Beyond replacement and absorption: exploring integrative pathways to accounting colonisation

Over the past couple of decades, the public sector has become subject to reforms aimed at advancing an economic agenda. These reforms, collectively known as the New Public Management (NPM) movement (Hood, 1991, 1995), have not avoided challenges (Kraus, 2012; Lapsley, 2007). The introduction of market-like structures (Broadbent et al., 1996), efforts to sanction and reward performance (Oakes and Berry, 2009) and stricter demands on financial accountability (Kurunmäki et al., 2003) have been found to conflict with values and norms characteristic of the public sector (Broadbent et al., 1991). For example, costs have been contrasted against needs (Kurunmäki et al., 2003), efficiency against equity (Musgrave, 1964), and markets against frontline professionals (Clarke and Newman, 1997). It is against this background that NPM has been considered a transformative movement (Lapsley, 2008) that seeks to challenge assumptions and norms previously agreed upon and commonly shared in public sector work (Broadbent et al., 1991; Clarke and Newman, 1997; Hood, 1991, 1995).

Since accounting has been found to be central to the expansion of NPM, being “very much the vehicle for economic reason in practice” (Power and Laughlin, 1992, p. 132), scholars have raised concerns about accounting and its role in the so-called colonisation of the public sector sphere. In brief, colonisation refers to the process whereby accounting practices shift social and traditional values within the public sector towards financial or economic considerations (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003). It concerns the “displacement of the social by an economic or managerial logic” (Lapsley, 2007, p. 369) and the development of “a new organisational language which displaces a previously dominant culture” (Power and Laughlin, 1992, p. 132).

In light of this view, efforts have been made to empirically explore the effects of accounting in the context of NPM-led reforms. While some studies have demonstrated colonising effects by illustrating how the emergence of accounting has led to the “displacement” (Lapsley, 2007, p. 369) of public sector values by an economic logic (Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003; Lawrence et al., 1997), others have shown that NPM-led accountability reforms have had only a moderate or even marginal impact on public sector work (Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003; Nyland and Petersen, 2004; Oakes and Berry, 2009). Findings to this effect in the latter group have been attributed to various mechanisms, such as absorbing groups (Broadbent and Laughlin, 1998; Kraus, 2012) or decoupling processes (Kurunmäki et al., 2003; Nyland and Petersen, 2004) that prevent public sector values from being displaced by economic concepts and imperatives. Together, these studies highlight that the “pathway towards colonisation” (Broadbent and Laughlin, 1998) is not a ready-made process, as there are various counter-measures that can be taken to hinder NPM-led accounting initiatives from substantially manifesting themselves in public sector work.

While studies on the colonisation of the public sector have gradually evolved into a “substantial body of literature” (Kraus, 2012, p. 1081), evidence suggests that our understanding of colonisation still requires some additional nuances or refinement. A literature review shows that previous studies have predominantly focused on a dichotomous view of colonisation and public sector values, as if both cannot succeed at once (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003) ^[1]. This creates the following theoretical problem. On one hand, we know that NPM-led accounting initiatives have been strikingly successful at expanding into various areas of public sector life (see Lapsley and Miller, 2019). Yet, we also know that values traditionally associated with public sector work continue to prevail (e.g. Laguecir et al., 2020; Wällstedt, 2020), despite ongoing efforts to transform them over the past thirty years (Hood, 1991). While these empirical observations of NPM’s success, on one hand, and the persistence of traditional public sector values, on the other, can be conceptually understood as a result of decoupling or absorbing mechanisms (Broadbent and Laughlin, 1998; Kurunmäki et al., 2003), research has shown that these responses are unlikely to be sustained over time (see Borell, 2019; Hallett, 2010; Kraus, 2012). Scholars in other domains of research have even argued that existing concepts for describing discrepancies between external pressures and operational practices seem outdated, due to “growing pressures on organisations to align their policies and practices” (Bromley and Powell, 2012, p. 483) in contemporary societies. Thus, if these claims are to be taken seriously, they imply that the “pathway to colonisation” (Broadbent and Laughlin, 1998) must diverge from what has often been outlined in prior research.

In light of this theoretical issue, efforts have been made to refine previous conceptions of colonisation (Campanale and Cinquini, 2016; Gebreiter, 2022; Oakes and Berry, 2009). Recent studies have argued that colonisation can be viewed as a phenomenon that must, in part, be integrated with traditional or social values to effectively “eclipse” (Power and Laughlin, 1992, p. 132) public sector work (Campanale and Cinquini, 2016; Gebreiter, 2022). However, despite offering an alternative approach to theoretically understanding modern developments in public sector settings, studies investigating the integrative aspects of colonisation remain scarce, with calls for further research in this area also being made (Campanale and Cinquini, 2016). The purpose of this paper is to address this gap by further exploring the integrative pathways to colonisation in public sector settings.

To address this aim, the paper will report on empirical fieldwork conducted within a Swedish compulsory school setting. This context is promising for exploring the integrative aspects of colonisation, due to the significant prevalence of market-like reforms combined with a history of traditional welfare values. Since the early 1990s, the Swedish compulsory school system has undergone economic reforms emphasising the need for “reconsideration, rationalisation, and improvements in efficiency” (Government Bill 1990/1991, p. 117, p. 11) through the implementation of structures that support competition, customer choice and costing. Alongside these reforms, other regulations, norms and structures have nevertheless been introduced that adhere to a welfare-oriented agenda; for example, stipulating that resources must be provided to pupils according to their “needs” (National School Act, 2010, p. 800), and attempts to establish more generalised aims of education, such as mitigating “socioeconomic inequities” (National School Act, 2010, p. 800). At the core of these aspirations are professional and administrative bodies of expertise, whose mandate to exercise judgements and operationalise political aims is considered to work as the extended arm of the welfare state (Clarke and Newman, 1997). From this perspective, the Swedish compulsory school system can be considered to operate neither solely under the principles of the market nor solely under welfarist norms, but rather in a welfare market: a context that requires actors to integrate different and seemingly competing values in their everyday work (see also Wällstedt, 2020). Given this empirical outlook, the paper seeks to empirically explore the following research question: how do actors integrate welfare with markets, and how does this particular form of integration simultaneously give rise to colonisation?

The paper is structured with respect to three types of situations observed during the fieldwork, each corresponding to distinct challenges commonly associated with the welfare market: (1) entrusting pupils or professionals as capable experts for evaluating schools, (2) allowing schools to operate in a self-interested manner while still providing equitable services, and (3) ensuring that schools are financially accountable for budgetary performance while also allocating resources based on the needs of their pupils. Based on these empirical vignettes, the paper demonstrates how welfare-oriented values were integrated with economic rationales, and how those forms of integration simultaneously allowed colonisation to occur. A reason for the latter, the paper argues, was because welfarist values had to be transformed into an economic, calculative reality in order to become integrated in practice. Professional views on school quality could only be integrated with a market structure when taking the form of quantitative performance metrics, as these enabled professional values to become an input to pupils’ calculative process of choice. Concerns about equity could only be integrated with competitive, self-interested forms of orientation in the form of monetary incentives, as this made it possible to compensate schools for providing services to pupils who were more expensive to educate. And values relating to needs could only be integrated with the budgetary aims of schools when taking the form of quantifiable numbers, as this made it possible for actors to identify costs that could be reduced outside the scope of needs. As will be shown, all of these examples encapsulated a technical, calculative form of integration that presumed an expansive scope and use of accounting instruments and concepts.

