Chapter 9. Measuring sustainability

Marleen De Smedt
Enrico Giovannini
Walter J. Radermacher

This chapter outlines the principles of the capital approach and of the systems approach to measuring sustainable development. In the capital approach, human, social, natural, and economic capital are considered separately, with indicators presented on their stocks and how they change over time. While significant progress has been achieved in operationalising this approach to sustainability, this approach, argue the authors, implicitly assumes the independence of these stocks, and do not easily lend itself to considering interactions between different parts of the systems that underpin human well-being and functioning ecosystems. The chapter considers how the systems approach should be taken forward to move from theoretical considerations to empirical applications. It explains the key notions underpinning the systems approach, including risk, vulnerability and resilience, arguing that sustainability remains the ultimate objective. The chapter proposes a measurement agenda, suggesting steps to improve consideration of economic, human, and natural capital in the capital approach; and to improve the measurement of resilience and other aspects of the systems approach.


Marleen De Smedt is former Adviser to the Director-General of Eurostat, Enrico Giovannini is Full professor of Economic Statistics at the Department of Economics and Finance (DEF) of the University of Rome “Tor Vergata”, and Walter J. Radermacher is Professor at the University of Rome “La Sapienza”, and former Director-General of Eurostat. The authors wish to thank Paul Schreyer as well as other members of the HLEG for their comments on previous drafts of this chapter. They also thank the participants to the two HLEG workshops held in Rome, Italy, organised in collaboration with the OECD, on: “Intra-generational and Inter-generational Sustainability” on 22-23 September 2014, hosted by Einaudi Institute for Economics and Finance and the Bank of Italy, and sponsored by SAS; and on “Measuring Economic, Social and Environmental Resilience” on 25-26 November 2015, hosted by Einaudi Institute for Economics and Finance, supported by the Bank of Italy and Istat, and sponsored by SAS. Section 9.4 largely relies on Manca, Benczur and Giovannini (2017); therefore, the authors of this chapter are deeply grateful to Anna Manca and Peter Benczur for their contribution.

The opinions expressed and arguments employed in the contributions below are those of the author(s) and do not necessarily reflect the official views of the OECD or of the governments of its member countries.

9.1. Introduction

Ensuring that individual and societal well-being can last over time requires preserving the resources needed by future generations. This approach (which underpins the definition of “sustainable development” provided by the 1987 Brundtland Report) implies that measures of economic and social progress must take into account changes in those resources that last over time (i.e. capital). These changes consist of depreciation, depletion and erosion (all of which diminish capital) as well as investment, innovations and discoveries (all of which increase capital). Without them, any account of sustainability is incomplete.

The very nature of “capital” – as a cornerstone of economic production, as an entrepreneurial code of conduct and as a fundamental of bookkeeping and accounting – implies a close relationship between capital and sustainable development. Conceptually, capital can be broken down into economic, human and natural capital (social capital is typically added too, but it is not discussed here as it is addressed in Chapter 10)1. At the same time, an exclusive focus on the measurement of different types of capital – and the deeply rooted belief that, with enough effort, signals about the future can be internalised via the right valuation of capital – may divert attention from investigating the measurement of sustainability from different but complementary perspectives.

One example of these complementary perspectives is provided by the notion of “footprints”. Sustainable development has transboundary implications: the environment crosses country borders and, in pursuing the well-being of its own citizens, a country might affect the well-being of people in other countries. This is why UNECE insists that

“… [Sustainable Development Indicator] sets should reflect the transboundary impacts of sustainable development, by highlighting how a country in the pursuit of the well-being of its citizens may affect the well-being of citizens of other countries.” (UNECE, 2014 p. 14).   
So increasingly, pursuing sustainable development means considering not just individual countries but the entire globe. A capital approach allows for an integration of time, the “future”, but it is less useful for the accounting of place, the “elsewhere” dimension, which is equally important when it comes to the monitoring of sustainability of national activities.

Another complementary perspective is to look into “systems”, their behaviour over time and their inter-relationships. In this perspective, monitoring capital stocks is only part of the story: ultimately, the challenge is to manage – nationally and internationally, and over a long-time perspective – those complex and inter-dependent systems (economy, society and nature) that shape the well-being of future generations. Such systems are complex, meaning they behave differently at different scales; they behave in non-linear ways; they are often self-organising; and they are characterised by uncertainties, resilience, tipping points and irreversibility.

In a systems approach, the focus moves from measuring the stocks of assets to coming to grips with the resilience of economic, societal and natural systems. Tackling these issues requires inter-disciplinary work, with a focus on the ability of the system to cope with risks and uncertainties in a broad and long-run perspective, and on the different ways to manage this coping ability (resilience) of systems. “Resilience” is indeed referred to in the Sustainable Development Goals (SDGs) and by the targets of the 2030 Agenda.

As this chapter will show, despite a range of national and international initiatives (Box 9.1), measuring sustainable development still confronts difficulties and limitations. Good and effective communication to all stakeholders about available as well as missing information is critical, especially for making decision-making processes more effective and participatory.

This chapter is organised as follows. Sections 9.2 lays out the principles of the capital and of the systems approach. Section 9.3 describes progress achieved in measuring economic, human and natural capital since the Stiglitz, Sen and Fitoussi (2009) report. Section 9.4 dives deeper into the systems approach, which should be taken forward to move from theoretical considerations to empirical information. Section 9.5 concludes by drawing up a measurement agenda for the years to come.

Box 9.1. Monitoring sustainable development in practice: some examples

Various initiatives aimed at improving the measurement of progress and sustainable development have resulted in international agreements and standardisation, such as the adoption of the 2008 System of National Accounts (SNA), the G20 Data initiative, and the adoption of the System of Economic Environmental Accounting (SEEA) in 2012. With the adoption in 2015 of the 17 Sustainable Development Goals (SDGs, United Nations, 2015a) and 169 targets, and the later agreement on a set of 232 global indicators, much headway has been made towards a common set of sustainable development indicators at world level.

Some countries, however, have tracked sustainable development for several years, even before the adoption of the UN SDGs, either through a national set of indicators and/or in a supra/international context. Examples of such international developments include:

  • The Recommendations of the Conference of European Statisticians (CES), providing guidance on how sustainable development indicators (physical and monetary) might be harmonised and made consistent across countries and institutions (UNECE, 2014).

  • In the EU context, the European Statistics Sponsorship Group on Measuring Progress, Well-being and Sustainable Development (European Statistical System, 2011), the Sustainable Development Indicators set, the biennial monitoring reports on the EU Sustainable Development Strategy (Eurostat, 2016a), the EU Commission Communication on the next steps for a sustainable European future (European Commission, 2016a), a proposal for a new European Consensus on Development (European Commission, 2016b) and the 2016 Eurostat report on sustainability (Eurostat, 2016b).

  • Various OECD analyses, tools and approaches, including the recent assessment of OECD countries on “Measuring distance to the SDG targets” (OECD, 2016b).

