Assetization as a mode of techno-economic governance: Knowledge, education and personal data in the UN's System of National Accounts

Abstract

Assets are made through the configuration of technoscientific and political-economic (or techno-economic) relations, claims and practices; a process increasingly conceptualized as ‘assetization’. The UN’s System of National Accounts (SNA) – a set of national accounting standards – defines assets as ‘entities that must be owned by some unit, or units, and from which economic benefits are derived by their owner(s) by holding or using them over a period of time’. Accounting standards like the SNA are implicated in the construction of assets through their ‘extension of the asset boundary’, which happens periodically as accounting standards are revised and updated to better reflect changing business practices. Assetization, then, entails more than an analysis of the transformation of something into an asset, it can also be conceptualized as a mode of governance in which social actors change their world. To make this argument, I examine the SNA’s treatment of knowledge, education and personal data: respectively, redefined as an asset (e.g. intellectual property product); treated as a quasi-asset (e.g. human capital); and subject to continued debate (e.g. digital data). In exploring the SNA’s accounting standards, I show how assetization reconfigures the governance of knowledge, education and personal data, often in problematic ways.

Keywords:

• assetization
• governance
• research & development
• human capital
• digital personal data
• national accounting

Introduction

Periodically, the UN’s Statistical Division releases a revision of the System of National Accounts (SNA), a little known but important conceptual and measurement framework for calculating global and national economic activity (OECD, 2017). The SNA is a set of national accounting standards organized by the Intersecretariat Working Group on National Accounts (ISWGNA), consisting of the European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations and World Bank. Following their revision, these national accounting standards are gradually adopted around the world. Following the rollout of the SNA, statistical concepts and measurements are standardized across different countries, thereby underpinning economic decisions and policymaking across countries, institutions and businesses (Rassier, 2013). The revision of the SNA – first published in 1953 and then revised in 1960, 1964, 1968, 1993 and 2008 – both reflects changing political-economic dynamics (e.g. responding to changing corporate accounting practices) and impacts those same political-economic dynamics (e.g. through macroeconomic policy decisions) (Lequiller & Blades, 2014). Analytically, these SNA revisions play an important constitutive role in political-economic governance (Miller & Rose, 1990).

An important revision of SNA2008 was the change in the treatment of research and development (R&D) and creative works. Previously, R&D was treated as ‘intermediate consumption’ (see SNA, 1993), while SNA2008 specifically ‘extended’ the asset boundary to conceptualize R&D as ‘capital formation’ (StatCan, 2015). This was significant because it changed the understanding, measurement and treatment of R&D from something that is used once (i.e. as an intermediate good) to being an asset that is ‘used repeatedly or continuously in production processes for more than one year’ (SNA, 2008, p. 196). A similar change had happened in the SNA1993 with software. In 2008, then, knowledge was turned into asset, in accounting terms at least (Birch, 2017a): it was reframed as an investment category (rather than intermediate consumption), eventually changing the way R&D spending is understood by businesses and governments. As Haskel and Westlake (2018) note, these national accounting changes have significant economic effects: ‘In the United States, for example, the capitalization of software added about 1.1 percent to 1999 US GDP and R&D added 2.5 percent to 2012 GDP’ (p. 43). Consequently, this conceptual and measurement framework changes the way we understand something like investment in knowledge production and changes the way we understand the governance of knowledge production.

My contention in this paper is that the extension of the asset boundary in SNA2008 is an example of ‘assetization’, defined as the transformation of something into an asset (Birch, 2017a, 2017b; Birch & Muniesa, 2020). Almost anything can be transformed into an asset with the right techno-economic arrangements, and accounting standards are important to this process (Chiapello, 2019; Mennicken & Power, 2015). Changes in accountings standards are implicated in the assetization of knowledge represented by the SNA2008 extension of the asset boundary to R&D. As noted, there is a dual change here: first, something (e.g. knowledge) is being reconfigured as an asset; and second, how we then understand and act upon (i.e. govern) that something is being reconfigured as a result of our understanding and framing of it as an asset. Assetization, then, can be conceptualized as a mode of techno-economic governance that is defined by this transformation process (Miller & Rose, 1990). Even though social actors construct assets through epistemic definitions and concepts, which are central to the making of assets, these epistemic changes have a very real material life of their own, such that social actors are subject to new political-economic imperatives. The question that motivates this paper, then, is not so much how something gets turned into an asset, but rather how does this transformation change the governance of the thing being assetized?

I focus on the ways that the SNA (i.e. national accounting standard) has extended the asset boundary through the conceptualization of three different political-economic objects policymakers and others have found difficult to define: (1) knowledge, redefined as investment in R&D in the SNA2008; (2) education and training, subject to ongoing debate and redefinition as ‘human capital’ in and beyond SNA2008; and (3) personal data, currently under debate in the ongoing SNA2025 revision. Focusing on these three objects helps to highlight the anomalies and societal implications of assetization. I start the paper with a discussion of the recent critical and constructivist literature on assets and assetization, positioning it in relation to debates about ‘governing economic life’ (Miller & Rose, 1990). I then turn to the SNA in order to examine how knowledge, education and personal data are being defined and reframed as assets. I analyse how the definition and conceptualization of assets and the asset boundary in the SNA have changed over time, focusing on the 1993 and 2008 SNA reports and the forthcoming 2025 SNA revision. I finish by considering the governance implications of this extension of the asset boundary.