These findings provide further insights within the literature of colonisation. They challenge prior studies (e.g. Broadbent and Laughlin, 1991; Broadbent and Laughlin, 1998; Kraus, 2012) by breaking up an overly dichotomous view between colonisation and public sector values, demonstrating that colonisation is an integrative phenomenon encapsulating economic and social dimensions, although on presumptions driven by an economic agenda. In line with this view, the paper also challenges prior conceptions of accounting and its role for public sector actors. While previous research has to a large extent assumed that public sector actors are reluctant to adopt accounting techniques, seeing them as a potential form of economic intrusion (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Campanale and Cinquini, 2016; Kraus, 2012; Kurunmäki et al., 2003; Holmgren Caicedo et al., 2019), this paper shows that actors can indeed consider the expansive use of accounting as a means to mitigate economic reasoning from having too much impact on everyday work. Overall, the study aligns with and complements recent studies that investigate the integrative aspects of colonisation (Campanale and Cinquini, 2016; Gebreiter, 2022), by highlighting the emergence of a technical, calculative form of integration that previously has been overlooked.

The structure of the paper is as follows. The second section presents a review of the existing research on colonisation, outlining its theoretical underpinnings and operationalisation in the context of accounting and public sector work. The third section provides the empirical background to the study, covering developments in the Swedish compulsory school sector during recent decades, as well as the empirical method used and processes of data collection. The fourth section presents the empirical findings and analysis, organised with respect to three illustrative cases of colonisation. The paper concludes with a discussion of the findings, as well as avenues for future research.

Studies of accounting owe the concept of colonisation to Jürgen Habermas' (1984, 1987) writings on the rationalisation of society. Central to this work are his concepts of “lifeworld” and “systems” (Habermas, 1987, p. 118). “Lifeworld” refers to a “culturally transmitted and linguistically organised stock of interpretative patterns” (Habermas, 1987, p. 124) representing “the lived experiences that give meaning to everyday life and reflect the norms and values of society” (Holmgren Caicedo et al., 2019, p. 375). The lifeworld involves elements of tradition, consisting of the “reservoir of taken-for-granteds, of unshaken convictions that participants in communication draw upon in cooperative processes of interpretation” (Habermas, 1984, p. 124), as well as elements of change. The reason for the latter is due to actors’ “innate capacity to achieve coordinated social action through language and discursive agreement” (Oakes and Berry, 2009, p. 345). Actors can freely test and articulate validity claims, causing their own convictions and presumptions to eventually be challenged. Thus, under such communicative conditions, the “rationalization of the lifeworld” (Habermas, 1987, p. 186) is always a possibility ^[2].

On the other hand, “systems” refer to the “self-regulating action contexts which coordinate actions around specific mechanisms or media, such as money or power” (Thompson, 1983, p. 285). They involve the establishment of rules, sanctions, incentives and feedback loops that serve to constitute individuals, organisations, or other units as stand-alone entities, each pursuing its own objectives while contributing to a coordinated whole ^[3]. Habermas’ prime examples of systems are markets and bureaucracies (Oakes and Berry, 2009). In the market system, actors are generally understood as autonomous entities each pursuing their own interests. Stability in markets is presumed to be attained once the interests of suppliers and customers intersect (Plant, 1982). Such intersections are made possible through money, which serves as a mechanism for creating exchange-based relationships (Broadbent et al., 1991). Similarly, in the bureaucratic system, actors are divided into functionally separate units, each corresponding to a position in the hierarchy. Authorities impose rules and regulations on subordinates, whose deviations are sanctioned or rewarded. Stability is here achieved through the mechanism of power (Broadbent et al., 1991), where each level in the hierarchy contributes to the whole by aiming to maximise its own rewards and minimising sanctions directed towards it. Systems are therefore organised according to an instrumental action orientation, where agents seek to enforce or incentivise one another in order to generate beneficial inputs to their own “calculation of success” (Habermas, 1984, p. 85).

For Habermas, lifeworld and systems are concepts relating to two perspectives whose interrelationships are crucial for understanding societies and their development (Laughlin, 1987). The lifeworld constitutes the “social reality” which gives systems their “meaning” (Laughlin, 1987, p. 97). Systems, on the other hand, are the “more tangible” or “technical” expressions of lifeworlds (Laughlin, 1987, p. 97), whose purpose is to “operationalize the lifeworld’s desires” (Holmgren Caicedo et al., 2019, p. 376). In modern societies, however, these relationships have come to change. As pressures to rationalise lifeworlds increase, actors become less able to revert to traditions in order to share a common worldview (Habermas, 1984). While the move away from traditional norms fosters a more rational and reasonable social order, Habermas also recognised that this could lead to the emergence of new problems. As actors are increasingly expected to reach common understandings through explicit justifications, they also become exposed to the risk of encountering dissensus (Oakes and Berry, 2009). Under these circumstances, systems tend to grow in complexity and scope, as their instrumental action orientation allows effective coordination to occur without the formation of a commonly shared lifeworld ^[4]. In other words, systems play an increasing role in modern societies because they “reduce the costs and risks of communication” (Habermas, 1987, p. 183). But in doing so, systems also get “further and further detached from the social structures from which social integration takes place” (Habermas, 1987, p. 154), and are gradually steered “into an autonomous logic of their own that does not reflect the lifeworld anymore” (Holmgren Caicedo et al., 2019, p. 376). This might cause systems to “turn back destructively upon the lifeworld itself” (Habermas, 1987, p. 186), and hence colonise it. Actors may then experience a loss of meaning, anomie, or personhood (Broadbent et al., 1991), or a “displacement of established norms of behaviour, alienation, and resistance” (Broadbent and Laughlin, 2009, p. 287).

Habermas’ ideas have later been picked up and used within the accounting literature (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Holmgren Caicedo et al., 2019; Kraus, 2012; Kurunmäki et al., 2003; Laughlin, 1987). His notions of lifeworld and system have been considered particularly relevant for addressing “the debate between those who see accounting systems as nothing more than technical phenomena and those who view them as having important social roots” (Laughlin, 1987, p. 98). On one hand, accounting can be perceived from a technical, “systemic” perspective. Performance indicators, resource allocation methods and costing techniques facilitate the establishment of exchange-based relationships (Broadbent et al., 1996), the sanctioning and rewarding of performance (Townley et al., 2003), hierarchical orders and the linkage of inputs to outputs (Broadbent, 1998). On the other hand, accounting can simultaneously be understood from the perspective of the “[life]world or social space” (Lapsley, 2007, p. 369) in which it operates. Unless accounting techniques reflect values and norms historically evolved and commonly shared within a given social setting, or facilitate new values being formed through rational deliberation, they can easily be steered into “an autonomous logic of their own that does not reflect the lifeworld anymore” (Holmgren Caicedo et al., 2019, p. 376). Accounting mechanisms and systems ^[5] thereby potentially colonise the lifeworld by “turning back destructively upon the lifeworld itself” (Habermas, 1987, p. 186). They can be used to construct a new “organizational language” that “displaces a previously dominant culture” (Power and Laughlin, 1992, p. 127) and to “redefine the world, or social space, which [they] enter” (Lapsley, 2007, p. 369).

