emphasize the nature of theory building case of the selected case study, a usage in line with consolidated research, as outlined by case study methodologists such as Yin (Reference Yin2018); we should add our aim is more modest, and we would qualify this study as ‘model building’, or mid-range theorizing approach, whereby we aim to elaborate warranted claims about the dynamics whereby systemic-level change processes may unfold under conditions of policy indeterminacy, notably with reference to the welfare sector. Case selection has been driven by the nature of the case, notably the characteristic of the case providing an instance of system-wide change, combined with the opportunity (complementary criterion of ‘opportunistic considerations’) of the vast accessibility we were granted to the multiple sites where and when the change process unfolded. Our main unit of analysis has been the change process, and in this Eisenhardt (Reference Eisenhardt1989) has been a guiding scholar.
The advisory role was run by a network of three universities, including a team of nine researchers, who worked as an integrated research group. There was one coordinator per university team – one of them being a co-author of this paper, the other two having commented on earlier draft of this paper for validation and approval. Access to administrative data and to all the relevant stakeholders was enabled by the strategic advisory role granted to the research team. Collected evidence was always first discussed within the research team, then presented and analysed with different local stakeholder groups, and finally re-presented to the regional policymaking body, thus gaining a deeper understanding of different stakeholders' perspectives.
The official administrative data recorded for the innovation project we had access to included: minutes of all official stakeholders meetings (both at central and local levels, as describe more in detail in the narrative); the list of general physicians (GPs) and privately contracted providers agreeing to the new scheme and the related signed contracts (also the list of those opposed to the new scheme); the regional planning and regulatory documents, both in their initial draft versions and their various modifications from bills to the finally adopted acts.
The research team was involved across all phases of the policy process and was able to attend the negotiation rounds led by the Regional Government with the different stakeholders: public health care organizations, private contracted hospitals, nursing home providers, GP associations and unions. The discussions occurred both at the central regional level, mostly about general principles and the regulatory framework, and at the local level in the eight local territorial planning areas in which the regional health care system was partitioned. The combined team was present at central level discussions with the stakeholders, whereas each of the three universities worked as the local advisor in two or three out of eight local planning areas.
Data collection included a combination of: participant observation (as advisors to the regional government and trainers to local groupings); direct observations (at stakeholders’ meetings) and interviews. More in detail, during the observation period, there have been seven official meetings at the central level between the regional Government policymaker and the research team (in average they lasted 2 hours), to discuss regional plans and proposals or research reports before presenting them to stakeholders. The regional policymaking group was of mixed nature including both politicians and top managers. A draft paper was always presented at the beginning of the panel and a new version was provided a few days later. The researcher role was rather active as discussant of the policy proposals or ex post collected evidences. Over the same period, regional policymaker run 12 stakeholders meetings (at these meetings the research team was present in an observer role), some sectoral with a specific category of stakeholders, others across the board; meetings lasted on average 2.5–3 hours; stakeholders’ meetings occurred in four waves during the 2017–2018 period, at every fundamental policy change or detailing step.
Each university also run three training rounds in every assigned local planning area, with all the different stakeholders' technicians as participant. The first round was focused on the health care context analysis and the premises of the reform, the second to present the reform and the third to discuss organizational and operational consequences. Every round was divided into three panels (public organization, private ones and primary care actors) and lasted one and a half day.
Additionally, four research focus groups with all the top managers and project team members of the eight local health organizations in which the regional territory was split were set up, which lasted 3 hours each, centred on the analysis of the criticism and strength faced by public health organizations to implement the reform. The panels of organizations were divided into two different groups, each one with four public health care organizations. The first focus group was, in both cases, about organizational issues and the second one about operational problems to implement the new chronic care model. Each health care provider was represented on average by five people, the top professional of each project team, led by one of the top managers (director general or health care director or integrated care director). The role of one of the authors of this paper was to moderate the focus groups; minutes of the focus groups were taken, both for research purposes and to report to the regional government.