Assetization as governance

The term governance can be confusing and contradictory, especially when it comes to disentangling its use as an analytical term from its manifestation as an empirical phenomenon to be studied (Bevir, 2013; Pierre & Peters, 2020). I therefore start this analytical discussion from Bevir’s (2013) point that governance relates to ‘all processes of governing, whether undertaken by a government, market, or network’ (p. 1). In particular, I am interested in governance as a concern with what Miller and Rose (1990) called the governing of economic life. Drawing on Foucault’s notion of governmentality, Miller and Rose (1990) were interested in how ‘power is exercised today through a multitude of agencies and techniques’, not all of which are bounded by the ‘executives and bureaucracies of the formal organs of state’ (p. 1). Moreover, they argued that there is a need ‘to pay particular attention to the role accorded to “indirect” mechanisms for aligning economic, social, and personal conduct with socio-political objectives’ (p. 2). Thinking about governance as ‘governing’ provides a useful analytical lens and starting point for examining national accounting standards that are not developed by ‘formal organs of state’ but rather established through the coordination of multiple intergovernmental agencies which draw on a range of experts and expertise to establish acceptable and credible accounting standards. My particular focus in this paper is on the configuration of knowledge claims, techniques, devices and suchlike that constitute the governance of the political-economic objects we call ‘assets’.

Assets have come under increasing analytical focus recently. The last few years have witnessed a growing intellectual interest in assets and especially in the process of assetization, defined as the transformation of things into assets. Much of this intellectual interest has been driven by critical and constructivist scholars examining the technical and political-economic intricacies of the assetization process, the asset form and the asset condition (Birch, 2015; Birch, 2017b; Birch & Muniesa, 2020; Muniesa et al., 2017). Building on an intellectual lineage stretching back over a century (see Birch & Muniesa, 2020; Muniesa et al., 2017; Nitzan & Bichler, 2009), this literature includes scholars in science and technology studies (e.g. Birch, 2017a, 2017b; Birch & Muniesa, 2020; Birch & Tyfield, 2013; Delvenne, 2021; Muniesa et al., 2017), critical data studies (e.g. Birch et al., 2021), feminism (e.g. Strauss, 2023), geography (e.g. Birch & Ward, 2022; Bridge et al., 2020; Ouma, 2020), political economy (e.g. Langley, 2021; Ward & Swyngedouw, 2018), socio-legal studies (e.g. Dreyfuss & Frankel, 2015; Pistor, 2019), political ecology (e.g. Ducastel & Anseeuw, 2017), sociology (e.g. Chiapello, 2019; Tellmann, 2020, 2022), history (e.g. Levy, 2017), education studies (Komljenovic, 2021), business and management studies (e.g. Geiger & Gross, 2021) and philosophy (e.g. Feher, 2018). As this literature demonstrates, there is an increasing research programme examining the ‘how’ of assetization; that is, how very different things are turned into assets with the right techno-economic arrangements.

Assetization also reconfigures how we frame, manage and govern societal resources. As noted above, my argument is that there is a dual transformation happening with assetization. While a ‘thing’ can be turned into an asset through the reconfiguration of the techno-economic arrangements that define something as a specific object or entity, and accounting is central to this process (Kurunmäki, 1999; Meyer, 1973), such a reconfiguration also changes how we understand, manage and govern that thing. Assetization, in this sense, is also a mode of techno-economic governance that emerges from the very transformation of a thing into an asset: that is, governance through assetization. As a mode of techno-economic governance, assetization entails four important shifts in the governance of societal resources (as assets). First, we need to understand the role of metaphors and narratives in the identification of things to be assetized; second, we need to examine how assets are defined and constructed; third, we need to unpack the temporal expectations that adhere to assets; and finally, we need to analyse the relational constitution of assets.

First, metaphors and narratives are central to an understanding of governance through assetization, reflecting the importance of language and discourse (see Miller & Rose, 1990). Chiapello (2015, 2019) highlights how metaphors frame things as capital – e.g. ‘human’, ‘natural’, etc. (see Komljenovic, 2021; Levidow, 2020; Levy, 2017). Metaphors like human capital can influence accounting frameworks when it comes to what constitutes an investment in and a return on education. Sociologists like Beckert (2016) have similarly stressed the need to analyse the symbolic value of tradable goods and services; here, Beckert argues that symbolic value is performative in that ‘fictional’ expectations configure things as economic entities (e.g. assets) in particular ways. Others have focused on the moral orders underlying assetization, including Ouma (2020) whose work on farmland investment highlights the morally contested nature of assetization. As noted by Power (1992), the deployment of metaphors and narratives reflects the establishment of ‘social credibility’ by experts making technical claims: he stressed that ‘the “how” of asset measurement and recognition is inextricably linked to the status of “who” does it’ (p. 40). For a thing to be assetized, it must be discursively framed and understood as having the potential to become an asset through the deployment of metaphors and narratives by credible experts to identify it as an object that can be governed in a particular way; for example, education can be identified as an investment by using the metaphor of ‘human capital’ (Gregory & Sadowski, 2021).

Second, thinking about governance through assetization entails analysing assets as techno-economic constructions (Birch & Muniesa, 2020; also Pistor, 2019). A thing has to be defined, bounded and constructed as an asset, which requires the delineation and enforcement of a ‘boundary’ that distinguishes it from other objects (Mazzucato, 2018; see also Meyer, 1973; and Kurunmäki, 1999, on the construction of ‘entity boundaries’). Such boundaries are effects of knowledge claims (Gieryn, 1999), especially accounting knowledge. As several scholars note (e.g. Chiapello, 2015; MacKenzie, 2009; Mennicken & Power, 2015; Power, 1992), accounting has this constitutive or performative character to it, in that it does not simply describe the world. Rather, it is implicated in the transformation of the world to reflect the assumptions of the knowledge claim itself (Callon, 1998). Accounting, moreover, configures value and valuation; by this, I mean the legal and political-economic operations that create rights to revenues and the recursive governance of those operations through the enforcement of those rights (Cherizola, 2021). As Muniesa (2012) argues, how we value something happens simultaneously with how we construct the thing we are valuing. Hence, how we construct assets configures the ways we govern societal resources, and the fact we value things in certain ways (e.g. as future revenues) ends up configuring how we understand the governance of those societal resources (Muniesa & Doganova, 2020). Consequently, assetization entails developing a particular way of ‘seeing’ things as if they are financial investments (in assets) and, consequently, governing them as such – what Muniesa et al. (2017) call the ‘asset condition’.