While scholars recognise Habermas' ideas as potentially useful for exploring the social and technical aspects of accounting, his concepts have also been found challenging to apply directly in empirical research. As stated by Holmgren Caicedo et al. (2019, p. 376), “Habermasian thought is difficult to put into practice because it remains at an abstract and impracticable level”. Consequently, scholars in the field of accounting have made efforts to translate Habermas’ notions into more operable terms. In the context of the public sector, refinements have mainly been made for the purpose of analysing how increasing pressures to become economic, encapsulated in various forms of accounting-led NPM reforms, have eclipsed values and norms which were previously dominant in public sector settings (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Holmgren Caicedo et al., 2019; Kraus, 2012; Kurunmäki et al., 2003; Lapsley, 2007; Oakes and Berry, 2009). This body of research typically starts from the assumption that the public sector has evolved into a distinctive “sphere” characterised by certain norms and values (Broadbent et al., 1991; Broadbent and Laughlin, 1998). It is a form of social space (Lapsley, 2007) dominated by values such as needs, equity, and professional or administrative expertise, which are grounded in claims of serving the common good (Clarke and Newman, 1997). These values have however become challenged. The introduction of NPM-led accountability reforms, which are considered to encapsulate “an instrumental view of management in action” (Kurunmäki et al., 2003, p. 114), can easily uncouple from previously established norms and values. From this point of view, colonisation refers to a process whereby NPM-related accounting systems turn back destructively upon the public sector lifeworld itself (Habermas, 1987), resulting in a displacement of social logic by an economic or instrumental one (Kurunmäki et al., 2003; Lapsley, 2007).

Essential for operationalising Habermas is thus to know what this “displacement” involves and how it occurs. Prior studies have mainly approached this problem from the perspective of resistance and change (for an overview, see Kraus, 2012; Oakes and Berry, 2009). Colonisation occurs when public sector values change due to pressures to comply with an economic agenda (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003; Lawrence et al., 1997). However, public sector organisations resist colonisation by employing various mechanisms that hinder NPM-led accounting initiatives from having any significant impact on core values and activities (Broadbent and Laughlin, 1998). One of those mechanisms is absorbing groups (Broadbent and Laughlin, 1998; Kraus, 2012). This refers to “individuals, specialist work groups or satellite organizations” (Kraus, 2012, p. 1080) functioning as a means to absorb accounting requirements “without interfering with the core values and work practices” (Kraus, 2012, p. 1085). For example, administrative burdens associated with measurement and financial control can be soaked up by a few individuals, hindering accounting-related concepts from being substantially employed at the professional core (Broadbent and Laughlin, 1998). Another mechanism is decoupling (Kurunmäki et al., 2003; Nyland and Petersen, 2004). This refers to a process whereby changes in accounting systems seldom result in a change in core work practices (Weick, 1982). Accounting is instead regarded as a technique that changes in response to the external environment, functioning as a means to portray public sector activities as more legitimate (Kurunmäki et al., 2003). For example, accounting information might be too aggregated to allow for any substantial monitoring of day-to-day work (Nyland and Petersen, 2004) ^[6]. Furthermore, actors may choose to use accounting solely for recording costs after they have been incurred, rather than as a tool for monitoring work according to pre-formulated targets (Kurunmäki et al., 2003). Such an approach permits other values that rarely coincide with costing, such as needs, to become more prioritised instead (Kurunmäki et al., 2003). Together, absorption and decoupling mitigate public sector values from being overtaken by an economic agenda by “making things look different while remaining basically as they have always been” (Smith, 1982, p. 318: see also Broadbent and Laughlin, 1998; Broadbent and Laughlin, 2005; Kraus, 2012).

Based on these operationalisations, studies of colonisation have grown into a “substantial body of literature” (Kraus, 2012, p. 1081). Empirical research has been conducted across various public sector settings, including health care (Kurunmäki et al., 2003), elderly care (Kraus, 2012), schools (Broadbent and Laughlin, 1998) and higher education (Oakes and Berry, 2009). Complementary to studies warning about the colonising effects of accounting in public sector work (Adcroft and Willis, 2005), more sceptical research on colonisation reminds us that this outcome is not inevitable (Broadbent and Laughlin, 1998). Responses aimed at mitigating NPM-led accountability demands from impacting public sector work remain a possibility. In other words, we have learnt that the “pathway towards colonisation” is not a determinate process, but rather one that requires researchers to remain open to new empirical insights (Broadbent and Laughlin, 1998).

Despite all these findings, evidence suggests that our understanding of colonisation still requires some additional nuance or refinement. Previous studies on colonisation have predominantly focused on a dichotomy between colonisation and public sector values: both cannot proceed at once (Broadbent et al., 1991; Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003). Yet we know that NPM-led initiatives have been strikingly successful over the past couple of decades, even as traditional public sector values continue to exist (e.g. Laguecir et al., 2020; Wällstedt, 2020). This empirical observation can be explained only by reverting to absorbing or decoupling mechanisms. However, studies have indicated that these mechanisms are unlikely to endure (see, e.g. Borell, 2019; Hallett, 2010; Kraus, 2012), and that their role may even be overstated in attempts to understand the relationships between external policies and organisational practices in contemporary societies (Bromley and Powell, 2012). In addition to these issues, it has been shown that accounting might serve purposes beyond facilitating NPM (Thomson et al., 2014; Wällstedt, 2020), complicating our understanding of its role in the colonisation of the public sector ^[7]. Thus, if such claims are to be taken seriously, it implies that the “pathway to colonisation” (Broadbent and Laughlin, 1998) will often diverge from what has been outlined in prior research.

In light of this theoretical issue, efforts have been made to refine previous conceptions of colonisation (Campanale and Cinquini, 2016; Gebreiter, 2022; Oakes and Berry, 2009). Recent studies have argued that colonisation can be viewed as a phenomenon that must, in part, be integrated with traditional or social values to effectively “eclipse” (Power and Laughlin, 1992, p. 132) public sector work (Campanale and Cinquini, 2016; Gebreiter, 2022). Gebreiter (2022) argues that our prior understandings of colonisation in health care settings cannot solely be understood as a result of NPM-inspired accountability reforms, but also as a reflection of historical developments within the medical field as such. Increasingly committed to natural sciences, medicine has developed into a field “much more open to economic calculation, and thereby underpin[ing] the increasing adoption of managerialist hospital accounting practices from the 1980s onwards” (Gebreiter, 2022, p. 1191).

