The Translation of Hospital Management Models in European Health Systems: A Framework for Comparison

The Translation of Hospital Management Models in European Health Systems: A Framework for Comparison University of Leeds, UK,1University of Innsbruck, Universitätsstraße 15, 6020 Innsbru

The Translation of Hospital Management Models in European Health Systems:

A Framework for Comparison

University of Leeds, UK,1University of Innsbruck, Universitätsstraße 15, 6020 Innsbruck, Austria,

2Bocconi University, Italy, and3Staffordshire University, UK Corresponding author email: bernadette.bullinger@uibk.ac.at

In this paper we develop a framework for comparing changes in the management of

public hospitals across different national health systems, drawing on insights from

institutional theory Using a range of secondary sources we show how one particular

form of hospital management, pioneered originally at the Johns Hopkins Hospital in

Baltimore, has been translated differently in four health systems: England, Denmark,

Italy and France This analysis builds on the notion of editing rules, which derive from

the institutional context, and illustrates how these rules broaden our understanding of

variable translations of global templates for hospital management The paper concludes

by highlighting wider implications for theory and policy.

In health systems around the world there has been

a common focus on strengthening the

manage-ment capabilities of hospitals following the model

of private corporations (McKee and Healy, 2002)

However, while there are strong indications that

healthcare management has become an

interna-tional trend there are risks of overstating

conver-gence Existing comparative research highlights

similar priorities that are driving reforms, but also

‘distinctive national or regional variants’ (Dent,

2006, p 624) A handful of studies, for example,

have noted differences in the implementation of

diagnostic related groups (DRGs), clinical

gov-ernance regimes (Burau and Vrangbæk, 2008) and

in the responses of clinical professionals to

budgets and leadership education (Jacobs, 2005;

Kurunmäki, 2004)

However, while this work suggests ‘alternative

change pathways’ in health reform (Jacobs, 2005,

p 158), with some exceptions (Dorgan et al.,

2010; Eeckloo, Delesie and Vleugels, 2007), less

attention has been paid to how this might apply

to the management of hospitals Although there

are strong indications that private sector

man-agement ideas and templates with a global pro-file have been interpreted differently, we know very little about the details of this process There are also deficiencies in our understanding of

why variations might occur between health

sys-tems and the factors that influence this process Much of the available comparative research has drawn loosely on notions of path dependence (Burau and Vrangbæk, 2008; Dent, 2003;

Kirkpatrick et al., 2009), which, although useful,

provide only a general starting point for drawing attention to different national outcomes of health management reforms

In this paper we address the question of how similar management ideas and models have been implemented differently across health systems and how one might explain varying outcomes To do so

we draw on recent advances in institutional theory and in particular the notion of ‘translation’ (Boxenbaum, 2006; Morris and Lancaster, 2006), which shows how actors engage in modifying tem-plates such as universal models of management Specifically we use ideas from Scandinavian insti-tutionalism (Boxenbaum and Pedersen, 2009) and

British Journal of Management, Vol 24, S48–S61 (2013)

DOI: 10.1111/1467-8551.12030

© 2013 The Author(s)

British Journal of Management © 2013 British Academy of Management Published by John Wiley & Sons Ltd,

9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA, 02148, USA.

comparative literature on new public management

(NPM) reforms to develop the concept of editing

rules for cross-national comparisons Actors

involved in translations implicitly follow editing

rules, which, we argue, are derived from the

insti-tutional context (Sahlin-Andersson, 1996)