  • The World Bank’s Genuine Savings Indicator, which is directly based on the capital approach (World Bank 2006, 2011), and the UNEP reports on “Inclusive Wealth” (UNEP and UN-IHDP, 2012 and 2014);

  • The Sustainable Development Solutions Network (SDSN).1

At the national level, since the 1987 Brundtland Report and the recommendations of the Stiglitz, Sen and Fitoussi (2009) report, a number of countries have started to collect data and to establish indicators on sustainable development. Among others:

  • In Italy, the National Statistical Office (ISTAT) and representatives of civil society developed a multi-dimensional approach to measure “equitable and sustainable well-being” (Benessere equo e sostenibile, BES), integrating measures of economic activity (GDP) and of the social and environmental dimensions of well-being, inequality and economic, social and environmental sustainability. The indicators are published on the BES website, with a detailed analysis of indicators available in the BES Reports.2 The BES system has now been incorporated in the annual cycle that drives economic and financial planning by the Government and Parliament.

  • Finland has a long tradition in using indicators on sustainable development: the first set of sustainable development indicators was published in 2000 and this list – available on the Findicators website3 – has been growing ever since.

  • Switzerland uses a measurement framework based on a system structure to monitor sustainable development. The system (MONET)4 comprises 73 indicators. Each indicator is published on the internet and evaluated (through “traffic light” symbols) according to the observed trend. The website also includes 9 indicators related to the global dimension of sustainable development.

  • The 2016-19 Federal Sustainable Development Strategy (FSDS)5 is Canada’s primary vehicle for sustainable development planning and reporting. It sets out priorities, goals and targets, identifies actions to achieve them, and links these to 12 global SDGs, with sustainable development indicators published on different webpages.6

  • In 2002, the German Federal Government adopted its “National Sustainability Strategy”. The Strategy was revised in 2016, setting out the challenges that have arisen for Germany from the 2030 Agenda. At the core of the strategy is a “sustainability management system” that defines objectives with a timeframe for fulfilment, indicators for monitoring, as well as regulations for the management and definition of institutional design. Since 2006, the Federal Statistical Office has reported on the indicators of the National Sustainability Strategy in six reports.7 The Federal Statistical Office has now been commissioned to continue its statistical monitoring based on the revised strategy.







7. The 2016 report by the Federal Statistical Office on the development of the indicators of the German sustainability strategy is available at:

9.2. Key concepts and approaches

9.2.1. The capital approach

Ensuring the well-being of present and future generations essentially depends on how societies choose to use their resources, i.e. their various forms of capital. These resources include physical elements, such as shelter and sub-soil assets or the quality of the natural environment, but also intangibles such as knowledge or the quality of social and institutional structures. The definition of capital should therefore include more than produced (economic) capital, to encompass social, human and natural capital.

Measuring capital requires constructing and examining balance sheets for different types of capital, for each country and for the whole planet, and assessing their changes over time. This type of “capital approach” may be characterised as reflecting what, in the field of financial markets, has been described as a “micro-prudential perspective”: all forms and appearances of capital are inventoried one by one. Measuring sustainability from a capital approach focuses on the net change in the volumes of the stocks of various assets, weighed by their “shadow prices” (i.e. a monetary value reflecting the true opportunity costs of all activities, taking into account all externalities and public goods generated by them); these shadow prices will in general differ from market prices, and depend on all other types of capital, technology and societal preferences. Shadow prices should also reflect future actions and their discounted consequences to make net changes in overall capital a true indicator of sustainability (for a full discussion of the theoretical issues associated with the capital approach, see Fleurbaey and Blanchet, 2013).

One important implication here is that adopting such a comprehensive capital approach would require reconsidering conventional distinctions between consumption and investment activities. For example, expenditures that contribute to a society’s human, and/or social capital (e.g. training of teachers) will enhance that society’s long-term sustainability, rather than simply representing final or intermediate consumption as assumed in the current System of National Accounts (SNA). Thus, measures of investment, consumption and national wealth in a sustainability perspective would generally differ from the one used in conventional economic statistics. As, in general, we lack a good understanding of the relevant flows of investment and depreciation, these difficulties were recognised in the Stiglitz, Sen and Fitoussi (2009) report, which recommended establishing a sub-dashboard to provide information about changes of those “stocks” that underpin human well-being. For economic, human and natural capital, the progress made in developing metrics made in recent years and the current challenges, as well as recommendations for the future, are described in Sections 9.3 and 9.4 below.

9.2.2. The systems approach

One of the lessons learnt from the financial crisis was that a micro-prudential approach of supervision and regulation is, in itself, insufficient to avoid financial crisis. A macro-prudential layer is also needed, as the interactions among individual financial institutions are, from a financial sustainability perspective, as important as the conditions of a single institution. To ensure overall sustainability therefore, a “systems approach” was needed.

The same dichotomy can be applied to complex systems in general, differentiating between the behaviour of single components considered in isolation, and that of the full system with all interactions. Such a systems perspective (Costanza et al., 2014) allows the identification of relevant actors who play a particular role in the system, their behaviour and their interactions with the other actors; and helps to address key questions of system dynamics applied to the society, economy and environment we live in.

While these systems are complex by themselves, their interactions with each other add to complexity. As mentioned above, besides dynamics, complex systems are often characterised by non-linearities (which could be negative or positive), emergent properties, self-organisation, tipping points, and transformation. These characteristics make the system subject to self-generated (endogenous) shocks in addition to shocks coming from outside the system (exogenous). The interactions of these characteristics help to define the degree of vulnerability of a system to risks as well as its resilience.

The key notions underpinning the systems approach could be described as follows:

  • Risk is here used with a double connotation: first, to describe threats for systems (i.e. systemic risks); second, the term risk is used when it is possible to quantify the probability of an unknown event. Contrary to this situation, uncertainty is used if the “unknown” cannot be quantified. For decision making and citizens’ participation in democratic societies, this distinction is critical. Individual well-being and social welfare are affected by these risks, but also depend on the availability of information about them. The availability of information – or the treatment of uncertainties – is particularly important concerning social inequalities linked to risk arising from economic activities.

  • Vulnerability manifests itself in damage from a disturbance which might arise from external or internal factors. It ranges from “low” to “high”, depending on the severity of the impact of the disturbance on the system. There are different definitions for vulnerability, ranging from vulnerability within a specific context, such as in an environmental context (climate change) or in a personal context (poverty, risky behaviour, sickness), to more integrated approaches relating to both physical and social systems’ susceptibility to multiple stresses generated by socio-economic and environmental changes. Complex systems may be vulnerable to all kinds of risks inherent in everyday life and which are the result of decisions, naturally occurring events, or a combination of both. In this wider context, vulnerability is of a multi-scale nature; it is not evenly distributed among social groups, spatial units or time. Vulnerability of a system is shown by its response to disturbances, and depends on the intrinsic characteristics of the system and on the severity and time of exposure of the disturbance.