Third, assets are defined by the temporal expectations that define them and define the governance of them (Tellmann, 2020, 2022). Accounting definitions like the International Accounting Standards (IAS) characterize assets as ‘a resource controlled by the entity as a result of past events and from which future economic benefits are to flow to the entity’ (quoted in Birch & Muniesa, 2020, p. 3). An asset is constructed and governed as an expectation and entitlement to a flow of future income, often through the deployment of techno-economic mechanisms to smooth out any lumps or bumps in that monetary flow over time (Leyshon & Thrift, 2007) or to align social actors across different geographies and temporalities (Miller & Rose, 1990). Governance through assetization entails an entitlement or claim to future income, which depends upon the enforcement of that entitlement or claim through time so it can have value and be valued (Cherizola, 2021): the enforcement of the claim makes the asset valuable. As Pistor (2019) notes, assets create privileges for their holders through their ‘durability’; that is, the extension of entitlements or claims over time. This means that future revenue expectations have to be made durable as entitlements, which entails a ‘durational’ temporal management and governance in Tellmann’s (2020) framing. Assets, then, are claims on the future that end up defining the governance of societal resources; that is, controlling or owning an asset is tied into enforcing a political-economic configuration that ensures an asset generates the revenues the asset owner/controller can collect (Adkins et al., 2020). This, necessarily, locks in social actors to meeting the expectations of owners/controllers, whether the former wish to be bound by those expectations or not (Dreyfuss & Frankel, 2015).

Finally, assetization does not happen in isolation. Political-economic knowledge claims (e.g. ‘the value of x is y’) and their performative implications (e.g. ‘go buy x’) are relational and situated. As an example, value and valuation are constituted by knowledge claims that reflect a set of collective expectations, entitlements and organizational processes (Birch, 2017b; Mennicken & Power, 2015; Muniesa et al., 2017; Styhre, 2015). Working out the value of an asset by doing valuation is relational. As historian Levy (2017) notes, a rethinking of asset value happened at the end of the nineteenth century in which social actors debated how to do valuation. Their answer was that ‘to capitalize the expected pecuniary income streams of the newly consolidated asset’ required it be ‘discounted against a uniform market interest rate’, which was ‘something historically novel, being made possible by the recent geographical integration of capital markets’ (Levy, 2017, p. 498). An asset’s value is always constituted by its relation to other assets and a generally expected long-running interest rate to which future expected returns on investments can be discounted (Doganova & Muniesa, 2015; Piketty, 2014). Asset values are, then, necessarily governed through the management of an array of social actors enrolled in this wider techno-economic ecosystem (Miller & Rose, 1990), including analysts, investors, financiers, market-makers, news media and so on (Horvath & Klinkmüller, 2019). As such, governance through assetization entails a societal transformation in the understanding of societal resources for them to accrue and retain value, leading to the entrenchment of particular social interests in the governing of the economy.

Assetization as a mode of techno-economic governance: Analysing the UN’s System of National Accounts (SNA)

Next, I examine how the SNA defines and conceptualizes assets in order to analyse assetization as a mode of techno-economic governance, especially in the treatment of knowledge, education and personal information as resources governable as assets (i.e. R&D, human capital and digital data).

The SNA and the asset boundary

The SNA has its origins in the post-World War II construction of an international liberal order, being established in 1947 by the UN’s Statistical Commission to account for national gross domestic product (GDP). Vanoli (2005) provides a more detailed history of national accounting and the early years of the SNA. The SNA was developed and promulgated by the UN’s Statistics Division, forming an ‘internationally agreed standard set of recommendations on how to compile measures of economic activity’.¹ The initial SNA was produced in 1953 and has subsequently been revised in 1960, 1964, 1968, 1993 and the latest iteration in 2008. A further revision is planned for 2025. At present, and at least since the 1993 revision, it is organized and managed by an ISWGNA representing the UN, IMF, World Bank, OECD and European Commission. Although the SNA standards are not mandatory, they provide the basis for many countries and international institutions to coordinate their national accounting practices and economic forecasts, including revising them in light of changing concepts and definitions.

Although concerned with national accounting, the SNA standards reflect business accounting practices, which feed into the revisions. Although there are differences between national and business accounting (OECD, 2017), the SNA standards play an important role in shaping how both governments and businesses conceptualize, understand and govern economic activity (Vanoli, 2005). For example, the similarities between business and national accounting are evident in their definition of assets. In business accounting terms, an asset is defined by IAS as:

… a resource controlled by an entity as a result of past events and from which future economic benefits are expected to flow to the entity.

And in national accounting terms, an asset is defined by the SNA (2008, p. 42) as:

… a store of value representing a benefit or series of benefits accruing to the economic owner by holding or using the entity over a period of time.

Both definitions conceptualize assets as a resource/entity that generates future benefits for its owner/controller over a period of time. Critical to these definitions is that assets are both revenues streams and claims on those revenue streams, which can be capitalized by the owner/controller. Assetization as governance rests on this dual quality. This has been theorized by Birch and Tyfield (2013), Birch (2017b), Pistor (2019) and Birch and Muniesa (2020) in their characterization of assets as legal constructs that are made through the configuration of techno-economic claims, practices, etc. Revenues streams have to be made, just as claims on those revenue streams have to be constructed; this entails the construction of boundaries to distinguish political-economic objects (e.g. asset) and entities (e.g. organization) from one another and underlying processes (e.g. production) (see Kurunmäki, 1999), but without necessarily entailing a separation of the revenue streams and claims on them from their constitutive techno-economic relations and contexts. This contrasts with commodities (see Braun, 2020), and highlights the importance of thinking through assetization as a distinct transformation process and form of governance. Castree (2003) notes that a commodity can be physically and morally separated from its producers and ‘supporting context’, such that the commodity form reflects a homogenized process meaning that it does not matter who produces it. This is not necessarily the case with assets, which may be valued precisely because of who produces them (e.g. music), their context (e.g. infrastructure), and their qualities (e.g. wine variety), distinguishing them from one another.