Applying these ideas, we focus on the

transla-tion of a particular model of organizatransla-tion,

pio-neered in the USA in the 1970s at the Johns

Hopkins Hospital (JHH) in Baltimore but which

later served as a template for how any hospital,

including the public or non-profit sectors, might

enhance their performance Specifically, it

empha-sized the need to strengthen the corporate

govern-ance of hospitals and sub-divide them into

business units (or ‘clinical directorates’) to

maxi-mize efficiency Focusing on this particular

tem-plate in the context of four health systems – the

English National Health Service (NHS),

Denmark, France and Italy− we pursue two main

objectives First, we explore the translation

process of the JHH model and whether this

resulted in different interpretations and practices

across national systems Second, using the

concept of editing rules, we explore how

differ-ences in the wider institutional and regulative

context might help to explain variations in the

translation of the JHH model A key contribution

of the paper is to advance understandings of

com-parative hospital management reforms and also,

drawing on concepts from institutional theory,

develop our knowledge of how these processes

might be theorized and explained

Translation as a model for

disseminating institutional templates

In many ways the notion of translation represents

a departure from institutional theory’s early focus

on isomorphism and conformity in organizational

fields More emphasis is given to the way

tem-plates such as lean management or diversity

man-agement – often available on a global scale – are

legitimated and enacted in local settings A

number of scholars have highlighted the

modifi-cations which actors introduce to make

institu-tional templates ‘fit’ in a local context Taking

into account the localized origin of templates, the

‘travel of ideas’ concept, for example, illustrates

how they can be translated into global ideas This

implies dis-embedding templates from their local

context in order to travel to other institutional settings, where re-embedding efforts are necessary

to translate the global idea into practices (Czarniawska and Joerges, 1996) These multiple translations depend on institutional actors like organization members, but also policy makers and professional bodies, who are no longer per-ceived as passive adopters but as actively modify-ing ideas as well as bemodify-ing modified by them This idea of translation helps to explain ‘how appar-ently isomorphic organizational forms become heterogeneous when implemented in practice in different organizational contexts’ (Boxenbaum and Pedersen, 2009, p 191) They are transformed both in verbal accounts and actual practices (Boxenbaum, 2006; Morris and Lancaster, 2006) While there is broad agreement within and between Scandinavian and North American institutionalism that actors modify spreading institutional templates, there is some disagree-ment concerning the degree of agency in this process As Boxenbaum and Pedersen (2009) suggest, the ‘strategizing’ approach places most emphasis on the strategic intentions of actors in the translation to promote their own interests By contrast, the ‘embeddedness’ approach focuses more on implicit and pragmatic dimensions of actors’ translations, which are unconscious efforts to make sense of and adopt templates in local contexts

Following the embeddedness approach, Sahlin-Andersson (1996) stresses the importance

of the institutional context for translation out-comes She found that actors do not arbitrarily modify or ‘edit’ practices Rather, their transla-tions are governed by non-formalized ‘editing rules’ which influence this process and may even be taken for granted by the actors themselves Thus, the outcomes of translation are not arbitrary structions but are linked to the way ‘different con-texts provide different editing rules’ (Sahlin and Wedlin, 2008, p 226) Implied here is that local history, traditions and institutions form the back-ground for how actors in a given setting engage with new templates More specifically, editing rules which arise from the local context enable and restrict how actors modify templates, how they translate them and make them fit However, it is important to note that editing rules cannot be conceptualized as prescriptive ‘rules to follow’ but rather they are implicit ‘rules which have been followed’ (Sahlin-Andersson, 1996, p 85)

To illustrate this idea Sahlin-Andersson (1996)

talks about editing rules concerning logic which

refer to reasons given for the introduction of a

template Stories about successful

implementa-tion of new practices such as, for instance,

corpo-rate codes of ethics in multinationals (Helin and

Sandström, 2010) often have a rationalistic plot

or logic, which presents the template as solving

clearly defined problems in line with models of

rational decision making Editing rules might also

concern the formulation, labelling or packaging

of new organizational templates, providing

attention-attracting rationales and moral

justifi-cations for change (Sahlin-Andersson, 1996) An

example would be rules actors implicitly followed

to promote lean management in the UK building

industry (Morris and Lancaster, 2006) These

practices were dramatized as morally superior to

the ‘outmoded’ wasteful forms of management

that existed previously and thus lean management

was labelled a modern management technique In

addition to this Sahlin-Andersson (1996, p 86)