A similar line of thought can also be identified in a study by Campanale and Cinquini (2016). Examining the role of accounting in a health care setting, the study showed how collaborative interactions between clinicians and controllers, aimed at making accounting systems “more clinically oriented than in the past” (Campanale and Cinquini, 2016, p. 60), simultaneously facilitated, rather than hindered, the occurrence of effective colonisation. As accounting figures became more accepted, clinicians became more open to a “managerial culture inspired by NPM” (p. 60). This led Campanale and Cinquini (2016) to propose a reciprocal approach to colonisation, i.e. viewing accounting systems as partially influenced by public sector values. Together, these studies serve to nuance our prior understandings of colonisation by showing that the expansion of economic reasoning is more complex than previously assumed: accounting systems colonise the public sector lifeworld by simultaneously reflecting elements or parts of it.

However, studies investigating this alternative form of colonisation in practice are still scarce (Campanale and Cinquini, 2016; Gebreiter, 2022). With the exception of Campanale and Cinquini (2016), no other study has explicitly addressed the integrative pathways to accounting colonisation in the daily operations of public sector work ^[8]. Calls have therefore been made for more studies that empirically investigate how actors maintain “congruence” between economic and social values (Campanale and Cinquini, 2016, p. 73). The purpose of this paper is to address this gap by further exploring the integrative pathways to colonisation in public sector settings ^[9].

To address the aforementioned theoretical aim, this paper will report on empirical fieldwork conducted within a Swedish compulsory school setting. This context is promising for exploring the integrative aspects of colonisation due to the significant prevalence of market-like reforms in Swedish schooling, combined with a history of welfare-oriented values. This section provides a background to these reforms, followed by a description of the empirical data collected for this study.

In line with general developments occurring within public sector services worldwide, the Swedish compulsory school sector has undergone a transformation aimed at advancing an economic, market-like agenda. This transformation dates back to the 1980s and early 1990s, a period when “reconsideration, rationalisation, and improvements in efficiency” (Government Bill 1990/1991, p. 117, p. 11) served as significant benchmarks for restructuring the educational sphere in Sweden. This agenda led to various reforms that continue to play a central role in the management of Swedish compulsory schools to this day. One of these reforms was the “freedom of choice reform” (Government Bill 1992/93, p. 230), whose purpose was to “achieve the greatest possible freedom for children and parents to choose schools” (Government Bill, 1991/92, p. 95, p. 8). Pupils and their parents were considered as autonomous agents capable of rationally assessing and choosing schools that best “match one’s own preferences” (Government Bill, 1992/93, p. 230, p. 27). Central to expanding freedom of choice was broadening the alternatives available for selection. To this end, the freedom of choice reform was coupled with another structural change, known as the “free school reform” (Government Bill 1992/93, p. 230). Schools that were privately owned, such as limited liability companies traded on the stock exchange, were permitted to compete alongside public schools to “break up the public school monopoly” (Government Bill 1992/1993, p 25–26). Attention was directed towards the issue of establishing “equitable rules of the game” (Government Bill 1992/93, p. 230, p. 27) by allowing private schools to access public tax funds on terms equal to or similar to those of public schools. This opening up of public provision was not only considered important for expanding the possibilities of choice but also for improving productivity and the quality of education at large:

The creation of atomistic, self-regulated entities (see Lapsley and Miller, 2019) was further driven by reforms and systems that evolved in parallel with this change. Responsibilities related to the funding and management of schools were decentralised from the Swedish state to the country’s 290 municipalities (Government Bill 1990/91, p. 18). Budgetary requirements, initially enforced on the municipalities, trickled down to the public schools, which now became regarded as their own economic units able to make relatively independent decisions concerning their finances (Borell, 2019). Efforts to improve productivity were enhanced by additional reforms that emphasised the importance of measuring and evaluating school performance in a more “systematic” fashion (National School Act, 2010: 800). National comparison lists were made available to the public in order to assess the performance of all schools in Sweden, both enabling customers to make an “informed” choice and incentivising schools to improve their measurable productivity. New allocative systems emerged that served to direct the flow of monetary resources according to the choices made by parents and their pupils, coupling financial flows to market demand (see also Broadbent et al., 1999). Together, these measures made the educational sector increasingly market-like (Lundahl et al., 2013), consisting of units whose focus on their own performance, finances and preferences was believed to generate benefits for the population at large. Stated in other terms, these reforms encapsulated an economic orientation that aimed to “autonomize the social sphere by giving the state the role of promoting and overseeing competition in the name of efficiency and choice” (Lapsley and Miller, 2019, p. 2212).

As a result of these reforms, the Swedish educational sector has been regarded as “one of the most liberal in terms of decentralisation and market elements” in the OECD area (Lundahl et al., 2013, p. 499). However, despite an unmistakable expansion of economic, market-like reasoning into compulsory school settings (Lundahl et al., 2013), the transformation has been far from complete. Even today, traces can be found in legal stipulations, everyday practices, aims and beliefs that reflect more traditional, social values previously established under the governance of the “welfare state” (see Ball, 2015; Clarke and Newman, 1997; Miller and Rose, 1990). According to the National School Act (2010, p. 800), schools are legally obligated to provide their pupils with resources based on their needs — a requirement that limits the focus on budgetary goals or efficiency gains in the planning of daily tasks. Needs represent an inviolable right that cannot be traded off (Lawrence et al., 1997).

In addition to this requirement, other similar stipulations can also be found in legislation, such as those concerning the more generalised aim of preventing social inequity in the provision of educational services (National School Act, 2010, p. 800). This means that schools are not permitted to deliver educational services in a manner that perpetuates socioeconomic inequalities in society, thus contradicting a more atomistic, economic approach to education. At the core of educational services are professional and administrative bodies of expertise, whose mandate to exercise judgement and operationalise political aims is considered to function as the extended arm of the state (Clarke and Newman, 1997). The latter stands in contrast to the idea of treating pupils and parents as knowledgeable customers capable of evaluating the activities and performance of schools. All of this suggests that the Swedish compulsory school sector can neither be understood as operating under the principles and orientations of the market, nor of the welfare state, but rather of a welfare market: a context that requires actors to integrate different and seemingly competing values in their everyday work (see also Wällstedt, 2020).

Thus, the Swedish compulsory school setting can be considered as a site relevant for exploring the integrative aspects of colonisation. The strong focus of this setting on market-led reforms, combined with the remaining prevalence of welfarist values, suggests that colonisation is likely to be found, yet in ways that still align with the public sector lifeworld. Within this kind of perspective, the question that this paper seeks to empirically answer is: how do actors integrate welfare values with market values, and how does this particular form of integration simultaneously give rise to colonisation?

The empirical fieldwork was conducted within one of Sweden’s 290 municipalities, here referred to as “City”. In Sweden, municipalities are self-governed local authorities responsible for the delivery of different services to their citizens. Some of these services include child care, social services, waste and water treatment, elderly care and compulsory schooling. Municipalities have a considerable degree of autonomy and have independent powers of taxation. Their main source of revenue stems from their own taxation systems, but they also receive funding from state grants and fees for certain services.

In brief, City is one of the largest municipalities in Sweden, encompassing more than 200 compulsory school units within its geographical/administrative area. Due to the marketisation of the Swedish compulsory school sector, these schools were either privately or publicly owned, with the latter falling under the direct oversight of City’s administration ^[10]. However, regardless of the form of ownership for schools, City was still responsible for funding all school units within its area. This meant that the activities in all school units ultimately had to be financed by taxes collected from City’s citizens. As will be shown, the resource allocation system was thereby an important instrument for City with which to influence both private and public schools, serving as the mechanism for creating the “right” incentives in the market.