The investigators were further granted access to most of the relevant social and professional events in the larger stakeholders arena, including: a panel discussion of the association of private contracted health care providers (an event which proved crucial for the forging of a common position towards the reform by this important group of stakeholders); the paediatric union board meeting, where it was discussed and decided how to react jointly to the reform as a homogeneous professional group; an open conference on the topic held by the major opposition party, an event which proved important for the unfolding of the events; and two official regional conferences to launch the reform. Finally, frequent informal and individual interactions with a wide range of actors (with a special focus on top managers of five of the biggest private health care holdings) in the policy process occurred, either face to face or on the phone.
4. The new chronic care model in Lombardy
4.1 Background: the management of chronic diseases in the ageing population of Europe and Italy
In Europe, an average of 30–40% of the population suffers from chronic diseases due to increasing ageing, and almost half of them live under multi-morbidity conditions, covering 70% or more of health care expenditures (Wyke, Reference Wyke2011). Chronic diseases require a completely different approach and set of services compared to traditional hospital acute care. Patients need to be treated over many years and the best possible outcome is slowing down the deterioration process (least bad). Most of the treatments are assumed alone by patients, outside formalized care settings, and so patients' compliance becomes crucial. In this perspective, understanding who the key actor for patient compliance is becomes crucial: it may be the patient him/herself, his/her partner, a child, friends or neighbours or a privately paid caregivers. Fostering patients' compliance is a core task of the chronic care case management function.
Most of the chronic conditions may be detected in advance, correlating with each patient her/his set of chronic illnesses and stages (Bates et al., Reference Bates, Saria, Ohno-Machado, Shah and Escobar2014). A chronic condition requires intense interactions amongst different specialists and other staff because multi-morbidity is prevalent and clinical synthesis is more effective than the sum of single specialized treatments. Addressing simultaneously all challenges together (permanent treatments, patient autonomy and compliance, proactive medicine and interdisciplinary work) demands a radical change in the organizational and service model as well as in the professional skills and the associated technical, professional and social competences and attitudes. Many health care systems, both at local and national levels, have therefore undertaken large scale experimentations to adopt new complex institutional and organizational approaches, which are generally classified under the umbrella label of chronic care management or disease management (Nolte et al., Reference Nolte, Knai and Saltman2014). Systemic change brings with it formidable change management challenges, due to the scale of change (system wide), the shift in the underlying paradigmatic approach (from reactive to proactive medicine) as well as in the locus of the services (ambulatory care instead of acute care), the different emphasis in the skills required (the clinical synthesizer takes prevalence over the specialist), the rise of new roles (case managers) and the introduction of new performance metrics (clinical adherence and patient compliance instead of productivity).
4.2 Lombardy: geographical and economic context
Lombardy, located in the north of the country, is the wealthiest Italian region per GDP and by far the largest per population with over 10 million residents, with Milan as capital city and 3.1 million residents in its metropolitan area. A significant percentage amounting to 33.5% of the residents is in some form of chronic condition, and they have been classified into three clusters, the green ones (1.9 million) with one illness, the yellow ones (1.3 million) with two to three illness and the red ones (0.15 million) with more than four diseases (Lombardy Region: DGR 6164/2017).
4.3 The institutional and political context
The Italian National Health Service is a decentralized Beveridge model with strong governance power delegated to regional governments. Within budget constraints, regional governments have significant leeway in the design of the institutional arrangements whereby health care is managed: number and type of public health organizations (local health care authorities – LHA – and independent hospitals), financing models, market share of private contracted providers and organization and service models. The regional government deliberates for each public and private provider the number of beds for each speciality, the kind and size of big technological instrumentation to be purchased by the provider and the total maximum annual budget allocated to each provider. The health care budget in the period of observation totalled ca. 18 billion euros per fiscal year.