As well as a conceptual process for standardizing national accounting terms, definitions and rules, the SNA is also political-normative in that it provides ‘the building blocks of macroeconomic statistics forming a basis for economic analysis and policy formulation’.² As such, the SNA has an explicit role in the governing of economic life as theorized by Miller and Rose (1990), being part of the ‘vast statistical apparatus’ they argued underpins the ‘installation of a new set of concepts by which to think of “the economy” as an economy’ (p. 12). In particular, the SNA plays an important role in defining and measuring the boundaries of economic activity, including the identification of what counts as an asset (Gieryn, 1999; Mazzucato, 2018). For example, the earliest SNA edition ‘confined’ discussion of assets to tangible assets in capital formation (SNA, 1953).

In accounting, boundary-making is a fundamental concern (Gieryn, 1999; see Mennicken & Power, 2015; Power, 1992). As Mazzucato (2018) argues, the way we define the ‘production boundary’, for example, around economic activity largely determines the way we understand, measure and make legible notions of value and constitute valuation practices of different social activities. Alongside a production boundary, the SNA2008 also defines an ‘asset boundary’; that is, an accounting designation to distinguish between intermediate goods and services (which end up accounted for in final production) from investments in things which generate future benefits (i.e. assets) (BEA, 2015). Assets are made through the development and deployment of accounting concepts, definitions and measurements (Birch, 2017a; Birch & Muniesa, 2020; MacKenzie, 2009; Mennicken & Power, 2015; Power, 1992); notably, these can and do change over time through an extension of the asset boundary, defined as:

The boundary line between those products that are retained in the economy and are used for consumption and those products that are used for capital formation is known as the asset boundary. (SNA, 2008, p. 198)

The distinguishing feature of an asset, here, is that it is a ‘good or service’ that gets used in production over a specific temporal period, namely beyond a single year. Because these goods and services are not used up in production, the issue of ownership/control becomes particularly pertinent, since it determines who can claim future revenues (see Pistor, 2019).

The significance of ownership/control is evident in a subtle shift in the language and discursive framing of the asset boundary between SNA1993 and SNA2008. Reflecting Power’s (1992) point about the importance of language and discourse in accounting, this shift illustrates how metaphors, narratives and expectations help to constitute assets (see Beckert, 2016; Chiapello, 2015, 2019). In the SNA1993, for example, ownership/control was important for defining the asset boundary, as shown below:

However, the ownership criterion is important for determining which naturally occurring – i.e. non-produced – assets are included in the System. Naturally occurring assets such as land, mineral deposits, fuel reserves, uncultivated forests or other vegetation and wild animals are included in the balance sheets provided that institutional units are exercising effective ownership rights over them. (SNA, 1993, p. 7)

Similarly, in SNA2008, ownership/control is important for defining assets, but the language changes. Instead of using the term ‘asset’, SNA2008 refers to ‘resources’ that can be turned into assets with suitable techno-economic arrangements (e.g. ownership claims). This reframing highlights the importance of the analytical lens we use to understand resources and how this lens plays an integral part in the transformation and governance of something as a political-economic object (e.g. asset). Consequently, this reframing affords the possibility of ‘extending’ the asset boundary rather than treating an object’s status (as an asset) as an essential quality:

However, the ownership criterion is important for determining which natural resources are treated as assets in the SNA. Natural resources such as land, mineral deposits, fuel reserves, uncultivated forests or other vegetation and wild animals are included in the balance sheets provided that institutional units are exercising effective ownership rights over them, that is, are actually in a position to be able to benefit from them. (SNA, 2008, p. 7)

Changing the asset boundary in SNA2008 illustrates the importance of conceptualizing the ‘making’ of assets undertaken in assetization studies (e.g. Birch & Muniesa, 2020), rather than assuming that assets have inherent or essential qualities. While SNA1993 used the terminology of ‘natural asset’ throughout, SNA2008 uses the terminology of ‘natural resource’. In defining ‘nature’ as a resource, SNA2008 illustrates how an asset is constituted – made legible and measurable – and governed in and through accounting concepts, metrics and practices; that is, it entails the techno-economic configuration and governance of something as an asset (e.g. property rights, future revenues, technology, relative prices, etc.).

SNA2008: Extending the asset boundary to include R&D

Although there are differences between the definition and conceptualization of assets and the asset boundary in SNA1993 and SNA2008, the major change implemented in 2008 was the extension of the asset boundary to include research and development (R&D) and creative works (i.e. ‘knowledge’).³ This change was gradually implemented in different countries: for example, 2013 in the United States and 2014 in the United Kingdom. Prior to 2008, R&D and creative works had been treated as intermediate spending or consumption in the production process. As SNA2008 (p. 8) notes:

The distinction between intermediate consumption and gross capital formation [i.e. creation of an asset] depends on whether the goods and services involved are completely used up in the accounting period or not. If they are, the use of them is a current transaction recorded as intermediate consumption; if not it is an accumulation transaction recorded in the capital account.