  • Resilience, rooted in the Latin word resilire (which means “jump back”, “rebound”), is the capacity to recover from adversity (from temporary shocks – such as sudden flows of migration – or from continuous threats/slow-burn processes – such as the ageing of societies), either returning to the original state or moving to a new steady state (from positive adaptation to transformation), often strengthened and more resourceful. Resilience, which is the centrepiece of the systems approach, can be a characteristic of a system, sub-system or individuals. This distinction is important, because system-wide resilience is typically the result from a series of interactions among individuals and sub-systems. It thus provides us with a concept and a structure that allows us to identify components, the monitoring of which could underpin a policy framework. At the same time, resilience also makes the role of dynamics more explicit, by looking at disturbances and their short- and long-run impact.

Focusing on resilience should not be misunderstood as an attempt to abandon sustainability, but instead as an approach to retain or restore it when responding to shocks and threats. Sustainability is and remains the ultimate objective. The fact that we begin to examine how systems respond to shocks and threats does not lead us to give up on preventing shocks or negative events. Similarly, if a resilient system rebounds from a shock to its initial path, this is not good enough if the initial path was unsustainable. While resilience and sustainability are not inter-changeable concepts, addressing resilience of systems is a way to build a system-level “macro-prudential” approach about how we can prevent and adapt to shocks and threats or transform our society. The “pressure-state-response” used in environmental economics provides an early example of the systems approach (Box 9.2). A more in-depth discussion is included in Section 9.5.

Box 9.2. The pressure-state-response approach as an example of the systems approach

The original pressure-state-response (PSR) approach developed by Statistics Canada and popularised by the OECD with reference to natural capital was later extended to Driver-Pressure-State-Impact-Response (DPSIR), adopted by UNDP in 1997 and used in the context of Environmental Economic Accounting (EEA). The DPSRI framework (Figure 9.1) contains the different elements describing the resilience of the State of the Environment (SoE) and allows the classification of data and indicators for the different elements (drivers, pressures, states, and impact). On the basis of such data, and using estimation and extrapolation techniques, the Intergovernmental Panel on Climate Change (IPCC, 2015) has estimated the planetary boundaries for climate change. The limits of a resilient ecosystem and these boundaries were again stated in COP21 (UN 21st Conference of the Parties, Paris, 2015): global temperature should be kept to well below 2ºC above pre-industrial levels, and we should try to keep it to 1.5ºC.

This example demonstrates that using the best available data and estimates, a scientific consensus could be reached on the planetary boundaries. Once these are set, it is possible to look at the scientific work and political actions needed to keep the system within these boundaries (see the discussion on carbon pricing in Section 9.3).

Figure 9.1. The pressure-state-response approach
Figure 9.1. The pressure-state-response approach

Source: European Environment Agency (1999), Environmental Indicators: Typology and Overview, Technical Report No. 25/1999, TNO Centre for Strategy, Technology and Policy, The Netherlands.

9.3. What is the state of existing capital measures?

The capital approach builds on the notion of preserving or increasing the different stocks (capital) that drive our welfare and well-being. This “stock” approach to sustainability can either look at variations in each stock in physical terms or convert all these assets into a monetary equivalent. So the capital approach could evolve in two directions: either as a “mainstream economic approach”, determining all types of capital and monetising them; or as an “organising framework” with physical indicators covering all the main assets. In the following, both interpretations are used, depending on the maturity of the data. Each type of capital is discussed separately.

9.3.1. Economic capital

Progress since 2009

During and after the global economic crisis, many of the issues raised in the Stiglitz, Sen and Fitoussi (2009) report gained traction, while at the same time the failure to appropriately measure economic sustainability by conventional statistics became apparent. Risks and vulnerabilities built up over time in the economic system through increased debt levels, higher financial leverage (supported by more liberal finance laws and regulations), deteriorating quality of debt through higher credit default risks, price bubbles and increased inter-connectedness across sectors and countries. This fragility went unnoticed due, partly, to the way that economic health was measured: it did not sufficiently measure risk.

During and immediately after the crisis, it was clear that many actors lacked appropriate and timely data to help them respond effectively. The G20 Data Gaps Initiative (DGI)2 has been an important source of progress in providing broader and more comprehensive measures of economic sustainability. This initiative supports government efforts to provide comparable statistics on the build-up of risks in the financial sector, cross-border financial links, vulnerability to shocks, and communication of these statistics. It emphasises the importance of having more and better information on how the assets and liabilities of one sector match those of others, as well as on currency and maturity mismatches. The G20 DGI provides templates for collecting balance sheet data that can be compared across countries. Many countries, especially within the EU, now make quarterly data available in a timely manner. Some even provide the institutional sector accounts discussed below.

One of the lessons on economic sustainability learned since the 2007 financial crisis is that there is no firm threshold for public debt beyond which we should expect GDP growth to fall significantly, even though high levels of public debt may raise concerns about the resilience of the economic system to shocks. Even studies showing correlations between public debt and GDP growth tell us little or nothing about causality. The only real test of the sustainability of public debt is provided by the market (i.e. the ability to sell government bonds), which is itself a function of the institutional setup of different countries (such as the existence of a central bank and the ability of monetary policy to maintain low interest rates) and of the assessment of future prospects for public finances given the evolution of demographic and other factors. There is also evidence that the relation between public debt and GDP growth may run from (low) growth to (high) public debt rather than the other way around in some circumstances.

The recovery from the crisis has been lacklustre in much of the world and, in the United States and some other OECD countries, has further concentrated income gains at the top. Part of the reason for the unsatisfactory recovery across the world was the implementation of austerity measures, enacted under the misguided belief that there is a critical threshold above which debt lowers growth. Erosion of human capital due to unemployment and underemployment, discussed elsewhere in the report, is likely to have lowered growth for years to come. Investments in crumbling public infrastructure could have helped millions of people, maintained human capital stocks, and gone further towards helping the economy to recover.

Outstanding issues and new questions

While the 2008 SNA includes full balance sheets for economic assets and liabilities, many countries are still guided by a very limited (and possibly misleading) approach to sustainability. As described above, comparisons of (gross) public debt to GDP are incomplete measures of economic sustainability. It may be that part of the appeal of using the debt to GDP ratio as an indicator of sustainability is that it is relatively simple to calculate and understand.

However sustainability has two additional aspects to be considered:

  • A full balance sheet approach (i.e. taking stock of a broader range of both assets and liabilities, and associated risks), by looking at:

    • the balance sheets of all sectors (banks, households, etc.) rather than the government alone;

    • both liabilities and assets (e.g. recognising that fire-sales of assets in depressed financial markets may worsen net worth);

    • distinguishing between types of economic capital that add to productive capacity and those that do not (e.g. land), and between changes in volumes and changes in prices (Box 9.3).

  • A long-term sustainability analysis which takes account of the impact of demographic and other factors on the evolution of public finances.

With regard to the latter aspect, models and scenarios provide valuable guidance to societies on the choices they have to make to achieve sustainability. Models will become increasingly important to assess interactions between different types of capital and their determinants (Box 9.4).

