Unpaid Work and the Governance of GDP Measurement
In what ways does GDP measurement impact international development policy and gender equality?
Few numbers are as ubiquitous in political and economic analysis as gross domestic product (GDP). It is enlisted to rank and compare national economies, it influences lending and investment decisions, and is often taken as a proxy for well-being. GDP and other indicators – from government debt and inflation to poverty statistics – are not only measurement tools but also framing devices. They make states and markets legible in particular ways that can never be fully neutral or objective (Mügge 2016). Internationally harmonized statistics are necessary for cross-country analysis and to have a sense of confidence in global policy-making and markets. This harmonization is valuable and commendable, but it comes with trade-offs. Statistical concepts designed for industrialized market economies do not necessarily fit neatly onto other kinds of socioeconomic settings – for example, low-income areas with high levels of subsistence and informal activity. This article gives a brief overview of the governance of GDP measurement. I zoom in on the politics underlying one particularly contentious issue area: the measurement of unpaid labor. While the international statistical system is a rather non-politicized field, the issue of unpaid labor is a rare example of statistics entering public debate due to its implications for international development policy and gender equality.
Why Do We Measure GDP the Way We Do?
The harmonization of GDP measurement around the world is an enormous, ongoing effort led by several international organizations and expert groups. GDP first emerged in a small number of industrialized countries in the 1940s and attracted the attention of policymakers through its role in economic planning during the Second World War (Kendrick 1970). Since then, it has become a global institution.
Although GDP grabs the most attention, it is only one of many indicators derived from the System of National Accounts (SNA). The SNA is an internationally harmonized ’set of recommendations on how to compile measures of economic activity in accordance with strict accounting conventions based on economic principles’ (ISWGNA 2008: 1). In other words, it is a framework for measuring the total economic activity of a country. The SNA was first published in 1953 and, since the disappearance of alternative systems in post-communist states, is now the only internationally-accepted standard (Herrera 2010: 18).
The revision, publication, and implementation of the SNA is a cooperative task between five international agencies: the United Nations Statistics Division (UNSD), the International Monetary Fund (IMF), the World Bank, the Organisation for Economic Co-operation and Development (OECD), and Eurostat. Since the early 1980s, statisticians from these organizations have worked together on the revision and global implementation of the SNA in the Intersecretariat Working Group on National Accounts (ISWGNA). Prior to the formation of the ISWGNA, the SNA was solely the task of the United Nations Statistical Office, later renamed the Statistics Division. All of this work is overseen by the United Nations Statistical Commission, the highest governing body of the international statistical system.
Unpaid Labor and the Production Boundary
The production boundary is a conceptual line drawn between economic and non-economic activity. In national accounts ‘what is defined as economic activity is, literally, anything deemed to sit inside a designated ‘production boundary’’ (Christophers 2011: 115). With some notable exceptions such as the inclusion of financial services, the production boundary has been one of the most consistent features of the SNA (Bos 2009: 40; Christophers: 2011). With regard to unpaid labor, this continuity has persisted in the face of contestation.
In SNA terminology, unpaid labor corresponds to the category of ‘own-account services.’ We might think of many of these services alternatively as housework, care work, or social reproduction. This includes activities such as childcare, cleaning, meal preparation, and care for the sick and elderly. Hoskyns and Rai (2007: 297) maintain that ‘[w]ithout unpaid services and their depletion being measured and valued, predictions are likely to be faulty, models inaccurate and development policies flawed’ (ibid.: 297). Since unpaid labor is disproportionately carried out by women, failing to measure it introduces a gender bias into economic data (ILO 2016; Miranda 2011: 6). The problem is particularly acute in developing countries, where the overall amount of time spent on unpaid work is comparatively high (ILO 2016:20).
These concerns are nothing new. In fact, they were a fundamental concern among early debates over national accounting principles. Margaret Reid (1934: preface, v), more than eighty years ago, warned that a singular focus on ’that part of our economic system which is organized on a price basis’ had blinded economists to productive work of the household, ‘our most important economic institution’. And by all indications, statisticians today are not blind to the problem. Yet, the inclusion of these activities inside the SNA production boundary (and hence in GDP) is off the table. To understand why this is the case, we can look at the revision process of the SNA.