Given its scope and size, City divided its public schools into different districts. Each district varied both in terms of the size of its population and of each population’s composition ^[11]. As a result, schools generally consisted of student bodies with compositions reflecting the socioeconomic composition of each district ^[12]. Some of the schools had a “high-end” socioeconomic structure, meaning that their pupils predominantly had parents with high-level academic degrees and/or who had Swedish as a first language. In contrast, other schools had a “low-end” socioeconomic structure, composed of pupils with the opposite characteristics. Classifications of this sort were all supported by calculations, usually performed by statisticians, that eventually served as inputs for other calculations, such as resource allocation systems or performance evaluation systems.

In terms of supervision and control, the municipality’s governance was structured according to a hierarchical model where each level was held accountable to the next. Political representatives formulated broader goals and directions, which were then operationalised by administrative departments in alignment with the specific context of each relevant social service. In terms of compulsory schooling, aims were operationalised by people working at the municipality’s Educational Department. Responsibility for evaluating the local school units was held by district managers, a role intended to serve as a “link” between the administration and the daily operations of schools by holding principals accountable for budgetary and academic performance. The principals at the local school units were in turn expected to plan and evaluate their daily operations in accordance with measurable targets.

The data collected throughout the empirical fieldwork reflects both of the aforementioned levels of control. It consists of interviews conducted with district managers, statisticians and controllers at the “administrative level” of City, all of whom worked at the Educational Department. The data was also generated through empirical investigations conducted at the “local level” of schools, encompassing observations made during various school meetings and interview responses from different types of personnel and at different school units ^[13]. In total, the data was generated through 41 interviews and 28 observation sessions conducted between 2016 and 2019. A summary is provided in Table 1.

The process of collecting data was informed by a qualitative research approach, characterised by “ongoing reflection on data and its positioning against different theories such that the data can contribute to and develop further the chosen research questions” (Ahrens and Chapman, 2006, p. 299). Driven by an interest in exploring the colonising effects of accounting in public sector settings (Broadbent and Laughlin, 1998; Lapsley, 2007; Power and Laughlin, 1992), the study focused on the technicalities and experiences of accounting in relation to a New Public Management (NPM) agenda. Interviews were conducted at the administrative level to gain a deeper understanding of the details of accounting systems used to manage schools. Observations and interviews were conducted at the local school units to provide a closer look into how accounting systems operated in practice, as well as the assumptions, values and commitments espoused by school participants. Data were complemented by document studies, covering national legislation as well as documents provided by City. The combination of data collected at both the administrative and local levels aligns with prior research on colonisation, insofar as the latter must be understood within the context wherein accounting is experienced and used (see, e.g. Broadbent and Laughlin, 1998).

Throughout the fieldwork, particular attention was devoted to situations where actors perceived that NPM-led accounting imperatives were steering educational activities “out of hand” (Holmgren Caicedo et al., 2019), i.e. “into domains that did not align with or reflect the lifeworld” (Broadbent et al., 1991, p. 5). This was empirically examined by investigating situations where actors expressed criticism of current accounting systems, such as criticism of their being “flawed”, “incomplete”, “unfair” or “coercive”. However, equal attention was paid to situations where actors expressed relatively positive attitudes towards the same accounting systems; for example, where actors perceived the current systems to be a “fair” way of allocating resources, that a particular performance indicator was considered to be a “good” measure of school quality, or other situations where accounting was believed to be a practical solution to problems caused by market-led reforms. Consistent with the aforementioned critique (Campanale and Cinquini, 2016), the objective was to remain open to potential ways that actors perceived accounting to be integrated with public sector values and ideals, rather than being solely a facilitator of NPM in practice (see also Wällstedt and Almqvist, 2015).

Data analysis was conducted by gathering quotes into a single document, sorted in consonance with situations where actors experienced conflicts between welfarist values and NPM, and how these conflicts could be mitigated by using accounting as a means to integrate them in practice. This resulted in rich empirical data consisting of various examples of both tension between and integration of accounting, welfare and NPM. Guided by the intention to unpack the uniqueness and multiplicity of the data rather than to portray its statistical generalisability (Payne and Williams, 2005), the analysis eventually resulted in three vignettes, each highlighting different integrative aspects of colonisation.

In line with general developments in the Swedish compulsory school sector, City was an entity that had implemented various forms of accounting systems facilitating a marketist agenda. For example, systems for allocating resources based on pupil enrolment created a link between customer choices and the financial rewards that schools received (Broadbent et al., 1996). Furthermore, performance indicators published online made it possible for pupils to compare schools and make putatively rational choices (Broadbent and Laughlin, 1998). Further, by delegating budgetary responsibility to local school units, economic entities could emerge – a central feature in the creation of a rational market supply. All of these examples illustrated how accounting operated as a “vehicle for economic reason in practice” (Power and Laughlin, 1992, p. 132), constructing various actors in the school setting — school units, professionals, pupils, parents — as relatively autonomous entities pursuing their own economic interests.

For this reason, accounting was seen as being associated with problems. Systems for facilitating customer choice could conflict with what welfare professionals considered to be a fair performance of their work. Additionally, systems that treated schools as independent financial entities could potentially compromise broader societal goals, such as addressing social inequalities or providing pupils with resources according to their needs. In other words, the economic systems could easily turn towards “an autonomous logic of [its] own that does not reflect the lifeworld anymore” (Holmgren Caicedo et al., 2019, p. 376).

However, actors attempted to mitigate these tensions by identifying strategies that aligned the lifeworld with economic imperatives. The rest of this section outlines three scenarios for how this could take place, each corresponding to a particular source of tension, and how they were managed through various types of integration. As will be shown, these forms of integration were still associated with colonisation, albeit in a subtle and specific form.

The first scenario pertains to the conflict between the choices made by pupils and their parents on one hand, and what teachers, administrators and other welfare experts considered to constitute good school quality on the other. We can see this scenario illustrated in the empirical insights from a specific performance meeting, where a district manager, a principal, assistant principals and teachers discussed a given school’s performance according to a report. The meeting began with a statement from the district manager. With the report projected on a screen, and the other school personnel present in the room, the district manager said:

The district manager here referred to a metric that was frequently encountered throughout the empirical work: the “eligibility metric”. This metric represented the percentage of pupils which had successfully passed all the subjects required to advance to the upper-secondary level in the academic system. This metric featured in various reports, such as in national comparison lists which served as a basis for pupils’ choice of school, or, as in this case, in reports used by City to compare public schools in different district areas ^[14]. As subsequent interviews revealed, it was even regarded as one of the most important metrics in the context of education, symbolising social mobility and serving as a predictor of pupils’ “future life chances” ^[15]. Given its centrality, it seemed no coincidence that the district manager singled out this particular metric during the performance meeting.

As stated in the quote above, the district manager thought that the school had demonstrated “fantastic” performance on this metric during the recent reporting period. This conclusion was made on the basis that the school had achieved significantly higher results compared to other schools within the same district.