Box 9.3. W versus K

Stiglitz (2015a), in his discussion of Piketty’s (2015) finding of a long-term increase of the capital-output ratio, notes that such an increase should normally be accompanied by a decline in the returns to capital relative to labour and by a declining capital share in income. Neither is the case empirically, though, as data on declining labour shares and real wages indicate. Stiglitz’ resolution of the puzzle lies in distinguishing between wealth (W) and capital services in production (K):

The distinction between W and K reflects the dual nature of capital, i.e. as a factor of production and a means of storing wealth. This distinction is a well-established feature in the literature on capital measurement (Jorgenson, 1963; Jorgenson and Griliches, 1967; Diewert and Schreyer, 2008; OECD, 2009) but is sometimes overlooked in the debate. Each aspect of capital is associated with a particular measure.

  • The wealth aspect of capital requires a measure that reflects the market value of capital goods. W is the conceptually correct entry into balance sheets. Balance sheets relate to particular points in time and valuation of wealth is at the prices prevailing at these points in time. The change in wealth between these points in time is made up of investment or other additions to the stock, minus depreciation or depletion, and revaluations.

  • To capture the production aspect of capital, a volume measure K is required to reflect the flow of capital services into production. Unlike the wealth stock, the price of capital service is identified with user costs, designed to capture the marginal productivity of the different types of capital.

Stiglitz’ shows that: “The wealth income ratio could be increasing even as the capital income ratio is stagnating or decreasing. Much of wealth is not produced assets (“machines”) but land or other ownership claims giving rise to rents. Some of the increase in wealth is the increase in the capitalised value of what might be called exploitation rents – associated with monopoly rents and rents arising from other deviations from the standard competitive paradigm. Some is an increase in the value of rents associated with intellectual property.” (Stiglitz, 2015, p. 8).

This distinction implies that the evolution of the wealth-output ratio in nominal terms can be very different from the capital-output ratio in volume terms. The distinction also raises a question of scope of the two concepts – the inclusion or exclusion of some assets can modify the entire profile of the wealth-output and capital-output ratio. One such asset is land, and in several countries the rise in the overall wealth-output ratio has been driven by the steep revaluation of land, confirming Stiglitz’ point. Despite the two distinct perspectives, the wealth and the production sphere are linked and so are its measures. Indeed, W and K should be constructed consistently and as part of an integrated framework as laid out for instance by the 2008 System of National Accounts (SNA) or in more detail by OECD (2009) and Jorgenson and Landefeld (2007).

Box 9.4. Modelling public finances

Policy-makers in some countries are becoming more interested in modelling the path of public finances, acknowledging that demographic evolution (particularly ageing) is a major factor in analysing fiscal sustainability, alongside structural reforms and productivity developments. Notable examples of the approach include the United States,1 Australia2 and European Union countries.3

This approach typically analyses the situation of countries over the short, medium and long term against their government debt level, their initial budgetary position and the projected evolution of ageing costs (notably old-age pensions, healthcare and long-term care). It uses a range of assumptions, including on demographic evolution, real GDP growth, inflation, real interest rates and labour market participation. An important aspect of the approach is to consider different scenarios, communicating clearly the sensitivity of the results with respect to the assumptions.

One advantage of this approach is that it is possible to analyse the projected path of sustainability over time, thereby identifying particular future periods of stress.




Arguably, macro-economic models should go further to reflect the joint determination of the paths of economic output and public debt levels by interest rates and the government primary balance (i.e. government net borrowing or net lending, excluding interest payments on consolidated government liabilities). This implies recognising that monetary policy and budget rules cannot be set independently of each other, and that fiscal consolidation in a recession may have large effects in terms of reducing GDP growth and limited effects in terms of reducing public debt (possibly further increasing it). Macro-economic models should also highlight the path that private demand is expected to follow under a given configuration of policy instruments, and the need to adjust such instruments when the path of private demand is inconsistent with macro-economic goals of full employment and price stability.

A full balance sheet approach to economic sustainability would also imply a more nuanced approach to sustainability, one that is not likely to rely on a “single number”. This makes it more difficult to decide, to take a pertinent example, when it is appropriate to engage in fiscal stimulus.

In this context, a complete balance sheet would have several important characteristics:

  • Private wealth should be considered alongside the assets and liabilities of the public sector, as private liabilities may be converted into public liabilities if particular agents fail (due to bank bailouts, for example). In addition, the tax base upon which the government can draw for meeting its liabilities depends on the net wealth of the private sector. In both of these respects, some sort of distributional information is important since aggregation may mask the fact that for many agents debt is not covered by assets. The share of households (or firms or banks) with negative net worth may be a useful indicator. There is also value in considering the transmission of wealth between generations through examining data on inheritance of assets more closely.

  • A better balance sheet would also take into account the fact that, even though the value of an asset (e.g. land) has increased, and overall measures of wealth have gone up, this is not the same as an increase in the volume of productive assets.

  • More detailed balance sheets of financial corporations and other institutional sectors are critical to understand risks and vulnerabilities. The G20 Data Gaps Initiative (DGI) recommends producing quarterly institutional sector accounts. We should also recognise that, when risk is not properly measured, we may underestimate the fragility of firms, households, and other institutions in the face of financial stress. Balance sheets should be more detailed, both in terms of showing more granular sub-sectors (to illuminate differences in vulnerabilities as measured, for example, by debt-to-income ratios) and more detailed data for each of those sub-sectors, while taking into account the costs and benefits of collecting and analysing more detailed data. More detailed balance sheets within a sub-sector would allow for a better analysis of risk through examining inter-connectedness by having breakdowns by counterparty sector, or breakdowns of debt by maturity and currency.

  • All relevant types of pension liabilities need to be included.

Even such improved and more detailed accounts, however, may not be sufficient to capture macro-economic risk. One reason for this is that there is no clear conceptual framework for capturing risks at a macro-level, and that a full understanding of the links between macro- and micro-level risks (such as the building up of sectoral risk) is still missing. Therefore, better indicators should be compiled to measure different types of risk (liquidity, solvency, maturity, currency, overexposure, contingencies and guarantees), and their concentration in specific segments of the economy. Aggregate data will not suffice. One also needs more granular information to assess what fraction of firms (or households) will face financial stress in the event of changes in asset prices, and the importance of these firms for the whole economy.

There are also other aspects to risk. For example, existing SNA conventions may lead to considering higher risks as adding to the value added of financial services. This issue is reflected in an on-going discussion about the measurement of “financial intermediation services indirectly measured” (FISIM). Financial intermediaries assume risks when they provide loans; hence the core question is whether the higher risk premiums that banks may incur increase their output, or whether risk is borne by other sectors or society at large and should not be reflected in the output of the financial sector. Doing so, one should make a clear distinction between developments in current and in constant prices.

These issues form part of the “systems” and “resilience” issues discussed in Section 9.5.

9.3.2. Human capital

Countries with higher human capital have stronger economic growth, and individuals with higher human capital and better capabilities to achieve better individual outcomes. At the country level, OECD (2010) estimates that increasing PISA scores (Box 9.6) by one standard deviation would increase GDP growth by 1.8 percentage points. At the individual level, people with higher education live longer, have higher earnings and accumulated wealth, better health, denser networks of connections and are more active citizens.