As this SNA2008 indicates, the difference between intermediate consumption and capital formation is that the latter reflects a temporal distinction between immediate use and use in production ‘for more than one year’. The SNA2008 change was meant to address a conceptual gap identified and recognized as a problem in SNA1993 (UN Statistics Division, 2006), but which was largely considered to be a technical issue to do with measurement, which could not be resolved at the time. For instance, SNA1993 stated that:

Research and development are undertaken with the objective of improving efficiency or productivity or deriving other future benefits so that they are inherently investment – rather than consumption – type activities … In order to classify such activities as investment type it would be necessary to have clear criteria for delineating them from other activities, to be able to identify and classify the assets produced, to be able to value such assets in an economically meaningful way and to know the rate at which they depreciate over time. In practice it is difficult to meet all these requirements. (pp. 179-180, emphasis added)

Notably, R&D was characterized as ‘inherently investment’, meaning that R&D should be treated as the construction of assets rather than intermediate goods and services. However, it was not treated as such. This section of SNA1993 highlighted the difficulty in identifying the criteria to characterize R&D as a produced asset. In particular, the inability to value the resulting ‘asset’ and to work out how it depreciates over time posed significant accounting problems. As with other scholars who analyse accounting knowledge claims (e.g. Chiapello, 2015; MacKenzie, 2009; Mennicken & Power, 2015; Power, 1992), these changes from 1993 to 2008 illustrate how such accounting knowledge claims come to constitute and construct something as political-economic objects (e.g. assets), underpinned by certain understandings of value and valuation practices (e.g. depreciation, discounting) (Doganova & Muniesa, 2015).

The SNA2008 change was not uncontroversial, nor was it uncontested. Discussions about extending the asset boundary to R&D and creative works was undertaken by the Canberra II Group, which is part of the UN Statistics Division. Operating between 2003 and 2007, the Group’s mandate was to do the research needed to update SNA1993. An Issue Paper prepared by the Group’s Secretary Charles Aspden, who was from the OECD, for the Advisory Expert Group (AEG) of the SNA recommended that ‘1993 SNA should be changed to recognise the outputs of R&D as assets’ and, as a result, that the ‘definition of an asset should be reviewed to ensure it covers the assets of non-market producers adequately’ (IMF, 2005, p. 2). Despite these recommendations, however, the Issue Paper contains an appendix in which it notes that ‘there is a significant minority who feel that some R&D does not result in new assets’ (IMF, 2005, p. 26). Debate centred on whether R&D that is ‘disseminated freely’ should be considered an asset: at issue was whether such freely distributed outputs – knowledge as a collective and relational process (Fuller, 2002) – meet the ‘economic benefit’ criteria of an asset, since it is likely they would not be ownable/controllable and, therefore, difficult to accrue economic benefit from without intellectual property rights (Birch, 2017b; Kang, 2020).

A few writers have outlined the implications of these SNA2008 revisions (e.g. Birch, 2017a; Frase, 2013; Haskel & Westlake, 2018; Mazzucato, 2018), as have a number of governance organizations (e.g. OECD, 2010; Rassier, 2013). In particular, the OECD has produced a Handbook (2010) that outlines how to apply the SNA changes, such as how to value intangible assets using market values; for example, using similar assets that have been traded or as a summation of production costs if there are no similar assets (OECD, 2010, p. 18). While it might seem like an arcane national accounting issue, this change has had significant policy and political-economic impacts. For example, in August 2013 The Economist magazine pointed out that as a result of the SNA2008 change, ‘This week America’s GDP rose by $560 billion, or 3.6%, mainly because the ‘boundary’ that defines what counts as an economic asset was moved’ (The Economist, 2013).

SNA2008 and beyond: Ambiguities around extending the asset boundary to ‘human capital’

While the asset boundary was extended to include R&D and creative works in SNA2008, the same was not done for ‘human capital’ (see Moulton & Mayerhauser, 2015). Human capital was specifically excluded from the extension to R&D and creative works (SNA, 2008, p. 206), but it was incorporated through so-called ‘Satellite Accounts’ covering issues outside the general system. Human capital is interesting because of this specific exclusion. SNA2008 noted that, ‘It is difficult to envisage ‘ownership rights’ in connection with people, and even if this were sidestepped, the question of valuation is not very tractable’ (p. 43). While knowledge, as a collectively produced and legitimated process (Fuller, 2002; Tyfield, 2012), was redefined as an asset, the human activities entailed in training and education were not because they are not deemed ownable, being embodied in individual persons.

Yet, education and training are still framed as an asset, specifically in terms of the metaphorical denotation of ‘capital’ to indicate an investment rather than intermediate spending (Chiapello, 2019).⁴ Human capital, as theorized by neoclassical economists like Gary Becker, is premised on the idea that spending on personal education and training is an ‘investment’ that leads to future returns to its ‘owner’ (Cooper, 2017). The Canberra II Group Issue Paper mentioned above, and dealing with R&D, stated that:

This paper only deals with the question of whether R&D should be regarded as investment. The ISWGNA has decided already that there should be no consideration given in the update as to whether human capital should be classified as an asset in the system, and by implication staff training. (IMF, 2005, p. 3)

Earlier incarnations of the SNA had also addressed whether ‘human capital’ should be treated as an asset. For example, SNA1993 considered whether ‘education’ was an investment and therefore asset. While SNA1993 noted that education and training increase the ‘productive potential’ of individuals, they are ‘not produced’ but rather ‘acquired through learning’ and therefore cannot be considered as ‘processes of production’. Specifically ‘embodied’ in individuals, not institutional units, human capital cannot be sold on a market or traded between units – at least, that is, in most cases (e.g. sports players might be considered an asset here – see Nappert & Plante, 2023). Even defining the value of human capital in the SNA was deemed problematic, since using proxies like remuneration incorporates time and effort, not just a payment for the use of an asset (Moulton & Mayerhauser, 2015).

While not much changed between SNA1993 and SNA2008 when it came to defining and conceptualizing human capital, there was some acknowledgement of dissatisfaction: ‘This treatment of education costs is consistent with the production and asset boundaries of the SNA but not all users of the SNA find it satisfactory in all instances’ (SNA, 2008, p. 8). Alternative conventions for defining human capital were encouraged in the 2008 revision using ‘satellite accounts’, which are outlined in Chapter 29 at the end of the overall report (pp. 523-544). Satellite accounts are meant to provide some flexibility for enacting the SNA accounting standards – reflecting what STS scholars call ongoing boundary work (Gieryn, 1999). Sector D of this chapter outlines the ‘extension of the scope of assets’ entailed in deploying satellite accounts:

The borderline between intermediate consumption, final consumption and capital formation may also be modified in various ways. Two often mentioned cases refer to human capital and consumer durables. If at least part of final consumption on education and health were treated as fixed capital formation, the corresponding central framework transactions would be reclassified from consumption to fixed capital formation resulting in human capital assets. (SNA, 2008, p. 527)

As this quote illustrates, treating education consumption as the creation of an asset requires a reclassification of consumption, thereby extending the asset boundary. It is an example of how assets are constructed through the identification and recognition of techno-economic categories (e.g. asset boundaries) (Power, 1992).