For this reason, the manager considered the school to demonstrate good quality – a view that was found to be shared by the principals, assistant principals and teachers present in the room. The subsequent discussions did not challenge the assumed relationship between the eligibility metric and quality; rather, they were grounded in it. An example was when an assistant principal cited the aforementioned metric as evidence of the school’s “great results”:

Thus, contrary to the belief that public sector actors find performance measurements too reductive and threatening to more contextual representations of quality (Adcroft and Willis, 2005), the meeting demonstrated a situation where actors agreed upon the significance of a particular metric. As indicated by the quote above, the problem was rather found to be related to how the indicator was used by those who articulated market demand: the pupils. Whereas the professionals considered their own school to be performing well, statistics about the popularity of the school told a different story. Few pupils had applied, leading to a situation where the school was having difficulty filling its classes. Since the municipality allocated financial resources based on the number of pupils enrolled, the school now faced difficulties meeting its budgetary requirements ^[16]. In other words, we can see a situation in this example where market demand and its associated accounting systems “uncoupled” (Broadbent et al., 1991) from “worldviews” shared across administrative and professional bodies of expertise.

As a result, a new conversation began to take shape during the meeting. After a few minutes of discussion, the principal came up with a suggestion. By publishing the scores from the eligibility metric on their website, and by spreading this news to parents, it was believed that the problem with application levels could be resolved:

The argumentation above illustrates a specific form of integration between welfarism and assumptions on customer expertise. By amplifying the visibility and use of the eligibility metric, actors intended to change customer choices by influencing their conception of what good school quality entailed. Money and resources would then flow according to what teachers, principals, and district managers already deemed to be a “fair” representation of school quality ^[17]. In this way, the market could exercise influence alongside that of welfare expertise without producing too much tension.

However, this was a form of alignment still colonised by an economic agenda. Without the possibility to translate welfare-oriented views on quality into a quantitative metric, there would be no way to influence customer choice. Further, even in cases where such a translation is achieved, there are no guarantees that customers actually use these metrics as inputs in their decision-making process. In sum, integration was presumed by the ability to align welfare-oriented views with the rational calculations performed by actors articulating market demand.

The second scenario concerns the establishment of economically self-interested school units on one hand, and the broader collective equitable goals of the education system on the other. The second scenario centres around interviews and observations conducted regarding the reimbursement system, used to allocate financial resources from the administrative to the local school units in the municipality. More specifically, it pertains to the critique and changes made to this system throughout the course of the empirical fieldwork.

In order to better grasp these changes, we will first consider how resources were initially allocated within City. As mentioned earlier, a key aspect of the allocation system was to link market choices with the financial funding received by the schools. Like many other municipalities in Sweden, City had developed a voucher system (see Broadbent et al., 1996). This system set a “price” that the municipality paid for every pupil enrolled in a given school. The system formed a direct link between market demand and financial rewards, since enrolment was determined by pupil choice. Thus, the more attractive a school was, the more funding would be received.

One issue with the voucher system, however, was that it failed to account for the varying costs of education associated with catering to different socioeconomic groups. A school located in a socioeconomically “low-end” district could therefore be at risk of facing budgetary deficits. This created a problem, since public schools were sanctioned and rewarded based on their budgetary outcomes, leading to an unfair system of evaluating their individual performance. Furthermore, since City was responsible for allocating resources to private schools as well, concerns could arise regarding how economic incentives were created in the marketplace. In other words, the voucher system could potentially steer school units “out of hand” (Broadbent et al., 1991), by formulating incentives that reproduced inequitable structures.

It was for this reason that the municipality had developed an alternative system alongside the voucher system, also known as the “socioeconomic compensation scheme”. This system worked as a mechanism to compensate schools based on the prevailing socioeconomic background of their students. The calculation of compensation amounts was quite complex, derived from a regression model analysing the relationship between historical grade performance and various socioeconomic variables. These variables included the number of years pupils had lived in Sweden, the educational backgrounds of their parents, social assistance and the like. The calculation resulted in a price increment that was added on top of the voucher price.

As City included schools located in areas with varying socioeconomic structures, the amount of compensation could differ quite significantly between schools. Based on the calculations mentioned above, a school situated in a socioeconomically disadvantaged area could receive almost twice as much funding compared to a school located in a more affluent area. However, despite these differences, actors generally perceived the socioeconomic compensation scheme to be a relatively “fair” way of allocating resources to the schools (INT#2 and INT#11). An example of this attitude is provided below, where a bursar — who worked at a school with a “high-end” socioeconomic structure – expressed the following opinions about the scheme:

However, even if the compensation scheme was generally understood as a fair way to allocate resources, it was also found to cause problems. During the empirical fieldwork, actors still opined on a number of occasions that the system ”mispriced” pupils and their socioeconomic backgrounds. One such example is provided below. Here, a principal criticised the system for treating “newly arrived pupils” (a term used to describe pupils who have recently immigrated to Sweden) too homogeneously. Consequently, this created market incentives to “profit” from having newly arrived pupils of a specific kind – those who had already received a significant level of education in their home countries:

From this example, we see that the ambition to integrate the economic interests of schools with equitable aims could easily fail. The socioeconomic compensation scheme relied on classifications that did not always correspond to the “true cost” of pupils, thus creating incentives that served to reproduce rather than mitigate social inequities in the school market.

As it eventually turned out, the principal in this example was not alone in raising this kind of concern. Due to an increasing stream of immigrating pupils between 2015 and 2016, the municipality decided that the socioeconomic scheme had to be changed. In line with the argument proposed by the principal above, this change was motivated by the system’s inability to capture heterogeneity among the newly arrived pupils. In an internal document produced in City, the following was stated:

As a result of these shifts, a new solution emerged. By creating another system that sought to better represent the heterogeneous background factors of the newly arrived, it was believed that the issue of inequities caused by the system could be addressed. This led to a new system called the “compensation scheme for the newly arrived” that stratified newly arrived pupils according to their home background and the number of years they had spent in Sweden. An illustration of this system is provided in the Table 2 below.

Taken together, these empirical insights illustrate a certain form of integration. Actors were repeatedly found to be reverting to the resource allocation system, making adjustments to it by developing finer categorisations. These changes were made on the underlying assumption that the costs associated with different social groups could be made to correspond more closely with the price that the municipality paid. Hence, integration was possible: schools could continue to pursue their financial interests without posing a threat to equity, because in a market where profits are uniform across different socioeconomic groups, there are no incentives to reproduce inequitable structures in society.

We therefore see another case of how integration was influenced by an atomistic, economic mindset. Equity could only be sustained by being translated into monetary terms, as this was the mechanism for creating the “right” incentives in the school market. Colonisation thus occurred by letting market-like structures dictate the premises for how social and collective aims could possibly be pursued.

The third and final scenario concerns the tension between costing and needs, drawing from observations and interviews conducted at a local school unit in City, here referred to as “Suburb” ^[18]. Specifically, it focuses on the process of staffing and scheduling, which is influenced by the twin imperatives of maintaining a balanced budget (costing) and adhering to national legislation outlined in the national time plan (needs). Moreover, this third scenario highlights how Suburb integrated these requirements into its daily operations – and how this form of integration eventually became subject to internal criticism.