Human capital has been defined as the “knowledge, skills, competencies and other attributes embodied in individuals that facilitate the creation of individual, social and economic well-being” (OECD, 2001).3 This OECD definition is all-embracing: it incorporates various skills and competencies that are acquired by people through learning and experience but may also include innate abilities. Some aspects of motivation and behaviour, as well as the physical, emotional and mental health of individuals are also regarded as human capital in this broader definition.

In practice, however, the measurement concentrates on a narrower definition. Knowledge, skills and competencies certified by formal education have been the object of earlier research in the measurement of human capital. More recent developments have looked at other approaches to complement educational attainment indicators.

Measurement of human capital has implications for understanding the fundamental processes of societal development and economic growth. It also matters for estimating and understanding inequalities within societies. Measurements of human capital are important for the accountability of the education and health sectors. They help in accurately accounting for the costs and benefits of societal phenomena such as unemployment, and of the proposed policies to address those problems.

Human capital significantly determines a country’s consumption and production possibilities, today and in the future. People’s knowledge, skills and competencies are “capital” in that they can be built up, but they can also decay, particularly during long periods of unemployment or sickness, or following shocks such as wars or migration. Failing to account for human capital could lead policy-makers to underweight investments in education, youth employment and public health, with detrimental consequences for the future.

This is why in this chapter human capital is examined through the lens of sustainability. One of the motivations for focusing on human capital is a concern that, during and in the aftermath of the financial crisis of the late 2000s, estimates of the cost of the crisis did not reflect the decrease in human capital from high rates of youth unemployment, workers’ layoff, loss of firm-specific human capital, and lower spending on training by firms. If these costs were underestimated, the response to the crisis in terms of fiscal stimulus or investment in education and skills may have been too weak.

While there is widespread agreement about the importance of maintaining and increasing human capital to ensure sustainability, there continue to be discussions on the best way to measure it, on the advantages and disadvantages of different definitions of human capital (what should be included and what can be measured given data limitations) and on different methods to value it (Box 9.5).

Box 9.5. Approaches to measuring human capital

Traditionally, the most common approach to measuring aspects of human capital has been to use non-monetary indicators of educational output, such as literacy or secondary school graduation rates. This type of indicator has the advantage of being widely available, both across countries and over time, even if data may not be fully comparable across countries. More recently, the indicators approach has been developed to take into account other aspects of human capital formation and stock characteristics. The use of standard classifications, for example by type of education programme, has improved the quality of these indicators.

Other approaches have aimed at producing a single summary measure of the stock of human capital in a country, expressed in monetary terms. These approaches have built on:

  • The Cost-based approach (Kendrick, 1976), where the stock of human capital is estimated as the depreciated value of the stream of past investments in human capital, such as teacher salaries and all other expenditures on education.

  • The Lifetime income approach (Jorgenson and Fraumeni, 1989, 1992a, 1992b), where the discounted value of the future labour income of individuals in the population for different education levels is calculated.

  • The Indirect or residual approach, which estimates the human capital stock as the difference between the discounted value of future consumption flows and the monetary value of other measured capital stocks. Because of its limitations, this approach is not recommended by the Guide on Measuring Human Capital (UNECE, 2016).

Summary measures of the stock of human capital may also be based on some combination of indicators and monetary measures. For example, the stock of human capital in a country may be measured as a weighted average of the mean years of schooling of different segments of the population (including those that are currently inactive), with weights based on estimates of the “rates of return to schooling” for various educational categories used to capture “quality”.

Progress since the 2009 Stiglitz-Sen-Fitoussi report

Some progress has been made in measuring human capital, as reflected in the Guide on Measuring Human Capital (UNECE, 2016). The guide provides reference and support for different strategies and approaches to measuring human capital, with an emphasis on preparing satellite accounts on human capital in line with SNA guidelines.

There is today general agreement in the statistical community on basic methodologies towards measuring human capital as related to education and labour market returns, though significant concerns remain. In general, in a National Accounts context, the most appropriate measures are either cost approaches or the discounted lifetime income approach (in the tradition of Jorgenson and Fraumeni).4

While the lifetime income approach is appealing from a theoretical point of view, it requires detailed data and a number of assumptions; in particular the value of human capital today depends on the assumptions you make about GDP growth in the future. This complicates the estimation, for example, of the impact of the economic crisis on future growth through the channel of reduced human capital. Other assumptions about the future must also be made, for example on life-expectancy, and these assumptions can have large impacts on the overall estimate of the value of human capital.

The cost approach is based on past expenditure, but is also requires assumptions, in particular about depreciation of, and future returns to, human capital. However, for data availability reasons, the cost approach is most often preferred.

Several National Statistical Offices have recently undertaken initiatives to develop monetary measures of human capital, to be used alongside indicators of education quality and achievement. One common finding of these studies is that, whatever approach is used, the value of human capital is high compared to economic capital, even if the size of the discrepancies between estimates based on lifetime incomes and the cost approaches remains a puzzle (Liu, 2011). However, beyond the numerical estimates produced by these studies, considering educational expenditures as investment rather than consumption would have large impacts on how capital formation is defined and understood.

In addition, more recently, there has been substantial progress in the direct measurement of cognitive skills – in particular by the OECD through the PISA and, since 2011, the PIAAC survey (Box 9.6). The PISA survey, in particular, has played an important role by bringing human capital to the attention of policy-makers in the educational community and beyond.

Box 9.6. PIAAC and PISA surveys

Two OECD-sponsored instruments are increasingly used as a basis for computing human capital indicators:

The Programme for International Student Assessment (PISA) was run in 2006, 2009, 2012 and 2015. While all waves of PISA included tests in mathematics, science and reading, the 2006 wave focused on science; the 2009 one focused on reading; and the 2012 PISA focused on mathematics. PISA testing also occurred in 2000 and 2003. In 2003 and 2012, tests were also offered in creative problem solving, while the 2012 wave included an optional test of financial literacy.

The Survey of Adult Skills, a product of the OECD Programme for the International Assessment of Adult Competencies (PIAAC), was designed to provide insights into the availability of some key skills in society and how they are used at work and at home. The first survey of its kind, it measures proficiency in several information-processing skills – namely literacy, numeracy and problem solving in technology-rich environments.

Outstanding issues and new questions

Progress still needs to be made to improve the lifetime income approach for estimating the value of human capital and, more generally, to expand our understanding and measurement of human capital. Most of the analyses have thus far considered human capital as formal education or cognitive skills, and its returns as increased labour earnings. Future work needs to expand the measurement of human capital to match the understanding that it is broader than education and cognitive skills, and that its returns are larger than individual earnings. The initial focus on education and labour returns was in part a function of data availability, and the fact that this concept was more straightforward to operationalise when limited to these aspects. Even so, measuring human capital in these narrow terms suggests that human capital investments are undervalued.

It is now important to build on this foundation to go beyond cognitive skills, education and remunerated activities. This broader perspective requires addressing difficult measurement questions such as how to measure non-cognitive skills and non-market benefits, both individual and social, and understanding and measuring specific human capital and networks.