The SNA discussion here is not about whether human capital should be treated in a particular way but that it can be treated in different ways – as already illustrated by the distinction between SNA1993 and SNA2008 with R&D. In a Guide on measuring human capital, produced by the UN Economic Commission for Europe at the request of European statistical agencies (UNECE, 2016), the authors highlight some of the ways to conceptualize and measure human capital as an asset. This Guide reflects on the definitions and concepts used in SNA2008, noting that there are several measurement approaches (and issues) with regards to human capital including: the scope of estimates, its heterogeneity and its aggregation. It goes on to emphasize that only returns that accrue to individuals are taken into consideration, and notes an inconsistency in SNA2008:

… human capital differs from the usual types of capital in that it is fully embodied in persons, and as such is an entity that cannot be sold as a separate item on the market. (UNECE, 2016, p. 12)

Human capital is, in this sense, acquired and embodied, rather than an alienable good or service (Birch & Muniesa, 2020). As such, it cannot be transferred to another person. Despite these attempts to think through human capital as an asset in the SNA, significant conceptual and definitional ambiguities remain unresolved.

SNA2025: Unpacking plans to extend the asset boundary to digital data

The ISWGNA that revises the SNA – with the assistance of an Advisory Expert Group on National Accounts (aka AEG) – is currently updating the SNA for the release of a full revision in 2025. As part of this revision, the ISWGNA is considering around 35 issues, which were identified in July 2020. Most of these issues focus on three substantive political-economic areas: digitalization, globalization and well-being and sustainability.⁵ Of these, digitalization represents an interesting case because of the growing importance placed on digital technologies and, especially, digital data by policymakers and others. There are frequent metaphorical allusions to digital data, especially digital personal data, as being the ‘new oil’ underpinning our economies; for example, in May 2017 The Economist magazine made the following claim on its frontpage: ‘The world’s most valuable resource is no longer oil, but data’ (The Economist, 2017). Consequently, a bevy of international policy organizations are engaging with the question of how to define and value digital data (e.g. OECD, 2022).

The ISWGNA started the revision process with the help of the AEG and is seeking input from various stakeholders before final publication.⁶ According to the ISWGNA, they are consulting widely through various outreach and consultation processes and events, trying to reach statistical experts, academics, policymakers and business. Many of these expert and consultation contributions and materials are already published on the SNA website, so it is possible to examine conceptual and definitional developments and differences. Under ‘digitalization’, the ISWGNA is considering issues ranging from ‘Framework for a satellite account of the digital economy’ – akin to the approach taken with human capital – to ‘Recording of data and valuation of free assets and free services’. The latter is further split between: (a) ‘Recording and valuation of data in National Accounts’ and (b) ‘Treatment of free digital assets and services’.⁷

Of particular interest to debates about assetization, especially of personal data (e.g. Beauvisage & Mellet, 2020; Birch et al., 2021), is the conceptualization and definition of digital data in this SNA2025 revision process. In 2021, the ISWGNA reported to the UN Economic and Social Council on their progress on a range of issues, noting the establishment of a task team to look at the ‘Role of data and the SNA asset boundary’ (UN ESC, 2021, p. 17). The conceptual discussion about the treatment of digital data as an asset is evident in the 14th Meeting of the AEG in October 2020. At this meeting, the ISWGNA Task Team on Digitalization presented a paper on the recording and valuing of data, making several recommendations and conceptual framings:

1. Data is the result of a production process.

2. For practical reasons, focus is on digital data. Non-digital data is out of scope.

3. Data is distinct from ‘observable phenomena’, which are input for data.

4. Observable phenomena are non-produced and in general have no value, except if purchased. They do not affect the core accounts.

5. ‘Long-lived’ data (i.e. those used in production for more than one year) is an asset and as such should be capitalized in the national accounts.

6. ‘Short-lived’ data also exist, is produced and have economic value … .

7. Data is subject to economic ownership, valuation and depreciation. (AEG, 2020, p. 1)

Here the SNA2025 discussions define digital data as an asset where it fits the temporal characteristics of the asset form (e.g. lifespan beyond one year), as well as fitting within the asset boundary defined by ownership, valuation and discounting (SNA2008). As part of the assetization of digital data (Birch & Muniesa, 2020), this conceptual definition of digital data – as the digital collection and storage of ‘observable phenomena’ – is being rolled-out and adopted in subsequent debates about the valuation of digital data (e.g. OECD, 2022). This reframing of digital data as an asset is distinct from the same process with R&D and human capital; in particular, there is a clearer emphasis on the relational dimensions of data. As scholars have noted, assetization is a collective or relational process in which an asset, and especially its valuation, are always contextual, specifically relating to other assets within a techno-economic system (Styhre, 2015). For example:

Ignoring the very rare occurrences of a single valuable observable phenomenon, most individual observable phenomena by themselves do not represent a ‘store of value representing a benefit or series of benefits accruing to the economic owner by holding or using the entity over a period of time’. … For all these reasons, single observable phenomena do not meet the definition of an asset. (AEG, 2020, p. 18)

As such, ‘data’ are an asset where it entails the collection and recording of multiple observable phenomena (e.g. user data). Here, data are very much relational (Viljoen, 2021), or at least its value is defined by its relationality (e.g. data analytics requiring population data).