refers to editing rules concerning the context and

claims that ‘the history of the local setting may

restrict the translating’ Here the focus is not just

on how new templates are framed and articulated

in rhetorical terms, but also on how new practices

get implemented that shape the strategies and

behaviour of actors as they engage with these

changes

An embeddedness approach therefore requires

a detailed understanding of the national context

in which translation occurs However, to date

this insight has not been fully developed in the

translation literature Most studies have only

looked at two institutional contexts: the context

of origin of a template (often the USA) and the

context of ‘destination’ Boxenbaum (2006), for

instance, studied the translation of diversity

man-agement from the USA for the Danish context

Morris and Lancester (2006) studied originally

Japanese lean management ideas being translated

for the UK construction industry, while Helin

and Sandström (2010) explored the travel of a

corporate code of ethics from the US parent to

its Swedish subsidiary While providing rich

descriptions of the inquired cases on

organiza-tional or field level, these studies have tended

to understate editing rules stemming from

the specific national context and how these

shape a varying potential for agency on different

levels

Given these limitations a fruitful line of enquiry for understanding the role played by national con-texts are other branches of institutional theory, such as the varieties of capitalism and business systems literature (Tempel and Walgenbach, 2007), which have tended to place more emphasis

on exploring the national institutional conditions that shape the reception of ideas With regard to our own specific focus on hospital management, ideas from the comparative literature on NPM reforms are also helpful (Dent, 2003; Hood, 1995) Pollitt and Boukaert (2011), for example, note that while socioeconomic forces and political pressures lie behind the spread of management reforms globally, crucially important at the national level are the perceptions of elite decision-makers both

of what is desirable and what is feasible The latter relates to what is considered possible given available resources, existing structures and likely obstacles such as ‘conservative forces which resist change’ (Pollitt and Boukaert, 2011, p 25) Perceptions of what is desirable are influenced

by political ideologies and cultural perceptions of the kinds of reform that are important and valu-able From these perceptions emerge editing rules that are more or less prescriptive in guiding how actors adopt global templates in each national context

Hence, to understand these particular national idiosyncrasies and their influence on the transla-tion process, we argue that editing rules need to

be extended from a merely symbolic and linguis-tic level of analysis to the level of structural im-plications and material practices As such our approach is not to focus on editing rules concern-ing logic and formulation, which refer to the symbolic level of editing stories, but rather to emphasize editing rules concerning the context (Sahlin-Andersson, 1996) Drawing on Pollitt and Boukaert’s cross-national focus, we specify such rules by looking for national differences in the perception (by elite actors) of the desirability and feasibility of a particular model of hospital management

Methods

To address the question of how global templates

of hospital management were translated we focused on the experiences of reform in four European health systems: the English NHS,

Denmark, Italy and France This comparison is

both meaningful and theoretically interesting On

the one hand all four represent health systems that

are heavily state regulated, with central

govern-ments being key actors in the top-down initiation

of management reforms In this respect, they

differ from more decentralized health systems

such as Germany where the diffusion of new

man-agement ideas has been less centrally directed On

the other hand our sample of cases is also

illustra-tive of different contexts that might shape the

process and outcomes of reform Hence, while the

UK and Denmark are unambiguously national

health systems with hospitals largely owned and

managed by the state, Italy and France both

operate more hybrid funding and provision

regimes, with a large proportion of hospital care

located in the private sector (see Table 1 for

details) These cases also illustrate the variable

timing of reforms, with France being a relatively

late starter

To conduct this analysis we drew on a range of

secondary data sources from the healthcare

man-agement, policy and sociology literatures A key

source was work already conducted by ourselves,

both on country-specific developments in health

management (Lega, 2008) and comparatively

(Dent, 2003; Kirkpatrick et al., 2009) A

system-atic literature review was also conducted in two

stages First, we used published summaries of

hos-pital management reforms to construct a general

narrative for each country Second, we carried out

a more focused search of the available academic

research – mainly published in English− relating

directly to changes in hospital management This

initially focused on peer reviewed journal articles,

manually reviewing titles for relevance to the

topic of hospital management Following a

snow-ball approach, the review was then extended

to include book chapters and reports Lastly, we drew on available comparative research on

hospital management (e.g Dorgan et al., 2010)

and information published by transnational agencies such as the Organization for Economic Cooperation and Development (OECD) and Eurostat