It also implies taking the demographic structure of the population and life expectancy into account, as those who live longer and healthier lives are more productive both in the market and in society. Migration also has to be included in the measurement of human capital: there is a cost for sending countries when the better-educated population migrates.

Improving the lifetime income approach

Quantifying the relationship of future productive potential to both levels of human capital and human capital investment is critical to establishing support for a more comprehensive treatment of human capital accounts, even though such a task faces clear methodological challenges. Better and more consistent estimates of the returns to education require longitudinal studies that can account for cohort effects, for example (Boarini, Mira d’Ercole and Liu, 2012).

In addition, measuring specific human capital is more complicated than measuring general human capital. Specific human capital includes, for example, firm-specific human capital, or networks, while general human capital includes schooling or non-specific work experience. Failing to take on-the-job training into account may bias the estimate of the returns to formal schooling.

Understanding the gap between cost-based estimates and income-based estimates requires simultaneous estimation of the two. Implementing satellite accounts for human capital would help in better matching the two types of estimates.

However, many of the data needed for implementing the lifetime income approach are not available for some countries, and not necessarily harmonised. A better understanding of human capital, and improvements in its measurement, will come from more cross-country research. For example, countries vary in the structure of the earnings reported (time period considered, what is included in the earnings definition), and in the reporting of educational attainment. Harmonised data, where available, would allow researchers to improve understanding of the role played by education.

Finally, there is a need to better understand what the lifetime income approach can be used for, given the sensitivity of estimates to changes in assumptions about the future. While these estimates serve a very important role in demonstrating that human capital forms a very large component of wealth, and that spending on human capital should be considered as investment rather than consumption, it is less clear what practical use can be made of the approach in terms of planning and measuring sustainability.

Box 9.7. Human capital consequences of recessions

A substantial concern with recessions is that unemployment, particularly youth unemployment, erodes human capital, or limits human capital acquisition through on-the-job training. If the full cost of recessions (the long-term lower GDP growth due to lower human capital) were recognised, policy responses might be stronger. While it is difficult to measure the loss of human capital due to recessions, these effects are likely to be important, as graduates who enter the labour market during a recession can be expected to have permanently lower incomes, these efforts are ignored by most applications of the lifetime income approach.

Expanding the measurement of human capital and its returns

A strategy that values human capital by estimating the impact of education on lifetime income is also insufficient as it omits many important features on both sides of the equation: first, the human capital acquired outside of formal education, as well as non-cognitive skills; and, second, the non-market benefits of human capital. A more comprehensive approach to human capital measurement would be important to lead policy-makers to recognise that education expenditure is an investment rather than consumption.

While the focus on formal education and market returns has been a function of pragmatically starting from where data availability and conceptual clarity were higher, it is important to increase data availability and conceptual clarity on human capital at all stages of capital formation, including how it is embodied in individuals, and its benefits. For example, human capital investment takes place not only through education, but also through on-the-job training, parenting and household production of non-market services (Box 9.8), as well as through participation in cultural activities. Destruction of human capital occurs for example in the presence of high youth unemployment, whose effects are not only lower consumption today but a lower long-run growth trajectory of the country tomorrow. Similarly, it is important to recognise that human capital is embodied in individuals not only through knowledge and cognitive skills, but also through non-cognitive skills and traits. Its benefits encompass not only labour market returns, but higher subjective well-being, citizenship, caring, social trust, co-operation and health. In this broader perspective, health care expenditures should be recognised as a kind of maintenance and repair flow for human capital, and better health conditions as non-market benefits from human capital. Human capital may also stimulate the accumulation of social capital: the norms and values that children develop at school will enable them to participate better in society as adults (OECD, 2010). In this broader perspective, developments in human capital can be seen in the context of the systems approach (outlined in Section 9.2.2), thanks to its links with developments in economic, social and natural capital.

Box 9.8. Measuring and valuing unpaid household service work

Even if unpaid household work contributes to preserve and improve human capital, most of it is excluded from the production boundary of the SNA. Various efforts have been undertaken in recent years to measure the amount and type of work carried out in the household and to estimate the monetary value of this work. Some countries have started to value these activities in a Household Satellite Account, which provides important information on the economy and society. Time-use surveys are an important tool to capture the amount of time spent by individuals to provide non-market services that benefit other household members or society more generally. Putting a value to household work is however not straightforward since the work is unpaid and because it often results in intangible services. A UNECE Task Force (UNECE, 2017) recently released a set of guidelines for valuing unpaid household services.

Beyond the average

As with income or wealth, it is important to go beyond the average when examining human capital: there are important inequalities in human capital, which can vary by country and among population groups. In addition to understanding whether overall human capital is increasing or not, it is important to look at inequalities in human capital as these play an important role in shaping life-time inequalities. For this reason, measures need to differentiate between adults and children, different groups in particular countries, and different household arrangements, to understand how patterns are shifting over time.

Better measuring of these inequalities (in education and health care, for example) will contribute to better understand inequality of opportunity. While gender inequalities in human capital are very important from a variety of perspectives (Box 9.9), so are inequalities by income, race, caste or religion.

Box 9.9. Gender inequalities in human capital

While reducing inequalities in education would help reduce inequalities in life chances, gender specific inequalities have an even broader impact as they affect fertility decisions, the health of children, gender relations in the family with regard to power, gender division of labour and authority within households.

The goal of achieving gender parity in education by 2005 has not yet been fulfilled, despite significant improvements. By 2011 only 60% of countries achieved this goal at the primary level, and 38% at the secondary level. In the world, more girls than boys are out of school: girls make up 54% of the total number of children out of school. In the Arab States, the share is 60%, unchanged since 2000 (UNESCO, 2015; UN Women, 2015). The gender parity index has increased dramatically in Southern Asia, where inequality was highest in 1990 and is now the lowest. However, while sub-Saharan Africa, Oceania, Western Asia and North Africa have made progress, girls are still disadvantaged relative to boys regarding enrolment in primary education. Social inequality also widened, and this inequality often interacts with larger social and economic cleavages.

Attending school does not necessarily mean achieving basic literacy skills. It is a particular concern for poorer countries with insufficient teacher resources, but also in rich countries. The partial closing of the gender gap in primary education has contributed to reducing the incidence of illiteracy among women, but women still account for more than 60% of all illiterate persons in the world.

In secondary education, progress has been even more uneven across countries. On average, across the OECD, 43% of 25-64 year-olds have achieved an upper secondary or post-secondary non-tertiary degree. The improvement from the older to the younger cohorts is particularly large for women. Across the OECD in 2015, 37% of 55-64 year-old women, but only 15% of 25-34 year-old women, had no upper secondary degree (OECD, 2015).

For developing regions as a whole, the gender parity index for secondary education increased from 0.77 in 1990 to 0.96 in 2012. However, there are large differences between regions, with girls enjoying an advantage in Latin America and the Caribbean, but lagging significantly behind boys in sub-Saharan Africa, Southern Asia, Western Asia and Oceania. Southern Asia stands out as the region where the greatest progress has been made, with the region’s index increasing from 0.59 to 0.93 between 1990 and 2012.