Implications of assetization as a mode of techno-economic governance

The SNA process has sought or is seeking to conceptualize and reframe knowledge production (i.e. R&D), education (i.e. human capital), and personal information (i.e. digital data) as an asset. This SNA process has not only contributed to the transformation of different things into an asset, it also changes the ways we govern these things (i.e. knowledge, education and personal information).

First, there are implications to using metaphors and narratives to redefine something like education as an investment, and hence ‘capital’. Both SNA1993 and SNA2008 emphasize that it is difficult to treat human capital as an asset for several reasons. As Birch and Muniesa (2020) note, assets are not the consequence of some inherent or embodied quality. If we want to consider human capital as an asset, then this necessitates a metaphorical or rhetorical move and the deployment of social credibility to define and thereby constitute something else – e.g. education – as an object of economic governance (Miller & Rose, 1990; Power, 1992). As a metaphor for education and training, however, human capital obscures the fact that it is the education and training providers who should really be considered as asset owners since they sell access to education and training to individuals (cf. individuals as investors in themselves as assets – see Feher, 2018). This is evident in the booming world of education technology, which is transforming a range of educational practices into assets (Komljenovic, 2021). We may be focusing on the ‘wrong’ social actors (e.g. learners) when it comes to governing educational accessibility and inclusivity; rather, we need to pay more attention to what is actually being turned into assets.

Second, how assets are made entails not only a conceptual orientation towards certain notions of ownership but also a need to restructure governance to then constitute that expectation: in particular, such restructuring necessitates new means for measuring, comparing, and ranking the asset performance of previously non-economic objects or entities (Mennicken & Power, 2015). On the one hand, an asset brings ‘economic benefits’ to their owners/controllers, which is not always measurable with knowledge, education or personal information. On the other hand, an asset is an ‘economic’ ownership right (i.e. claim on revenues), which can be separated from the thing under consideration (e.g. knowledge) and its embodiment (e.g. as a collective and common resource). Hence, the SNA2008 notes that it is not possible to define human capital as an asset because ‘It is difficult to envisage ‘ownership rights’ in connection with people’ (SNA, 2008, p. 43). Transforming knowledge or education into intellectual property or human capital respectively, even metaphorically (Chiapello, 2019; Power, 1992), requires that we contractually limit a person’s capacity to work; for example, through an exclusive license like intellectual property (Kang, 2020), or work contract (Nappert & Plante, 2023). Consequently, governance through assetization further centres contractual arrangements in political-economic life, reinforcing an increasing emphasis on transactional relations between social actors to the detriment of collective political engagement in pursuit of shared goals.

Third, assets are defined by their temporality (Beckert, 2016; Muniesa & Doganova, 2020; Tellmann, 2020, 2022). An asset is a claim on future revenues resulting from past actions and valued through temporal discounting (Mennicken & Power, 2015). This is why Pistor (2019) notes that the ‘durability’ of claims is a defining feature of assets. The SNA’s conceptualization of digital data, for example, differentiates between ‘long-lived’ data (an asset) and ‘short-lived’ data (not an asset) (AEG, 2020). Configuring something as an asset necessitates the creation of a temporal ‘revenue structure’, which entails the roll-out of ‘devices of obligation’ according to Tellmann (2022, p. 2). These devices tie revenues to the claims over a temporal duration and represent an important shift in the governance of political-economic objects; specifically, an asset is an investment, meaning it falls under international investment law rather than trade law. Consequently, assets are governed by adjudication mechanisms like investor-state dispute settlement panels, which are focused on protecting investor expectations against direct or indirect ‘expropriations’ (Dreyfuss & Frankel, 2015; Horvath & Klinkmüller, 2019). Governance through assetization can entail a temporal lock-in to investor logics, in which policy and political changes are deemed illegitimate if they impact expected future revenues. To date, digital data is not considered an asset or investment (Horvath & Klinkmüller, 2019), but as the SNA2025 process unfolds this could change with potentially problematic consequences for states’ enforcement of data protection and privacy rights.

Finally, the relational dimension of the asset form further complicates the relationship between how things are reconfigured and governed as assets. As noted, things like knowledge, education and personal information are all relationally constituted: for example, knowledge is collectively produced, constituted and legitimated (Fuller, 2002); education is relationally organized (Komljenovic & Robertson, 2016); and personal information is relational (Viljoen, 2021). This relationality is best illustrated with personal information in the debates about digitalization in SNA2025. For example, the AEG discussion about digital data ends with an important question for further consideration by others in the SNA process: ‘Do you agree that data belong to their producer (unless they are sold/licensed) and not to the persons/households to which the underlying observable phenomena refer to?’ (AEG, 2020, p. 19). This question points to an issue with personal information being something that is collected, recorded and stored, and thereby holding value, versus it being something that the person it is about having a legitimate claim to (e.g. privacy rights, property rights). Although personal information is currently not claimable as property does not mean that it cannot be governed as such with the right techno-economic arrangement, like contractual agreements (Horvath & Klinkmüller, 2019). Being relational, personal information can provide ‘economic benefits’ but only where it is made excludable so others cannot access it, despite its origins in those very relationships that are then occluded; this is a point emphasized by the OECD (2022, p. 10) in their discussion about measuring the value of data.

Conclusion

I have analysed the extension of the so-called asset boundary in the System of National Accounts (SNA) as part of the assetization of knowledge, education and personal information: that is, their transformation into intellectual property products, human capital and digital data respectively. This transformation is not uncontested – as illustrated by debates about human capital – and is also ongoing in the case of digital data. Building on the interdisciplinary literature examining assetization (e.g. Birch & Muniesa, 2020), I outlined how changes in accounting standards and frameworks like the SNA represent part of the techno-economic configuration of new asset classes (Chiapello, 2019). My aim was to unpack the anomalies, contradictions and societal implications of assetization, illustrating how the redefinition and reframing of things as assets can lead to problematic societal outcomes; for example, the importance of an investor logic replacing other social values in the governance of societal resources (Muniesa et al., 2017).