The regional health care system of Lombardy has a reputation for being one of the best in the country (CERGAS, 2017). Lombardy is the region which attracts more patient mobility from other Italian regions. Since the mid of the 1990s the Lombardy region is the one with the highest market share of private contracted health care providers (28% of the total public health care budget), most of which are large Italian for-profit health care groups. The system is based on a regulated competition between private contracted providers and public providers and patients are free to choose the provider they prefer with no restrictions.
4.4 Political and policy background: recent history before the reform
Since the establishment of a regional elective government, in 1970, Lombardy has been governed by a centrist or centre-right majority, with very few and short-lived exceptions (latest 1993–94). The dominant character in recent political history has been Roberto Formigoni, who performed as regional governor uninterruptedly for nearly 18 year from 1995 to 2013. He belonged to the political expression of a catholic movement, and governed in coalition with Berlusconi's party (named ‘Forza Italia’) and the Northern League (a previously secessionist party later reshaped into an anti-immigrant, sovereigntist and populist party). He strongly operated to increase the market share of private providers from 10% of the regional health care budget up to 30% (CERGAS, 2017), by introducing the ‘purchaser-provider split’ whereby LHA acquired the role of purchasers of health services on behalf of patients and transforming all public hospitals in autonomous service providers, financed on a fee-for-service basis and in competition with private contracted providers, both for acute care, home care and LTC, which gained, respectively, the 30% of the health market share, 95% of the provision and 80% of the nursing home beds (CERGAS, 2017): private providers represent a key group of stakeholders. The slogan of the regional health care model was ‘patient choice freedom’, meaning the possibility for patients of choosing the provider of health services. The model was embedded in a regional background with strong institutional capacity, a high level of medical competences and a good degree of social capital; even the pressure of the fiscal crisis, hitting hardly Italy and most Mediterranean countries in those years (for an overview of its perceived impact by public managers see Ongaro et al., Reference Ongaro, Ferré and Fattore2015; Longo, Reference Longo2016), was felt quite assuaged in the richest region of Italy.
The final stage of the Formigoni era was tainted by massive corruption scandals, with the Governor also personally hit, centred on the allegedly illegal benefits that people close to the government extracted from big private contractors. These scandals led to the forced resignation of Mr Formigoni and anticipated elections, which sets the stage for the episode of systemic change accounted for in this paper.
4.5 The emergence of a narrative for radical change and the role of policy indeterminacy
A new Governor, Roberto Maroni, was elected in 2013, from a different party (Northern League), but supported by the same political coalition. His governmental course of action was driven by the ambivalent imperative to both distinguish himself and mark his mandate as distinct from the Formigoni era, in order to leave behind the shadow of the previous scandals, but also guarantee a certain degree of political continuity, given a similar constellation of stakeholders was supporting his governing coalition. Given the political symbol par excellence of the Formigoni era was embodied by his health care model, the new governor undertook the design of a profound reform of the health care system, as a way of signalling discontinuity.
In formulating the argument for change, political discontinuity could only limitedly be evoked, because of the continuity of the supporting coalition, so the image of the policy issue was depicted in a scientific-technical way: he underpinned his argument on the epidemiological shift brought about by the ageing society and the consequent arising centrality of chronic disease, which demands for a strong development of primary care and community services, a deep shift in focus from the previous model, very much centred on acute care hospitals and nursing homes.