Sub-Saharan Africa, on the contrary, stands out as the region where girls have the worst chances, particularly in the poorest sections of the population and in rural areas. Only 9% were completing lower secondary education by the end of the 2000s, a share that has been declining over time. Based on recent trends, girls from the poorest families in sub-Saharan Africa are only expected to achieve lower secondary completion in 2111.

Overall, at past rates, low-income countries would not achieve universal primary and secondary education before the end of this century. Around half of 15- to 19-year-old girls and boys are expected to complete lower secondary education in low income countries by 2030, while only 33% of boys and 25% of girls would complete upper secondary.

While gender specific disadvantages still remain in primary and secondary education, in tertiary education the gender gap was closed by 2015 and in some countries even started to reverse, with women outnumbering men. In two out of five OECD countries, as well as in Lithuania and the Russian Federation, one out of every two young (25-34) women has a tertiary diploma. Only in Canada, Korea, Luxembourg, the Russian Federation and the United Kingdom do men have such high rates of tertiary education. Gender differences, however, still remain in field of specialisation, with women concentrating in humanities and men on the scientific and technical sectors. Furthermore, the gender balance again reverses at the upper tertiary level, with more men than women obtaining a PhD. Finally, although most tertiary graduates are women, men still have better labour market outcomes in terms of both participation and earnings (United Nations, 2015b).

9.3.3. Natural capital

At a social and political level, concern about climate change and environmental sustainability has continued to grow since the Stiglitz, Sen and Fitoussi (2009) report, with real progress in addressing these issues on a meaningful scale being slow. At the same time, there has been a shift from thinking about capital to thinking about quality, biodiversity and ecosystems – so not just about quantity and volume anymore.

The development of the capital approach as applied to natural capital has followed four historical episodes (and corresponding measurement tools), each of them driven by environmental crises:

  1. 1. Measuring volume and price changes, driven by energy crises and depletion of natural resources.

  2. 2. Measuring local changes of environmental quality, linked to growing degradation of air, land and water quality, and poorer waste treatment.

  3. 3. Focus on measuring global phenomena, linked to awareness of ozone layer depletion and climate change.

  4. 4. Measurement of ecosystems and planetary boundaries.

We are now in this fourth episode, moving beyond the measurement of individual stocks of natural capital and towards ecosystems. This considers the “interplay of different assets (for example, within a forest, there is an interplay between water, timber, soil, and wildlife)”. This definition, provided by the System of Environmental-Economic Accounting (SEEA) discussed below, makes clear that, in order to measure environmental sustainability, more than the measurement of stock is required. Ecosystems are not a collection of different stocks but, more fundamentally, systems and, as such, they can have greater or lesser degrees of resilience. They provide a multitude of services to society (for example, a forest not only supplies timber but may also provide water retention and flood or landslide protection, air filtration, carbon sequestration, habitat for rare species, and recreation).

A capital approach, applied to nature, could, in principle, allow values in the environment to be compared to values in the economy, providing a bridge between environment and economics. However, a common measurement framework is not easily adapted to the measure of ecosystems. This section summarises progress in measuring environmental assets since the Stiglitz, Sen and Fitoussi report (2009); identifies areas that require work most urgently; and sets out a path towards measurement of ecosystems as a key part of natural capital.

Progress since the 2009 Stiglitz, Sen and Fitoussi report

Several advances in the measurement of natural capital have taken place since 2009, some of them codified in international frameworks and recommendations. In particular, the System of Environmental-Economic Accounting – Central Framework (SEEA CF) was adopted as a statistical standard by the United Nations Statistical Commission in 2012 (United Nations et al., 2014a). It covers the first three episodes described above: measuring volume and price changes, local environmental quality, and global phenomena.

The SEEA extends national accounting to include a broader set of environmental assets, for example fish stocks. It is designed to produce comprehensive and systematic information on environmental conditions linked to the economy to help guide policy-making, to understand the drivers of environmental change, and to assist with modelling and scenario building. The SEEA Central Framework also covers initiatives taken to measure carbon emissions embedded in a country’s imports and exports (“carbon footprints”), based on multi-country input-output tables. Box 9.10 describes the progress that National Statistical Offices have made in implementing the SEEA Central Framework.

The SEEA Central Framework defines environmental assets as the “naturally occurring living and non-living components of the Earth”, together constituting the biophysical environment, which provide benefits to humanity. In the SEEA Central Framework, environmental assets are viewed as individual components (including land, mineral and energy resources, timber and aquatic resources, and water resources) that make up the environment. For these assets, physical as well as monetary asset accounts can, in principle, be compiled to describe the opening and closing stocks as well as the changes in these assets. In practice, many conceptual and data problems limit our ability to both quantify several of these assets in physical terms and to value them in monetary terms.

A significant further development in the field of measuring environmental sustainability that occurred since 2009 was the development of the SEEA-Experimental Ecosystem Accounting (SEEA-EEA) published in 2014 (United Nations et al., 2014b), and which corresponds to ecosystems and planetary boundaries, the fourth episode described above The SEEA-EEA represents initial efforts to define a measurement framework for tracking changes in ecosystems and linking those changes to economic and other human activity. In this framework, an ecosystem is a dynamic complex of plant, animal and micro-organism communities and their non-living environment interacting as a functional unit.

Human activity influences ecosystems across the world and significantly modifies many ecosystems. Several countries have begun to set up experimental accounts that describe ecosystem assets and the flows of services from these ecosystems. Ecosystem services include provisioning (e.g. food, water), regulating (e.g. flood protection, air filtration) and cultural (e.g. recreation) services.

Box 9.10. Adoption of the System of Environmental-Economic Accounting

A global assessment of SEEA implementation undertaken by the United Nations Statistics Division in 2014 indicated that 54 countries have established a programme on environmental-economic accounting as part of their national statistical programme, with 15 more planning do so in the short term. Topics covered by these current and perspective accounting programmes differ between countries. In a nutshell, the UN assessment shows that developed countries’ accounts tended to focus on air emissions, environmental taxes, material flows, the environmental goods and services sector and physical energy flow accounts; while developing countries focused on water and energy. In the EU, the focus has been on physical and monetary flow accounts, while outside the EU the focus has been on natural resources accounting.

These differences in compilation practices may reflect differences in national priorities. The policy demand in developing countries may be understood as stemming from the need for better managing their endowments of natural resources and from specific security issues related to water and energy.

Natural capital accounting (NCA) considers natural capital as an important element in decision making for national development and economic growth, complementing GDP data with stock measures in particular of natural resources and ecosystems. The Wealth Accounting and Valuation of Ecosystem Services (WAVES) Partnership led by the World Bank and involving many UN agencies, national governments, academia and NGOs aims to ensure that natural resources are mainstreamed into development planning and national accounts. WAVES has adopted the SEEA as the underlying statistical framework to inform policies.