Revising GDP and Justifying the Production Boundary
The SNA has undergone three official revisions, leading to publications in 1968, 1993, and 2008. The revision process of the 1993 SNA was especially important for the issue of unpaid labor. During this process, which began in the early 1980s with the creation of the ISWGNA, the production boundary was discussed several times in both expert group meetings and within the UN Statistical Commission. Ultimately, very little changed. The ISWGNA reconfirmed the previous definition, wherein some forms of non-market activities were included and others, encompassing various forms of ‘housework’, were excluded. But the 1993 SNA made these exclusions explicit for the first time, cementing this historical idea into a formal institution.
The SNA states that the biggest problem in determining the activities included in the production account is deciding how to treat ’…activities that produce goods or services that could have been supplied to others on the market but are actually retained by their producers for their own use’ (ISWGNA 2008: 6). One of these gray areas is own-account production, which includes activities such as subsistence farming. Another is own-account services, a category that includes ‘the preparation of meals, care and training of children, cleaning, repairs, etc.’ (ibid.). The 1993 SNA confirmed the inclusion of own-account goods, and added some activities (including water collection and repairs to buildings) inside the production boundary on these grounds (Harrison 2005: 150). Services, in contrast, were explicitly excluded.
Reports of the ISWGNA and the UN Statistical Commission during the 1993 revision process indicate two main discursive justifications for this exclusion. The first can be labeled the ‘third party criterion’. The third party criterion – or simply the market criterion – originates from Margaret Reid’s definition of household production as consisting of unpaid activities that could be delegated to a paid worker or replaced by market goods (Reid 1934: 11). By the 1970s, this criterion became a consensus among economic statisticians (Sakuma 2013: 5F56). The second justification is a distinction between non-market goods and non-market services – a distinction that is in many respects arbitrary but leaves no ambiguity about the status of these services.
The third party criterion is implicit in expert deliberation during the revision process. The topic of non-market activity was first put on the agenda of the Statistical Commission by ‘the delegations representing the developing countries’ in the early 1980s (UNSC 1983: 10). But a cautious attitude prevailed, as several delegates expressed ‘a strong resistance to losing sight of the transactions-oriented base’ of the SNA (UNSC 1981: 12-13). Ultimately, as the UN Statistical Commission acknowledges, ’[t]he distinction between what is considered to be subsistence output and what is not is essentially an arbitrary one. It reflects mainly the traditional limits of marketed output in developed countries’ (ibid.: 14).
The second justification is the distinction between household goods and services. In a 1987 meeting of the ISWGNA, members of the expert group acknowledged the difficulty of defining the production boundary: ’In general it was not felt possible to have a single succinct definition of the production boundary that would explain why some items were included and some excluded…’ (Harrison 2005: 148). To get around this ambiguity, the group decided ‘to give fairly general indications followed by specific lists of examples that would make clear where the boundary should be drawn’ (ibid.). Such a list appears in the SNA (ISWGNA 1993: 149):
- The cleaning, decoration and maintenance of the dwelling occupied by the household including small repairs of a kind usually carried out by tenants as well as owners;
- The cleaning, servicing and repair of household durables or other goods, including vehicles used for household purposes;
- The preparation and serving of meals;
- The care, training and instruction of children;
- The care of sick, infirm or old people;
- The transportation of members of the household or their goods
These activities were explicitly excluded, and remain so in the most recent version of the SNA. In making these choices, the members of the ISWGNA relied on historical precedent, noting that ‘the only extensions to the production boundary previously accepted are for the production of goods’ (Harrison 2005: 150). Services provided within the household, on the other hand, ‘are always immediately consumed by those producing them and therefore do not add to the pool of goods and services available for redistribution’ (ibid.).
The Agency of Experts
As several scholars have pointed out (e.g. Wood 1997, Waring 2003), neither of these lines of reasoning – the third-party criterion nor the goods/services criterion – are consistently applied. For instance, washing clothes or taking care of children can be (and frequently are) done by paid domestic workers and daycare centers (Wood 1997: 51). Likewise, there is little meaningful distinction between non-market goods and services in a subsistence household, as Waring (2003: 36) demonstrates with an example:
The woman goes to collect water. She uses some to wash dishes from the family evening meal (unproductive work) and the pots in which she previously cooked a little food for sale (informal work). Next, she goes to the nearby grove to collect bark for dye for materials to be woven for sale (informal work), which she mixes with half a bucket of water (informal work). She also collects some roots and leaves to make a herbal medicine for her child (inactivity). … She will also collect some dry wood to build the fire to boil the water to make both the medicine and the dye (active and inactive labour). All this time she will carry the baby on her back (inactive work).