The paradigm shift was significant and it took 2 years of internal negotiation and harsh infighting within the coalition before the reform championed by Maroni was passed, in summer 2015, met by strong resistances from the near totality of private hospital providers, a major stakeholder of the governing coalition constituency. Other events imperilled the future of the reform: a few days after the reform bill had been enacted, both the regional minister for health care and Maroni's party health care expert and official advisor were arrested on charge of (another) corruption case, this time related to some public purchasing procedures and the outsourcing of public dentistry services. After a tense few months, a new regional minister for health care was appointed, but the situation remained troublesome: a new reform had been adopted on paper but not implemented, strong headwinds against it blew both from inside the governing coalition and from the contracted private health care providers; the inspirer of the new reform lay in jail, and the founding fathers of the previous model that had lasted nearly two decades were politically delegitimized due to the corruption scandal. The government was in a quandary: it was politically extremely challenging either to move forward with the new reform or to backtrack to the previous status quo, and the governor was already beyond mid-term, with only 24 months before the subsequent regional election campaign (due in March 2018). In these circumstances, there was a political drive for radical change, in order to distance the political course from both older and more recent scandals. On the contrary, the political ambiguities in the governing coalition hampered the consistency of the policy design. The combination of a need for radical change rhetoric, weak coalition and a short time span before the following elections led to a somewhat bold but very generic reform package, highly vague in its content, and additionally the same reform unfolded in four distinct acts, a regional law (law no. 23 of summer 2015) and three secondary legislation executive acts, the latter providing some details to the reform, the former representing but a broad framework to be filled. As a consequence, in our terminology, it may quite plainly be stated the reform was characterized by a high level of indeterminacy in policy design.
4.6 Outline of the design of the reform
According to the design of the reform of health care in Lombardy, community and hospital services were integrated into 27 comprehensive public health care organizations, and the purchasing organizations were merged into only eight large commissioning authorities, with an average catchment area of 1 million inhabitants (with a separate arrangement for the 3 million resident Milan area), assigning to it a role as local planner, regulator and controller of the system. The majority of the 105,000 civil servants employed in the regional health care system were relocated into the new 27 comprehensive health care organizations.
The new ‘mantra’ was about integration, holistic approach to patients, case management in order to tackle the chronic condition challenge. In a highly symbolic move, the regional ministry for social care was merged to the ministry of health in order to establish a new integrated ministry for welfare, with the ‘chronic patient’ – ultimately one every three residents – central in the rhetoric and the focus of political communication (the slogan was ‘from providing care to taking care’). Between January 2017 and August 2017, three significant planning documents were issued:
Regional Decree X/6164: Steering health care needs: implementing the chronic care model (30/1/2017).
Regional Decree X/6551: Providers network plan and case management for chronic patients (4/5/2017).
Regional Decree X/7038: Additional decisions to assess chronic care model providers (3/8/2017).
Jointly, they outlined the pillars of the new model. The first act was a systematic health population management exercise, clustering all patients in 65 different disease types, each one diversified along three levels of severity. The published data were aggregated and anonymous, whereas every public commissioning agency received the nominative list of its chronic patients, all classified into one of the 195 diseases clusters. For each of 195 different chronic conditions, a maximum number of treatments and health consumptions was set, correlated with a bundle-payment tariff.
The second planning document designed the governance architecture of the health care system, offering new potential roles both to public and private providers. Providers could put forward their candidature to become ‘case-managing organizations’ and ‘commissioning organizations’ for chronic patient. Provider could also apply to be service producer for chronic treatments, under case management organizations' commissioning. Providers could play simultaneously both parts in the comedy, for different patients. The new arenas were regulated by the local public planning and controlling agencies. A central part of the design lay in the strong push to GPs (which operated on an individual basis) to merge into cooperatives, which would take charge of case management and commissioning of chronic patients, with a pre-emption for all 1.9 million low level patients. This act paved the way to a major public tender (in June and July 2017), run by each of the eight planning and controlling agencies, offering to any public and private organization in the system to apply both for the case manager and the commissioner role, and the service provider role. The goal was to offer to every chronic patient a list of case managing and commissioning organizations to choose from, in order for patients to be supported by these organizations in the care of the diseases through all the spectrum of services: proactive medicine, annual treatment plan, administrative admission procedures run directly by the case manager, compliance control and support.
The third official act defined the required modalities and contractual forms and obligations that health care organizations had to adopt – reciprocally binding each other to the provider/purchaser role – in order to establish the integrated care networks able to take care of the entire gamut of the patient needs.