However, before we examine this instrumental, technical form of integration, it might first be necessary to elaborate on the budgetary and time plan requirements. Similarly to areas where the other public schools were in City, Suburb was constructed as an independent economic entity responsible for securing its own financial ends. As it turned out, one of the most important aims was to achieve a “balanced budget”: a term often mentioned throughout the empirical fieldwork, meaning that costs should not exceed revenues. In practice, this implied that the cost of personnel (which is by far the largest cost item in a school’s budget) had to be planned for within the revenues received from the voucher system, the socioeconomic compensation scheme, and the reimbursement system for the newly arrived.

Whereas the budget pertains to the issue of costing, the national time plan is related to the more social value of allocating resources according to needs. In brief, the time plan is governed by national regulation (National School Act, 2010, p. 800), and thus has to be complied with not just in City, but for all schools located in Sweden. Specifically, it regulates the number of hours of education that a pupil is entitled to throughout their compulsory schooling, sorted into different subjects (such as mathematics, English, Swedish, art, music, and so on). In this way, the time plan had a significant impact on the staffing of schools by stipulating how many hours a pupil “needed”. Thus, contrary to the idea of costing, which required schools to allocate and reduce resources in ways that enhanced efficiency, the time plan portrayed resources as a legal right that could not be traded off against the time spent on someone else (National School Act, 2010, p. 800).

As both of these requirements were connected with sanctions, Suburb therefore had to identify ways to integrate both in its everyday work ^[19]. This brings us to the next set of empirical observations. Here, we find ourselves in a room where the principal of Suburb is planning the staffing and scheduling for the upcoming term. With a decisive gesture towards the computer screen, he explains the process that led to the formulation of Suburb’s current plan:

The scheduling and staffing of Suburb were described here as an iterative process informed by the national time plan and budgetary goals. The process began with the principal identifying pupils’ needs according to the figures stipulated in the national time plan, and then developing a local plan accordingly. The principal then tested the local plan against the budgetary requirements. In cases where the local plan did not hold, the principal revisited Suburb’s staffing, making adjustments to reduce costs without compromising the national time plan. As illustrated above, this involved two main strategies. The first strategy focused on rationing costs related to personnel not directly involved in classroom work, such as psychologists or counsellors, as these groups did not contribute to the hours included in the time plan. The second strategy was to ration the cost of teachers by increasing the number of hours that the teachers in Suburb taught in class, minimising the total number of teachers needed to meet the time plan requirements. Together, these two strategies eventually led to a final local plan that fulfilled both budgetary targets and time plan requirements.

However, despite integrating costing and needs in practice, the local plan eventually became subject to internal criticism. To exemplify with an observation session, such criticism emerged at one annual meeting. After the principal had presented the local plan to the staff, a pedagogue raised her hand with the following concern:

As indicated here, teachers believed that the time devoted to mentoring — an activity that was considered wasteful from a budgetary and time plan viewpoint — had been reduced too much. In fact, teachers believed that mentoring was something that their pupils needed more of. However, as the principal later explained, there was no room for making such changes to the current staffing plan. Expanding the time devoted to mentoring would mean diverting resources from classroom teaching, making the plan “too expensive” (OBS#20) to pursue. The concerns raised by the teachers were thereby silenced in favour of efforts to integrate the time plan and budget in practice – a form of integration that itself presumed the stipulations of the latter to be true and correct.

In fact, this was not the only objection raised against the schedule. During the meeting, several teachers raised concerns that the plan would lead to too much stress. Being obliged to teach 18 hours per week was considered too burdensome given the time needed for preparation, resolving conflicts between pupils, contact with parents and other activities that had to take place outside of class. This was also later explained by a special pedagogue as follows:

Thus, teachers expressed a need to reduce teaching time from 18 to 16 hours per week in order to reduce stress among the staff. However, despite these calls, the principal still considered the plan to be valid. If the number of teaching hours were to be reduced, more teachers would need to be hired in order to fulfil the time plan, resulting in budgetary overspends that badly needed to be mitigated.

Taking these issues and conflicts together, it becomes evident that the integration between costing and needs was implemented in a particular way. The management of Suburb made use of definitions of needs provided by the time plan in order to identify costs to be reduced – a strategy that was heavily reliant on definitions uncoupled from local perspectives at Suburb. In other words, this was an example of integration that was purely technical in form, dictated by calculative overlaps rather than fostered by locally shared perspectives.

This paper aimed to theoretically explore alternative pathways to how accounting colonises the public sphere. Informed by fieldwork conducted in a Swedish compulsory school setting – a site characterised by the prevalence of both traditional, welfare values on one hand, and an economic, market-like agenda on the other – the paper addressed this aim by empirically exploring the following research question: how do actors integrate welfare with markets, and how does this particular form of integration simultaneously give rise to colonisation?

The abovementioned cases each contribute to answering this question in the following ways. Addressing the first part of the research question, they provide a rich empirical account of how welfare was integrated with markets in an educational setting. For instance, the first case provided insights into how welfare experts attempted to integrate their own views on quality with customer choices in the market. The second case illuminated how the more collective aim of equity was integrated with the economic interests of independent schools. The third case showed how one particular school aligned matters of costing and needs when planning its staff schedule. Continuing to the second part of the research question, the three scenarios also demonstrated a particular form of integration that had a colonising dimension. The reason for this, the paper argues, is that welfarist values had to be transformed into an economic, calculative reality in order to become integrated in practice. Professional views on quality could only be integrated with choices in the market when taking the form of quantitative performance metrics, as this enabled professional values to become an input in pupils’ calculative process of choice. Concerns about equity could only be integrated with competitive, self-interested forms of orientation in the form of monetary incentives, as this made it possible to compensate schools for providing services to more costly pupils. And values relating to needs could only be integrated with the budgetary aims of schools when taking the form of quantifiable numbers, as this made it possible for actors to identify costs that could be reduced outside the scope of needs. In other words, all three of these examples encapsulated a technical, calculative form of integration that presumed an expansive scope and use of accounting instruments and concepts.

Underlying this technical, calculative form of integration is also a certain realisation about the colonisation of the public sector sphere. The process of colonisation is a subtle one, wherein displacement does not necessarily involve a fully-fledged replacement of public sector values with economic concepts and ideals. Instead, colonisation occurs by reconstructing the public sector into a domain where social and collective values can only persist by becoming integrated into economic systems. Thus, the power of colonisation lies in its ability to frame disintegration of economic systems as a risk that must be mitigated in practice. Central to managing this risk is accounting, as numbers, statistics, costing and prices enable actors to translate social and collective values into technical, economic terms.