Assetization is a dual transformation. First, something is being transformed into an asset through the reconfiguration of the techno-economic arrangements that constitute assets, including: the narratives and metaphors that frame things as political-economic objects; the configuration of devices, practices and boundaries that constitute revenue streams and claims too them; the temporal dimensions to the asset form; and the relationality of those assets with broader economic systems. Second, how we understand, organize and manage that transformed thing – now an asset – also changes as a result of rethinking how we should govern societal resources. Although assetization has primarily been deployed conceptually to study how things are turned into assets, it also provides a useful political-normative approach for understanding the governance of technoscience and political economy: that is, governance through assetization. Muniesa et al. (2017) call this political dimension of assets and assetization the ‘asset condition’, which concerns the configuration of the political-economic future. An asset has to be ‘delineated’ by their techno-economic boundary and constituted by legal claims, technologies and materialities, and an ‘attributable scope’ (see also Pistor, 2019). Most importantly, an asset has an economic value defined by its future revenues and often entailing the capitalization of those future revenues through various discounting practices; as a result, an investor logic comes to form a new common sense for how to govern our economic lives.

The specific case of the SNA highlights a number of important governance and analytical implications. From a governance standpoint, the SNA changes are not simply conceptual or definitional adjustments in an ongoing statistical process of tweaking national accounting practices so they become more accurate over time. Changes to the SNA have real world impacts, although whether they are performative or not is a question for future research. For example, Haskel and Westlake (2018) highlight the political-economic impacts these SNA changes have, especially in terms of increasing national GDP. Whether something is configured as an asset or not has real import when it comes to governing the economy (Miller & Rose, 1990). Assets have become central political-economic objects in many societies (Adkins et al., 2020; Birch, 2015), suggesting that assetization underpins an emerging political-economic regime. My analysis provides one illustration of the implications this has for our lives. In particular, the SNA2008 change in treatment of R&D is transforming how social actors understand and manage innovation: technoscientific knowledge is treated and governed as an asset, entailing the insertion of investment logics and calculations into R&D operations (Birch, 2017b). As I argue elsewhere, this impacts how R&D is financed and the rationale for R&D, promoting innovation that reinforces ownership/control rights and the generation of economic rents (e.g. digital rights management) (Birch & Muniesa, 2020). Extending the asset boundary to education and training and personal information would have similar impacts.

Analytically, the SNA raises questions about how we govern societal resources as assets and how we organize a society of assets. Such questions help us understand assetization as a new mode of techno-economic governance. In theorizing assets as claims to future revenues, we examine how those revenues are constituted by the particular techno-economic configuration of organizations, practices, knowledges and devices, which has a temporal duration to it to ensure those future revenues happen and which is secured by techno-legal limits on challenges or contestations of those initial claims. I am highlighting all of this to emphasize that assetization can refer, like governance (Bevir, 2013), to both an analytical approach and an empirical phenomenon, as well as a political project; that is, as a way to intervene in the highly contingent practices, claims and devices that constitute the transformation of things into assets (Langley, 2021). Consequently, there is a growing need to examine assetization as a mode of techno-economic governance for organizing our societies, requiring, for example, more analyses of the implications underlying the specific temporal coordination and organization of the economy (i.e. focused on the future) (see Beckert, 2016; Tellmann, 2022); of the financial framing of social actions and subjectivities, such as a focus on housing ownership (Adkins et al., 2020; Birch, 2015) or human capital (Cooper, 2017; Feher, 2018); and of the normative limitations imposed on political and policy decisions resulting from the dominance of investor logics and interests (Birch & Muniesa, 2020; Dreyfuss & Frankel, 2015). A final call might be to think about how we can step back from the asset condition to find ways to manage our societal resources outside of the asset boundary; this would necessitate a rethinking of what resources, assets and investments are and what they are for.

Acknowledgements

My thanks to the organizers and participants at the following events: Perspectives on the Fourth Industrial Revolution Workshop, KAIST, South Korea (2017); European Association for the Study of Science and Technology Conference and 6^th Changing Political Economy of Research and Innovation Workshop, both Lancaster, United Kingdom (2018); Society for Social Studies of Science Conference, Sydney, Australia (2018); Technoscientific Constitutionalism Joint DFG-NSF Workshop, Washington DC, United States (2019); The Assetization of Work: Varieties of Human Capital Workshop, University of Sydney, Australia (2019); Geography Seminar Series, University of Newcastle, Australia (2019); Science and Technology Policy Webinar, University of Athens, Greece (2020); and Department of Science & Technology Studies Seminar Series, University of Vienna, Austria (2021). My thanks also to the guest editors (especially Veit Braun), reviewers, journal editors and journal manager for their comments and suggestions. Usual disclaimers apply.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Social Sciences and Humanities Research Council (SSHRC) of Canada under Grant [Ref. 435-2018-1136].

Notes on contributors

Kean Birch

Kean Birch is Director of the Institute for Technoscience & Society and Professor in the Graduate Program of Science & Technology Studies at York University, Canada. He has a new book coming out with Palgrave Macmillan called Data enclaves.

Notes

1 UN Statistics Division, https://unstats.un.org/UNSD/nationalaccount/sna.asp

2 UN Statistics Division, https://unstats.un.org/UNSD/nationalaccount/sna.asp

3 This type of extension had happened before as SNA committees debated whether one thing or another was best conceptualized as an asset or not. Haskel and Westlake (2018, p. 43), for example, outline how SNA1993 had extended the asset boundary to include spending on software, which was then subsequently enacted in the EU (1995), United Kingdom (1998) and elsewhere.

4 A related, but tangential, issue here is the assetization of education through student loans, although I do not have space to go into in this paper (see Milyaeva & Neyland, 2020).

5 See here: https://unstats.un.org/unsd/nationalaccount/update_Issues.asp

6 See here: https://unstats.un.org/unsd/nationalaccount/uProc.asp

7 See here: https://unstats.un.org/unsd/nationalaccount/aeg/2020/M14.asp