Patients would enjoy freedom of choice both with regards to their commissioning and with reference to the case management organization. Commissioning organizations select providers and sign binding agreements (legally they take the form of contracts); still, patients were free to opt out from the contracts, as they could more broadly opt out of the entire scheme. For the patient operating within the new chronic care model, all treatments would be reserved in back office by their case management organization, hence allegedly making access to services much easier and more comfortable.
The emphasis on patient free choice offered a good balance within the governing coalition to coalesce around a reform rhetoric combining continuity with the previous ‘Formigoni model’ (pivoted around the ‘patient choice’ slogan), on the one hand, with a major, potentially disruptive, innovation which could anyway be claimed to be made necessary by ‘objective’ epidemiological trends (tied to the ageing society), on the other hand.
5. Engaging stakeholders: how systemic change unfolded
A brand new game opened up for purveyors; whether each health care provider would have gained or lost in the new chronic care arena was an unknown, especially given the range of potential entrants in the new market of chronic care: any kind of health care organization could potentially put itself forward as case management and commissioning player, and also as service provider within the system. Public and private organizations were placed on a level playing field, with equal market access opportunities: public providers, public teaching or research hospitals, cooperatives of GPs, private contracted hospitals or ambulatory providers, and nursing homes could all become case management and commissioning organizations.
The players of the new game were in the dark about what share of the market they would obtain in the end, and no robust forecasts were available given the radically different scenario where they had to operate. Citizens under chronic conditions received in January 2018 a letter from the local planning authority indicating a list of case management and commission organizations within the catchment area of the patient, and patients were free to step in and choose the case managing organization of their like, or to opt out of the model and maintain the previous arrangement. Even when a case management organization signed contracts for their patients, still patients retained the freedom to opt for any other provider of their choice. This high level of indeterminacy in the reconfiguration of the health care provision market and sheer uncertainty for each individual provider triggered a process whereby actors were incentivized to opt into the system, in order to secure the chance to become a purveyor to the system at a later stage, should the market for chronic diseases, intermediated according to the new governance model, grow in the future. In a way, any player could project and attach to the new scheme either a major opportunity or a daunting threat. Uncertainty was not only related to the market share of each provider and between public and private organizations, but also between different health care settings. The competition between primary care professionals, acute care specialist and the large nursing home world (60,000 beds in Lombardy) was opened up, especially for the case managing and commissioning role, bringing more attention to the reform and participation by actors. In short, a combination of design feature (the very indeterminacy of the policy design) and market-type competition triggered by the funding regime (which basically set out one, comprehensive envelope of money for all the actors in the system) can be purported to have triggered the social mechanisms of attribution of opportunity and attribution of threat to the new circumstances by the stakeholders, leading to the entirety of the providers – public and private – to ultimately subscribe to the scheme.
Another event enabled to attract to the scheme two other sets of actors. Three major universities were jointly contracted as regional government advisors for securing the implementation phase and for showcasing the innovation at the local level. Each of the three universities had a solid reputation in health care management. They were also picked up as gatekeepers towards different categories of stakeholders. The Catholic University of Milan, placed mainly in Milan and Rome, has developed a robust tradition in health care management, both in teaching and research. Bocconi University is a major provider of both executive education and research in the health care sector and the hub of important networks of public and private health care top managers. The Polytechnic of Milan has a strong expertise in health care management and is very well established in the field of health care operations and ICT management. The universities advised jointly the regional government, shuttling local stakeholders' worries from the territory to the heads of the regional ministry of health in Milan. They were in charge of communicating the reform locally. Each university addressed territorially different clusters of both public and private stakeholders, in each of the eight LHAs at least 10 meetings were held with different categories of stakeholders: GPs, private contractors, public hospitals, associations and local governments. The three universities became key interlocutors in informing and ‘explaining’ to professionals the new governance model envisaged by the regional government: besides private acute hospitals and private ambulatory companies, also GPs, paediatrics, nursing homes were involved in systematic discussions of the proposed reform by the three teams of the universities – who acted in a very coordinated way due to previous personal acquaintances amongst themselves. Every group held discussions within peers and then with public policy makers, both at local and central levels, building a cohesive web. The action of the three universities, by providing their reputation and legitimacy to the process and operating so diffusely on the territory addressing all the categories of stakeholders mentioned above, triggered the mechanism of actor certification: attesting the soundness and providing legitimacy to the new system of governance arrangement for chronic care management.