Our analysis of these data involved two main stages First we sought to identify differences in the translation of the JHH model across our four case study health systems As we shall see, varia-tion occurred along two key dimensions: the nature of the authority structure of hospitals and development of non-clinical management func-tions Second, we analysed the data on the actual process of reform, noting how change was influ-enced by perceptions of elite actors of desirability and feasibility and inputting from this different editing rules The results of this analysis are pre-sented below, although prior to that it is impor-tant to describe briefly the nature and emergence

of the JHH model itself

Hospital management reforms: a case study of translation

The origin of an institutional template

As noted earlier, the drive to reform the manage-ment of healthcare has been present in many developed economies since the mid-1970s Hospi-tals in particular became a target for these reforms given the high proportion of resources they absorbed and the apparent difficulty of coordinat-ing different priorities of care, cure and adminis-tration (Glouberman and Mintzberg, 2001) In this context an alternative model emerged for how

Table 1 Hospital sector characteristics

Proportion of health

expenditure

accounted for by

hospital sector (%)

Percentage of total acute care beds in publicly owned hospitals (%)

Hospital payment system (public)

Autonomy of hospitals to recruit medical and other health professionals

Autonomy of hospitals to decide the remuneration of other health professionals

UK 40 96 Per case/DRG (70%) and

global budget (30%)

Denmark 45 97.7 Global budget (80%) and per

case/DRG (20%)

Sources: Healthcare expenditure, 2008; Paris, Devaux and Wei, 2010.

hospitals might be run to maximize efficiency,

originating from the JHH, a teaching hospital in

Baltimore in 1972 (Chantler, 1984; Heyssel,

Gainter and Kues, 1984)

In its original form this model moves away

from the practice of governing hospitals through

parallel hierarchies, with doctors represented by a

senior medical committee, sometimes with powers

to veto management decisions Instead, the focus

is on all clinical staff (doctors and nurses)

report-ing through a sreport-ingle, unitary chain of command to

a clinical director who in turn is accountable to

the chief executive or general manager of a

hospi-tal Closely related to this were changes to the

governance arrangements of hospitals, moving

away from an exclusively trusteeship logic,

with boards focusing on conformance and

exter-nal accountability to a management logic with a

stronger performance orientation (Eecklo,

Delesie and Vleugels, 2007)

In organizational terms, the JHH model

involved a break from the traditional functional

structure (with medicine, nursing and other

func-tions organized separately) with the hospital

rep-resenting a kind of ‘holding company’ of

semi-autonomous divisions (based around product/

service lines or clusters of activities) (Braithwaite

and Westbrook, 2004, p 142) At the middle tier

this meant grouping resources, with specialties

and doctors aggregated in clinical units, each

managed by a team (or triumvirate) headed by a

medical chief, supported by a lead nurse and

administrator Each group (or, later, directorate)

was given responsibility for budgets and made

accountable for direct costs, and the operational

performance of their units, delivery against

targets and human resource management

Proponents of this change in hospital

organiza-tion have highlighted a number of advantages At

JHH an explicit goal was to mimic practices in the

corporate sector to drive down the costs of

inpa-tient care (Heyssel, Gainter and Kues, 1984) By

merging clinical specialities into larger

directo-rates and sharing other costs associated with

administration, nursing and ancillary staff, the

model offered potential for economies of scale

and scope, as well as better integrated services

These changes also represented a way of

stream-lining lines of accountability and strengthening

the authority of managers to make decisions

Lastly, an added advantage might be to co-opt

clinicians (especially doctors) themselves more

fully into the ‘world’ of management (Eecklo, Delesie and Vleugels, 2007)

This model of hospital management, or at least various translations of it, has subsequently been adopted in health systems around the world According to Braithwaite and Westbrook (2004,

p 142): ‘The clinical director (CD) concept dis-persed relatively rapidly, in ways that innovation diffusion theorists would find predictable of an attractive idea’, such that ‘every large hospital now has some form of CD structure as a key component of its governance arrangements’ This process began in the USA and Canada (Fitzgerald and Dufour, 1998) but quickly spread to Aus-tralia and Europe (Neogy and Kirkpatrick, 2009) This rapid dissemination of the JHH model was aided partly by the existing strength of interna-tional professional networks in the health sector

In England, for example, the model was champi-oned by Professor (later Sir) Cyril Chantler of the United Medical and Dental Schools of Guy’s and

St Thomas’ Hospitals, who had previously been a visiting Professor at Johns Hopkins (Chantler, 2012) Also important was the status of JHH itself, one of the elite university hospitals in the USA, and the publicity which leading clinicians gained by publicizing their experiences in the

highly prestigious and widely read New England

Journal of Medicine.