The findings of this paper thus provide an alternative story about the pathway to colonisation. First, the paper stands in contrast with studies by Broadbent et al. (1991), Broadbent and Laughlin (1998), Kraus (2012), and Kurunmäki et al. (2003) by breaking up an overly dichotomous conceptualisation of how colonisation occurs. While prior studies have predominantly viewed the expansion of economic reasoning as incompatible with public sector values, depicting colonisation as a process of supplementing the latter with the former, this paper nuances this perspective by demonstrating the integrative nature of colonisation in public sector settings. This finding is arguably important in order to understand why we have seen such a broad expansion of NPM over the past couple of decades (Lapsley and Miller, 2019) and yet still observe a strong prevalence of traditional public-sector values in everyday public-sector work (Laguecir et al., 2020; Wällstedt, 2020). Treating colonisation as a form of integration makes it possible to understand the coexistence of the two without having to revert to decoupling or absorbing mechanisms (see Broadbent et al., 1991; Broadbent and Laughlin, 1998; Kraus, 2012; Kurunmäki et al., 2003).

Second, and in line with critiques of explanations based on decoupling and absorption as proposed in other domains of research (Bromley and Powell, 2012), this paper diverges from the perspectives of Broadbent et al. (1991), Broadbent and Laughlin (1998), Kraus (2012), Kurunmäki et al. (2003), and Nyland and Petersen (2004) by highlighting another relationship between accounting and the continuation of public sector values. Previous studies have mainly understood the persistence of public sector values as an outcome of absorbing or decoupling mechanisms, which serve to minimise the influence and use of accounting in everyday work. However, as this paper has shown, actors can also consider an expansive use of accounting as a means to make traditional welfarist values prevail. Under integrative–instrumental pathways to colonisation, accounting becomes the primary means to integrate and thereby protect more socially and collectively oriented aims and values. It is precisely in the scope of this integration that the more subtle forms of colonisation can emerge.

This paper therefore partially aligns with more recent studies, sharing the view proposed by Campanale and Cinquini (2016) and Gebreiter (2022) on the need to explore the integrative aspects of colonisation. In particular, this study aligns with the work of Campanale and Cinquini (2016), who also investigated alternative pathways to colonisation in public sector practice, albeit with an empirical focus on health care rather than education.

Aside from its empirical focus, this paper also differs from the aforementioned study in one theoretical aspect. Campanale and Cinquini (2016) proposed that the integrative pathway to colonisation mainly is navigated through “collaborative interactions” (p. 60) wherein different parties enter into dialogue to reach some form of agreement. This led them to discuss colonisation from the viewpoint of an “integrative-interactive approach” (p. 59) wherein actors strive to test and justify the plausibility of economic and public sector values – a form of compromise which, however, serves to make the former more prominent and articulated than the latter (Campanale and Cinquini, 2016). However, the findings of the present paper reveal a more technical or instrumental approach towards integration that caused colonisation to differ in both its mechanisms and its character. This study shows how welfare experts strived to integrate their views with market demand by attempting to alter the inputs to customers’ calculations in their rational process of choice – a form of integration that did not necessarily involve a commitment to enter into “collaborative” (Campanale and Cinquini, 2016) patterns of interaction, where both parties participate in discussion in order to reach an agreement. Furthermore, this study demonstrates how actors attempted to reconcile the establishment of self-interested suppliers in the school market with the collective aims of education by developing a price that corresponded with the cost of serving different groups of pupils. Although involving elements of dialogue (as actors had to identify a price that was considered “fair”), discussions were mainly guided by an underlying motif of maintaining, not “compromising” (Campanale and Cinquini, 2016) economic calculations in the market. Lastly, and perhaps most differently from Campanale and Cinquini (2016), this study shows how schools integrated costing and needs by developing a plan that exploited calculative overlaps defined by budgets and regulations – a form of integration that did not even presume locally shared “agreements” (Campanale and Cinquini, 2016) between professionals and managers to succeed. Together, all of these examples illuminate a technical, calculative form of integration, where consensus was neither sufficient nor necessary for aligning welfare values with an economic rationale. The primary mechanism of integration was through intersecting calculation rather than mutual or participatory argumentation.

From this point of view, the findings of this paper do not sit very well with the integrative–interactive approach as laid out by Campanale and Cinquini (2016). An integrative–instrumental approach seems more appropriate instead. While this concept also encapsulates the integrative aspects of colonisation, it does so from a more “instrumental” (Kurunmäki et al., 2003) or “technical” standpoint (Laughlin, 1987). It emphasises how actors integrate the public sector lifeworld with an economic agenda by calculations rather than participatory argumentation. In Habermasian terms, such a process entails a “systemic” approach focused on the establishment of rules, sanctions, incentives, or feedback loops for the purpose of maintaining entities as stand-alone units, pursuing their own calculations while contributing to a coordinated whole (Habermas, 1987). The concept thereby complements Campanale and Cinquini’s (2016) work by operating as a specific form of “reciprocal colonisation” distinct from the integrative-interactive approach.

The integrative-instrumental pathway to colonisation thereby encapsulates certain realisations about systems and lifeworlds. As indicated by its integrative character, such a pathway involves economic systems reflecting the public sector lifeworld (see Broadbent et al., 1991). Whereas this form of integration allows actors to mitigate a full replacement of social with economic values, it is still an approach that tends to overstate systemic over lifeworld premises and mechanisms. As Habermas reminds us, forming social values under an economic logic is still a reductive approach from the lifeworld’s perspective. For example, as this study has shown, attaining integration through intersecting calculations withdraws actors from truly engaging in consensus-forming obligations. Making professional worldviews fit into a rational process of choice easily falls short, since this entails certain premises for how accounting should be used. And putting a price on social ends can easily obscure their multifaceted meanings. Thus, the integrative-instrumental approach downplays the superiority of lifeworld over systems, assuming that the former can only be sustained when being incorporated into the mechanisms of the latter. It is precisely this belief that creates colonising outcomes: it turns “back destructively upon the lifeworld itself” (Habermas, 1987, p. 186) by making it operate under instrumental conditions ^[20].

While studying the integrative-instrumental pathway to colonisation offers valuable insights into the tensions between economic reforms and the public sector lifeworld, these insights warrant deeper exploration to fully understand their broader implications. The integrative-instrumental approach seems most likely to emerge when pressures to adopt economic imperatives are strong enough to make absorption or decoupling unsustainable in practice, yet weak enough to avoid fully replacing public sector values. More studies are needed to investigate the interrelationships between different pathways to colonisation in order to understand their emergence, stability, and disappearance over time. Additionally, further research is required to examine the relationships between instrumental and interactive pathways to colonisation and the process of rationalisation, i.e. when the lifeworld evolves through an open and deliberate process.

This study suggests that integrative-instrumental forms of colonisation are further removed from rationalisation than interactive ones. However, greater attention should be paid to the underlying mechanisms that lead actors to follow colonising pathways despite their constraining nature. In this regard, Habermas’ concept of communicative relief (Hove, 2008) presents a promising avenue for theoretical exploration, as it builds on the idea that systems grow in complexity and scope due to their inherent capacity to “reduce the costs and risks of communication” (Habermas, 1987, p. 183). These are issues that future research should explore further.

The author gratefully acknowledges the respondents for their valuable access throughout the empirical work of this paper. The author also acknowledges the anonymous reviewers, Roland Almqvist, Johan Graaf, Fredrik Svärdsten Nymans, Niklas Wällstedt, and colleagues at SBS for their valuable comments and insights throughout various phases of this paper.