Beyond the involvement of the universities, opportunities for communications and listening were actively sought through a communication campaign targeted to various categories of stakeholders and at different territorial levels, through meeting and panels with the managers of all public organizations, the association of private health providers, the unions, and patients' associations at the central level. Each of the eight planning authorities had formal meetings with local public and private providers, with GPs, with the unions, with patients' associations and with local governments' representatives. Every Tuesday there was a meeting between all directors of the planning units and the director of the regional ministry of health and his staff. This communication campaign – a provision deployed along the course of the implementation process by the regional authorities – further reinforced acceptance of the new governance model by a range of stakeholders (it became a design feature set up during the implementation phase that acted as a reinforcing factor in the change process).
Other stakeholders proved much more resistant to being swayed, though. The powerful private hospitals association was initially very sceptical, fearing that public health care budget would shift from acute, where it had a strong presence, to primary care, where they had initially poor market opportunities. Also, the influential medical society of Milan started a robust media campaign against the new chronic management model, which lasted several months. The main argument was that case management models, like the one proposed by the regional government, ultimately lead to economic interest to drive health care processes, to the detriment of the ultimate wellbeing of the patients. This strong intervention by the medical society of Milan resulted in a much lower percentage of GPs adhering to the new scheme in the Milan area. However, the percentage of adhesion was much higher in the rest of the Lombardy region, tallying an average of 48%, which may be deemed to be quite significant considering the average age of GPs in the region is 55 years, an age at which somebody is quite reluctant to change her/his way of working (in the hindsight, the weakness inherent in the system of territorial care, with too few GPs in charge of too many residents, operating in isolation and often approaching retirement age, proved to be detrimental to the capacity of the Lombardy health system to tackle the major shock of the COVID-19 crisis which broke out in winter/spring 2020 – though this is analytically unrelated to the process under observation in this contribution).
However, other key actors were fully won to the cause of adhering to the new scheme. Although in May 2017, a board meeting of the paediatric union officialized opting out from the scheme, because of the wide uncertainty attached to it, the union then re-opened the option of opting in at a later stage. Notably, paediatrics were not used to work in cooperatives, and associating into cooperatives was a requirement for GPs (paediatrics are GPs in the Italian health care system, and each person aged 14 or less has to register with a paediatric as the GP of reference). They turned to the universities for advice, from which they drew two suggestions that affected their final choice: first, that the model is so open-ended and unpredictable in its impacts that it may powerfully be shaped by those who will first join it and the course of action they will concretely undertake; the balance of power between the professional groups in the health profession will also be widely affected by first movers. Second, the lack of experience with working in cooperatives is not an insurmountable hurdle, as a professional group can be hosted within the cooperative arrangements already set up by other GPs (so a paediatric could join of cooperative of GPs rather than have to set up her/his own with other paediatrics). Eventually, the consistency of the argument and the authoritativeness of the source of advice (three high reputed universities) won the argument, and the association of paediatrics totally reversed its initial opposition and opted to join the scheme. The experts from the three prominent universities provided legitimacy to the new scheme in the eyes of these professionals. The fact universities were interposed between the politicians (much less legitimate in the eyes of professionals, notably in the then scandal rife regional government) and professionals acted as means of decoupling mechanism empowering the legitimacy of the scheme in the eyes of this key stakeholder. The social mechanism of actor certification appears to have been at work providing a positive spin to the message beyond the low reputation of the initial sender of the message itself (the scandal-hit regional government). Also, another social mechanism appears to have been in operation: the mechanism of appropriation of mobilizing structure, whereby the universities acted as platform for mobilizing professional groups, such as paediatrics, to join the reform (making them shift from the initial oppositional stance).