Although the justifications for excluding unpaid labor do not hold up to this kind of close scrutiny, they are nonetheless perceived as necessary by national accountants. Drawing the line at household services is seen as avoiding a ‘slippery slope’ that would compromise the monetary transaction-based core of the SNA and threaten international comparability. The 1993 SNA states that this reluctance is explained by a combination of factors: the isolation of these activities from markets, the difficulty of estimating monetary values, and ’adverse effects…on the usefulness of the accounts for policy purposes and the analysis of markets and market disequilibria’ (ISWGNA 1993: 149). In short, unpaid labor would introduce an unacceptable degree of uncertainty into GDP.
This pragmatic discourse clashes with the arguments of feminist political economists who point to this as a blind spot in economic theory and policy. Yet, while statisticians are certainly aware of and engaged in these debates, the governance of GDP remains relatively insulated from external preferences and constraints. This insulation is possible due to the highly specialized and technical nature of economic statistics. Member states have delegated the methodological work to international organizations, which have in turn created specialized statistics departments. The definition and harmonization of GDP is separated by yet another level of governance through the establishment of the ISWGNA, a small and cohesive expert group.
From GDP figures to the Sustainable Development Goals and corruption indices, there is a growing recognition that numbers and rankings shape global politics in important ways. Increasingly, indicators and statistics have not only been used in quantitative research but studied on their own. Yet, the origins of these numbers and the governance of the international statistical system have largely been neglected. The example of unpaid services in GDP shows that statisticians, due to their expertise, possess a great deal of agency over global rules and standards for economic measurement. Yet, professional norms and shared ideas remain important drivers of stability in how economies are quantified.
Bos, F. (2009) The National Accounts as a Tool for Analysis and Policy: History, economic theory and data compilation issues. Munich Personal RePEc Archive. Available online: mpra.ub.uni- muenchen.de/23582/
Christophers, B. (2011) Making finance productive. Economy and Society, 40(1), pp. 112-140.
Harrison, A. (2005) The background to the 1993 revision of the System of National Accounts. Available online: unstats.un.org/unsd/sna1993/history/backgrd.pdf
Herrera, Y. (2010) Mirrors of the Economy: National accounts and international norms in Russia and beyond. Ithaca: Cornell University Press.
Hoskyns, C. and Rai, S.M. (2007) Recasting the Global Political Economy: Counting women’s unpaid work. New Political Economy, 12(3), pp. 297-317.
ILO. (2016) Women at Work: Trends 2016. Geneva: International Labour Organization.
Intersecretariat Working Group on National Accounts. (1993) System of National Accounts 1993. Washington, D.C.: Inter-Secretariat Working Group on National Accounts.
Intersecretariat Working Group on National Accounts. (2008) System of National Accounts 2008. Washington, D.C.: Inter-Secretariat Working Group on National Accounts.
Kendrick, J. (1970) The Historical Development of National-Income Accounts. History of Political Economy, 2(2), pp. 284-315.
Miranda, V. (2011) Cooking, Caring and Volunteering: Unpaid work around the world. OECD Social, Employment and Migration Working Papers, No. 116, Paris: OECD Publishing.
Mügge, D. (2016) Studying macroeconomic indicators as powerful ideas. Journal of European Public Policy, 23(3), pp. 410-427.
Reid, M. (1934) Economics of Household Production. New York: J. Wiley & Sons.
Sakuma, I. (2013) The Production Boundary Reconsidered. Review of Income and Wealth, 59(3), pp. 556 567.
Scott, J.C. (1998) Seeing like a State: How certain schemes to improve the human condition have failed.New Haven, CT: Yale University Press.
United Nations Statistical Commission. (1981b) Report on the 21st session. Available online: https://unstats.un.org/unsd/statcom/reports/
United Nations Statistical Commission. (1983) Report on the 22nd session. Available online: https://unstats.un.org/unsd/statcom/reports/
Ward, M. (2004) Quantifying the World: UN ideas and statistics. Bloomington: Indiana University Press.
Waring, M. (2003) Counting for something! Recognizing women’s contribution to the global economy through alternative accounting systems. Gender and Development, 11(1), pp. 35-43.