Translation process and outcomes in four health systems

Focusing on our exemplar countries it can be seen that versions of the JHH model have been imple-mented in public hospitals in all four cases The timing of this process varied between countries In the English NHS, a version of the JHH model was introduced following legislation in 1991 (NHS Community Care Act), which also led to the establishment of semi-autonomous foundation trusts with corporate style boards (Harrison and Pollitt, 1994) In Denmark, major changes to hos-pital governance were first introduced following legislation in 1984, with a second hospital com-mission promoting the model of clinical directo-rates based on ‘unambiguous management’ in

1997 (Kragh Jespersen and Wrede, 2009) In Italy regulatory pressures were also important Here a key piece of legislation in 1992 allowed some public hospitals the opportunity to convert to

semi-independent enterprises (Aziende

Ospe-daliere) with a chief executive officer (CEO) and

board structure and actively promoted clinical

directorates, although these only became

manda-tory in 1999 (Lega, 2008) In France, although the

idea of strengthening management in hospitals

had been attempted in 1983, it was not until 2002

that a sustained push in this direction began with

the introduction of the ‘Hospital 2007’ plan An

ordinance of 2005 established a new governance

structure for hospitals, establishing management

boards Hospitals were also encouraged to

rear-range clinical units into larger ‘activity centres’ (or

Poles) with delegated budgets, very similar (on

paper at least) to the English model (Dent, 2003;

Or and de Pourourville, 2006)

Hence, while there have been differences in the

timing of reforms, versions of the JHH model

have been adopted in the public hospital systems

of all four countries As indicated in Table 1, this

has also been associated with some move towards

variable funding per case (based on DRGs), away

from global budgets for hospitals It has also

manifested itself in new hospital governance

arrangements − formalizing the role of chief

executive officers or equivalent − and the

estab-lishment of a middle tier of management around

departments (or directorates) with devolved

responsibilities (see Table 2) However, the

avail-able evidence suggests that the degree of

conver-gence should not be exaggerated and the JHH

model has been translated by actors operating

both at the national (policy) level of each country

and locally, within hospital organizations

This is most obviously the case if we look at the

degree to which reforms in each country have

been formally implemented In Italy, for example,

Lega (2008, p 255) reports that, even in 2004,

only 66% of hospitals had fully adopted clinical

directorates Perhaps unsurprisingly, studies also

point to wide variations in the size of clinical

directorates (measured by staff, beds or turnover)

and in the level of authority clinical managers

might exercise over budgets (Bellanger, 2007;

Cantù and Lega, 2002; Kragh Jespersen and

Wrede, 2009) Clinical directorates are even

con-figured in different ways according to their size or

differing logics with respect to the size of hospitals

(Braithwaite and Westbrook, 2004)

In addition to this are variations in translation

outcomes between countries, most obviously with

regard to nomenclature In the English NHS a

corporate language of CDs, boards and CEOs has

been adopted quite explicitly By contrast, in other countries the terms used to describe new roles suggest much greater continuity with profes-sional norms and pre-existing models of hospital organization Hence, ‘clinical directors’ are for-mally heads of centres in Denmark, of activity

centres (or Chefs de Pole) in France and, in Italy,

chairs of departments (Cantù and Lega, 2002) Table 2 outlines further differences in transla-tion looking at four key areas: strategic (govern-ance), middle management, the nature of authority structures and development of non-clinical management roles Concerning the former

in England, the decision was made to establish a unitary governance arrangement both at board and clinical directorate levels (Harrison and Pollitt, 1994; Shortland and Gatrell, 2005) A similar situation applied to Denmark, especially