Another key actor was also won to the cause: private providers of acute services. In May 2017, the major opposition party organized a political conference on the reform. Rather than lashing out at the reform, the leading leftist party seemed to stress the positive side of the reform, notably its being focused on chronic diseases and the development of primary and ambulatory care service, a component of health services historically weak in Lombardy and a traditional battle horse of the left. However, party exponents did lash out to the overly constrained role for public providers in the new designed scheme. The takeaway lesson for private providers was that they did need to enter this new ‘market’ in the time window before the subsequent elections (fast approaching in less than 2 years), otherwise in case of reversal of the political fortunes, the space for private providers in the new model would have shrunk (dramatically, in the viewpoint of this category of stakeholders), whereas the new governance model of chronic disease management was anyway destined to last, as it met also the favour of the main opposition party which was the only one realistic potential alternative party in government. The new model came to be framed by this set of stakeholders as both an opportunity and a threat: an opportunity for the expectation it would have lasted; a threat because under the scenario of the opposition party gaining power the space for private providers in the market would have dwindled: a void had to be filled promptly. Ultimately, all private providers of health services joined the reform deeming it more convenient to join the process, rather than being side-lined and hence inhibited to enjoy the prospective benefits if the outcome of the process turned out to be favourable: the bandwagon effect may explain the building up of sustained momentum for the new governance system to develop and institutionalize.
6. Analysis and discussion
The combination of design features – most notably the very indeterminacy of policy design, our analytical focus, but also deliberate design provisions like the market-type competition triggered by the funding regime – and context factors (ranging from the reputation of local universities in the eyes of the clinical operators to the political stability of the regional government, whose political majority had remained unaltered over the two previous decades and most stakeholders expected it to remain the same at the subsequent elections, as it happened) led to the hypothesized activation of a number of social mechanisms, in certain concatenations, that has enabled, in the case observed, a systemic change in the governance of chronic care management, and, complementarily, defused alternative courses of action that might have led to stall or dampen the reform.
It should be noticed that most of the triggers of the observed social mechanisms appear to have been non-deliberate, at least not deliberated in the design of the policy: indeed, we emphasize the indeterminacy of the policy design as one of the key enablers of the change process. More specifically, alongside the indeterminacy of the expected governance arrangements, it is the uncertainty about the expected impact – in terms of on the role and market share each provider of health services could expect to win out of the new governance arrangements for the management of chronic disease – which acted as enabler, due to it being accompanied by the triggering of other social mechanisms enabled by the multi-level communication and negotiation process and notably by the interposition of the three universities between policy-makers and the key stakeholders.
In itself, indeterminacy of policy design could have well triggered attribution of threat to the new circumstances, interpreted as a menace, hence stalling the change process. Instead, indeterminacy led to attribution of opportunity to be attached to the new circumstances. Central to the outcome was the multilevel communication and negotiation process towards stakeholders, run in a diversified and bespoke way, whereby some communication and negotiation channels operated on a collective basis as a function of the professional community involved (GPs and paediatrics) and some others on a local, decentralized basic, province by province. This way, there was enough room for some ‘adaptive incoherence’ in the messages offered by the different level of government to the diverse stakeholders. In addition, crucial to the effective touting of the reform was the partial ‘contracting out’ of stakeholder management to three major regional universities. The universities offered a sort of open discussion space, where the prestige of the researchers and the formal absences of the official policymaker offered a setting for direct observation and comments, which supported a better understanding of stakeholders' expectations. Universities acted as mobilizing structures. It worked also as a robust sense-making process able to explain the vision of the reform, without the need to immediately come to political negotiations.