after 1997 (Kirkpatrick et al., 2009) By contrast,

in France and Italy, notwithstanding the rhetoric

of reform, the management authority of CEOs (or equivalent) is far less clear-cut In both cases the translation of the JHH model has resulted in tri-partite decision-making structures involving both external and internal (notably medical) stakehold-ers In France, for example, while supervisory and medical committees are formally consultative, they have powers to nominate Chefs de Pole and

in many cases are locked in to relationships of

‘collective bargaining’ with the director (Laouer, 2011; Vinot, 2012) As Bellanger (2007) notes, while ‘both General Director and Hospital Direc-tors do manage physicians’, this is primarily ‘by influence’ such that ‘the decisional process is gen-erally based on consensus’

Similar differences are apparent when one turns

to the changes at middle management (clinical directorate) level While in England and Denmark the focus has been on developing clinical directo-rates based on a single line of management accountability, this has been less obvious in France and Italy In the latter, the decision was made to establish executive committees within each department (or directorate) made up of the chiefs of clinical units with powers to nominate the chair and veto key decisions (Tousijn and Giorgino, 2009)

These differences in the way hospitals have reorganized management also have implications for the nature of authority While in England establishing a unitary chain of command in hos-pitals was a central plank of the reforms

organization England

authority structure

non-clinical management

(Harrison and Pollitt, 1994), this has been less

true for both Italy and France According to Lega

(2008, p 260) in Italian public hospitals many

chairs of departments still lack formal authority

over chiefs of clinical units and consequently

behave more like ‘project managers, responsible

for special projects ’ Indeed, it is suggested

that ‘change following the introduction of CDs

was more formal than real’ (2008, p 255)

Similarly, in France, while Chefs de Pole have

‘hierarchical authority over medical, nurse,

administrative teams’ they have ‘no decision

con-cerning the nomination of doctors or the quality

of their clinical activity’ (Vinot, 2012, p 6)

Lastly, one can note differences in development

of non-clinical management roles in areas such as

finance, procurement and human resource

man-agement within hospitals In the USA, both

private and non-profit hospitals tend to invest

heavily in these areas, with non-clinical

‘adminis-tration’ making up a significant part of the health

labour force: 27% according to one estimate

(Woolhandler, Campbell and Himmelstein,

2003) In our own cases the picture is quite

differ-ent with ‘administration’ accounting for a much

lower proportion of the workforce and

expendi-ture However, as can be seen from Table 2, this

does not rule out some quite marked differences

between countries At one extreme, in England,

specialist managers are employed in large

numbers within clinical directorates (Jacobs,

2005; King’s Fund, 2011) and make up a majority

of board members (over 70% according to one

calculation (Veronesi, Kirkpatrick and Vallascas,

2012)) At the other are Italy and Denmark where

hospital governance is dominated by clinicians

with very few specialists employed at lower levels

(Barbetta, Turati and Zago, 2007; Jacobs, 2005;

Kirkpatrick et al., 2009; Lega, 2008).

These conclusions are supported by other

comparative research Dorgan et al (2010)

for example note that non-clinically qualified

managers/administrators make up approximately

42% of all managers in the hospital system of the

UK, 36% in France and only 10% in Italy This

study also assigns a composite ‘management

prac-tice score’ (rating capabilities in the management

of operations, performance and talent) on a

five-point scale, with the UK scoring 2.82, coming

ahead of both Italy (2.48) and France (2.4) (US

hospitals in the sample scored 3.0) In this

research no figures are provided for Denmark,

although Eurostat data suggest that the Danish health system operates with administrative over-heads of only 1.2% of total expenditure, which is low by international standards

Hence, while ideas originating from the JHH model have clearly influenced health policy in the four countries, leading to a broadly convergent move to restructure hospitals along corporate lines, there are also differences in the outcomes of this translation This process in turn, we suggest, has resulted in pathways of change that vary along two key dimensions, as depicted in Figure 1: (a) the extent to which management authority within hospitals has been streamlined (or left ambiguous); and (b) the extent to which management work itself is performed either by clinicians or non-clinical specialists Viewed in this way it can be argued that the English NHS comes closest to the original corporate model of the JHH, while in both France and Italy signifi-cant compromises have been made which essen-tially preserve key elements of professional bureaucracy and consensus administration within hospitals By contrast Denmark represents a hybrid case in which management authority struc-tures have been streamlined but without large investments in non-clinical managers