The mechanisms that we conjectured to be at work in a combined way include: the mechanism of actor certification by interposing authoritative institutions between (the partly discredited in the eyes of stakeholders) policy-makers and the key stakeholders for the change process to gain traction; the appropriation of mobilizing structure mechanism for decoupling and recoupling of communication and negotiation channels, by means of the multiple channels of communication occurring at both local and central levels, in such a way that attribution of opportunity by key groups of stakeholders shaped the dominant stance – defusing the initial stance characterized by the potential attribution of threat to the new circumstances opened up by the reform and its indeterminacy; and finally, the band-wagon effect mechanism based on rational imitation, whereby successive actors opted for joining the chronic care management scheme and opted in, sustained the change process and enabled to build momentum to the change process and reinforce a self-sustaining dynamic.
Ultimately, the combined activation of these mechanisms performed as enabler of substantial systemic change, as measured both in terms of the number of players which signed into the model (the entirety of private providers as well as of the professional groups), and the overall shift in focus of the health care agenda from acute to chronic disease management. This specific concatenation of social mechanisms may provide an explanation about why system-level change occurred in the way recounted here.
7. Conclusions
We propose the approach of the analysis of social mechanisms as a conceptual tool to explain systemic change in public governance and public services management, and we undertake to develop such approach by applying it to the case of systemic change in chronic disease management in the Italian region of Lombardy. Specifically, our analysis sheds light on system-level change under conditions of policy indeterminacy, and our core argument is that certain concatenations of social mechanisms, like the ones highlighted in the case study of chronic disease governance in the Italian region of Lombardy, may enable to explain systemic change, and why it occurred under otherwise unfavourable conditions.
In particular, we claim that policy indeterminacy may be an enabler of systemic change – contrarily to what might be expected – and we provide an analytical explanation for how this unexpected course of events may occur. Hence, the analysis of social mechanisms provides a valuable repertoire of conceptual tools for explaining the occurrence of courses of action which would have been highly unlikely to happen otherwise: notably, major, systemic change is generally assumed to be hard to put into effect, and absence of policy clarity is generally seen as recipe for stalled or aborted, rather than implemented, change to occur. In terms of the generalizability of the findings, we would consider most public policies to be generally characterized by ‘determinacy’ of the policy design, and even more so in sectors where technical knowledge and expertise is a key factor as is the case of the health care sector. And yet there are instances in which the level of indeterminacy is so relevant for the policy design to be considered as indeterminate: this condition may open us a set of process dynamics which differ from the (more common) circumstances under which this is not the case. This study aims to shed light on what happens under these circumstances, and its findings might enable a better understanding of how system-level change in welfare systems may unfold, under conditions similar to the ones we have outlined.
More broadly, we argue that the analysis of social mechanisms may be a powerful approach for generating knowledge about effective change management in public governance and public management. The analysis of social mechanisms is an approach which may usefully complement other approaches to change management in this field. In this perspective, we join the call by scholars in the field of public governance, policy and management (Barzelay, Reference Barzelay2003; Barzelay and Gallego, Reference Barzelay and Gallego2006, Reference Barzelay and Gallego2010; Ongaro, Reference Ongaro2006, Reference Ongaro2013a, Reference Ongaro, Capano, Howlett, Ramesh and Virani2019; Mele, Reference Mele2010; Asquer, Reference Asquer2012; Mele and Ongaro, Reference Mele and Ongaro2014; Capano et al., Reference Capano, Howlett, Ramesh and Virani2019) for the systematic application of the analysis of social mechanisms to the study of institutional, policy and administrative change.
Acknowledgements
The authors are grateful to Eugenio Anessi Pessina and Cristina Masella for their valuable comments on an earlier version of the article, as well as to the anonymous reviewers for insightful comments on the submitted version.