Accounting for comparative differences

in the context of translation for the clinical directorate model

In this section we turn to our second question of how one might account for variable translation outcomes When discussing this topic it is first

Figure 1 Translation outcomes

important to note how different actors were

involved in the translation process In some cases

change followed initiatives that were taken locally

by managers and clinical professionals who had

been influenced by the model In England, for

example, Guy’s Teaching Hospital in London

first experimented with clinical directorates in

1984, with other hospitals also copying this model

before legislation was introduced (Harrison

and Pollitt, 1994) However, a more important

driver for change has been top-down regulatory

demands from governments responsible for the

bulk of hospital funding Indeed, one can identify

almost coercive institutional pressures linked to a

broader agenda of reforming the management of

public services more generally (Pollitt and

Boukaert, 2011) A key question, though, is how

these reform agendas and their interpretations

were shaped in ways that led to the different

translation outcomes described in Table 2 and

Figure 1

To address this question we return to the notion

of editing rules described earlier and to the work

of Pollitt and Boukaert (2011) on the importance

of elite actors (dominant coalitions of

decision-makers) in each country mediating economic,

political and ideational pressures for reform

Spe-cifically we argue that the nature of rules in a

given context – how much leeway they leave for

local interpretations and deviations from the

tem-plate − will depend upon elite perceptions of the

desirability and feasibility of reforms With regard

to desirability, it is often noted that the level of

commitment of policy makers in different

coun-tries to the restructuring of public services has

been highly variable Hood (1995), for example,

differentiates between ‘high’ and ‘low’ NPM

groups of countries depending on how forcefully

they have sought reforms Others note differences

in the objectives of reform and how far these have

been influenced by neo-liberal ideas emphasizing

the risks of public monopoly and the need to

weaken professional ‘provider power’ (Greener,

2002)

Concerning editing rules relating to feasibility,

perceptions of elite actors of the likely obstacles to

radically changing the existing health

manage-ment system by introducing ‘foreign’ templates

are important In some contexts these may be

con-siderable depending on a number of factors, such

as the ‘countervailing power’ (Light, 1995) of

clinical professionals, the nature of administrative

cultures and the wider political governance of public services (including health) Where profes-sionals are concerned Light (1995) compares systems along a continuum of professional or state dominance In the former, the medical pro-fession ‘controls not only its own work but also a range of related institutions, services, privileges and finances’ (Light, 1995, p 30) arguably making

it harder for governments to impose radical change without consent Thus, in contexts with professional dominance professional actors will engage in a significant editing and modifying of government-introduced templates Closely related

to this are the ‘administrative cultures’ of public services, which have particular consequences for the status of clinical professionals, either as sala-ried employees (or contractors) of the state or (under what Pollitt and Boukaert (2011) term the

Rechtstaat model) tenured civil servants The

wider governance of health systems may also have consequences for the feasibility of reforms, espe-cially when, as is the case of many federal states, hospitals are technically owned and managed by regions that possibly have differing political agendas (Reay and Hinings, 2009)

These considerations, we argue, have direct implications for editing rules that apply in differ-ent contexts and which shape not only the contdiffer-ent

of reforms (e.g whether all aspects of the JHH model are adopted or more loosely translated) but also their timing and pace Indeed, one might even place national health systems along a continuum ranging from those in which radical reforms are considered to be both desirable and feasible and those at the opposite extreme In the former, editing rules will be far more prescriptive in speci-fying new management models for hospitals, offering actors at the policy and local level far less room for interpretation and translation By con-trast, in the latter, perceptions of limited desirabil-ity and feasibildesirabil-ity will result in editing rules that emphasize elements of the national context and thus require translations that considerably deviate from global templates As a result, policy makers have been more selective in how they adopt models of hospital management, and reforms show more continuity with established practices and structures

Turning to our own cases, it is possible to argue that the English NHS sits at the high desirability/ feasibility end of this continuum, with editing rules that are generally more prescriptive and thus

result in translations that do not deviate

consid-erably from global templates Since the early

1980s there has been a strong push by

govern-ments, both Conservative and New Labour, to

adopt the practices of private firms, with clinical

professionals often viewed as ‘part of the problem

rather than the solution’ (Greener, 2002) Linked

to this has been a drive to recruitment of

non-clinical specialists with commercial expertise

(Veronesi and Keasey, 2012) The command and

control structure of the NHS (with professionals

directly employed) has also made it easier for

gov-ernments to legislate and enforce these changes,

most recently through the use of targets and

per-formance management techniques (King’s Fund,

2011) As a consequence, in the English case,

editing rules have been highly supportive of a

more literal translation of the JHH model than

elsewhere

This situation is in stark contrast with the

French and Italian cases Here, although

politi-cians have paid lip service to the NPM, the goals

of policy have often been ambiguous This is

notably true in France, which according to

Bellanger and Mossé (2005, p 119) has adopted

‘one of the least market-oriented models for

reforming its health care system’ Despite the fact

that a stated rationale for reform was to drive up

efficiency, ‘Ironically the word “competition” was

hardly ever mentioned’ (Or and de Pourourville,

2006, p 22) Similarly, in Italy, although attempts

to implement NPM reforms including quasi

markets are more long-standing, here too there

has been considerable ambiguity about objectives

Competition was immediately discouraged

through the arrangement of funding caps for

indi-vidual hospitals Under the close scrutiny of local

politicians CEOs were required to focus on

improving weak areas of their hospitals, rather

than incentivizing their competitive advantages

(Tousijn and Giorgino, 2009) Indeed, at one

point in 2003 the Minister of Health, himself a

doctor coming from the largest teaching hospital

in Milan, explicitly linked quality problems in the

Italian NHS to the introduction of too many

managers (Anonymous, 2003)

This mixed commitment to reform in the

French and Italian cases has been further

exag-gerated by the existence of certain barriers to

radical change In both countries, medical

profes-sionals have considerable ‘countervailing powers’

and make up a higher proportion of the clinical

workforce than in either England or Denmark (OECD, 2008) In Italy doctors have also laid a claim to the general administration of hospitals, with a sub-specialization in ‘Hospital hygiene and organization’ dating back to 1938 Over the years, this category of physicians (referred to as ‘hygien-ists’) emerged as an independent medical speciali-zation, taking care of hospital hygiene, hospital organization, medical archives and epidemiologi-cal analysis (Cantù and Lega, 2002) A related point is that in both France and Italy doctors are effectively state functionaries (or civil servants), their contracts held centrally, with no direct employment relationship with hospitals (see Table 1) Lastly, while the French public health system is relatively centralized (theoretically allowing for the top-down imposition of new management models), this is not the case in Italy where regional governments also play a signifi-cant role in negotiating policy Indeed, as Mattei (2007) suggests, it was largely to avoid a head-on clash with these regional governments (worried about the loss of control over hospitals) that key aspects of the JHH model, strengthening the executive authority of hospital directors, were deliberately watered down in the 1990s

From these conditions we impute editing rules

in France and Italy that have been less prescrip-tive and placed more emphasis on crafting reforms in ways that ensure conformity with the local institutional context and thus play down key aspects of the original corporate model Indeed, one might argue that the JHH template was adopted as a ‘label’, which was loosely combined with existing practices Influenced by the editing rules in these two health systems, actors de-emphasized the initial idea of market orienta-tion and management authority in favour of tra-ditionally legitimate forms of (professional) organization of hospitals

Finally, using this framework it is possible to argue that the Danish case lies somewhere between these two extremes Since the mid-1980s governments have supported the objectives of NPM reforms, increasingly so after 1997 (Kirk-patrick, Kragh Jespersen and Dent, 2011) A command and control healthcare system, broadly similar to the English NHS, also makes it less problematic to implement changes However, in Denmark, long-standing political traditions that emphasize decentralization and partnership with key stakeholders (notably the clinical professions)