As mentioned above, developing methodologies that allow valuing different systems (economic, social, and environmental), and that allow these values to be compared to one another, is important, and monetary valuations are often called upon to serve this role. However, pricing natural capital is difficult, not only conceptually but also technically. There are still no agreed methods to estimate the monetary value for many environmental assets, because the market prices of environmental assets are inadequate or non-existent for several reasons. In a market context, economists use market prices to evaluate trade-offs, implicitly assuming that the price for a good or commodity obtained from the market reflects their marginal value for society as a whole. However, this relationship breaks down in the presence of externalities, which are large in the environmental sector, or when prices are not observed (as there are no transactions).

While accounting for non-financial, non-produced assets remains a hurdle that has not yet been overcome, progress has been made on measuring land and subsoil assets (a set of issues belonging to the first historical episode in the development of the capital approach). A number of countries already produce monetary estimates for these assets. Other forms of natural capital remain, however, uncharted territory, in particular when it comes to ecosystems. A number of countries have started experiments to systematically describe ecosystem capital, and first experimental estimates of their monetary value have been made.

Outstanding issues and new questions

The need to measure environmental assets and, in particular, ecosystems and planetary boundaries (episode 4 above) is now being recognised. Lots of research over the last 50 years has gone into measuring these assets and into developing valuation techniques that would allow monetary values to be estimated for them. However, the most critical measurement issues have still not been resolved.

Uncertainty remains over both how to measure quantities and conditions (of sub-soil assets, public goods, ecosystems and their services) and on how to price them. In the case of global phenomena, and in particular for ecosystems and planetary boundaries, these assets are non-traded, and so the market system will generally not provide the metrics for measuring and pricing. Reliable, widely-accepted alternatives have yet to be established.

While all OECD countries produce measures of produced assets, only a handful of countries produce complete SNA balance sheets that include also the value of land and sub-soil assets: Australia, France, Korea, and the Netherlands. The value of other environmental assets is generally excluded.

More countries should apply the SEEA, and produce more timely and reliable environmental-economic indicators based on it. These could include measures of “resource productivity” (the amount of material consumed per unit of GDP) and measures of the “circular economy” (e.g. recycling rates or cyclical use rate, measuring the amount of materials that is reused in relation to total material use). These indicators do not need to be in monetary units, but they should meet the same quality criteria as GDP.

It should be clear from the above discussion that measuring sustainability, in the narrow sense of keeping the total monetary value of the stock of natural capital constant, is plagued with uncertainties that result from a combination of difficulties to find price estimates in the absence of markets; difficulties to predict future demand; uncertainties about system behaviour including lack of knowledge about the inter-dependence of different systems; and so on. A more comprehensive approach should aim to develop an information system that enhances our knowledge about all components of natural capital, including subsoil assets, land and the way it is used, and ecosystems. This is the ambition of the systems approach presented in Section 9.2.2.

Box 9.11. Carbon pricing

The conceptual and data issues related to carbon pricing are relatively simple, and the underlying phenomena well understood compared to many other types of natural capital. However, even for such relatively simple case, the following would have to be addressed.

  • Carbon prices should fully reflect the social cost of emissions. This social cost is hard to predict and should include not only the costs of providing the global public good in question, but also take into account tipping points and non-linearities in the damage from each additional unit of emissions.

  • Carbon prices should incorporate time discounting appropriately. The discount rates applied should reflect both an element of “pure time discounting” (i.e. how much consumption tomorrow is valued by a person relative to consumption today) and an assessment of how much better off future generations will be relative to the current one. Deciding on this second element of the discount rate is not straightforward, but has a large impact on carbon pricing.

  • Distributional effects complicate pricing carbon emissions. The effects of climate change will fall disproportionately on certain social groups and places, so the most appropriate carbon prices should reflect the degree of “aversion to inequality” of the community.

  • Finally, the price of carbon emissions should take into account cross-border externalities, i.e. the effects of emissions in one country on other countries, as well as risks and resilience.

9.4. Towards a systems approach to inform policy

In the capital approach, the different forms of capital (human, social, natural, economic) are considered separately. This implicitly assumes their independence and, therefore, substitutability. Since we know, however, that they are not truly independent, a more adequate measurement approach would call for a further step, going beyond independently measured balance sheet items. To properly describe phenomena that are shaped by the interaction between complex systems (be they social, economic or ecological), a more macroscopic approach is needed. In practice, the ambition to build balance sheets cannot be achieved for some assets, in particular when assets are difficult to value in monetary terms, either because of non-economic benefits that flow from the asset or because the valuation of an asset is complicated or involves uncertainty (as in the case of sub-soil assets that have not yet been discovered). Further, certain government activities provide benefits to society as a whole (they may be public goods or the provision of commodities generating large externalities), whose value may bear little relation to the cost of the assets providing these services (unlike private sector activities, where in equilibrium, marginal benefits should equal marginal costs).

Another limitation of this approach is that deciding whether a particular situation is sustainable is difficult in the presence of risk and uncertainty. Whether a given situation is sustainable depends on the risk posed by that situation, implying that some evaluation is needed of whether that level of risk is acceptable. Apart from the fact that people are more loss averse than risk averse (i.e. they will take considerable risks if they don’t expect to lose much), estimating the level of risk itself is difficult, and in the case of uncertainties it is impossible. People may have different preferences regarding risk and the best way of dealing with uncertainties, and those preferences may differ across generations, making it difficult to ascertain whether a particular development (as described by successive balance sheets) is sustainable or not. The global nature of sustainability adds a further layer of difficulties (issues related to causalities, property rights, etc.) and complexity to attempts of measuring the different stocks of capitals.

Most likely, the “holy grail” of a society-wide balance sheet, incorporating all types of capital and permitting the different sectors to “talk” to each other by assigning monetary values (under the assumption of “weak” sustainability) will never be fully achieved. Or it will be achieved only at the price of heroic assumptions. Setting shadow prices entails evaluating the future, which is a daunting task falling outside the remit of official statistics (Fleurbaey and Blanchet, 2013). Further, these estimations and assumptions would have such a large influence on the conclusions of the exercise that the exercise itself would likely be neither successful nor helpful as a contribution to a democratic debate on the societal choices related to sustainable development.

To better understand the complexity of our world we should look at it from a systems perspective (Walshe, 2014; Borio, 2009; Fiksel, 2006 and Costanza et al., 1997) and examine how these systems – the ecological-social-economic-political systems – cope with changes and shocks. This broadens the notion of sustainability with the dimension of the system’s ability to cope with future, known and unknown, disturbances. This should ensure that the system remains sustainable, or at least that it has the ability to restore its sustainability after a temporarily unsustainable period.

The two main dimensions of shocks and of slow-burn processes (such as demographic changes) that determine how the system could respond to them (hence their resilience) are intensity and persistence. The interaction of these two dimensions determines the system’s ability to sustain a resilient behaviour; in turn, such ability can be classified as “absorptive capacity”, “adaptive capacity” and “transformative capacity”. Each of the three can then be linked to different types of interventions aimed at enhancing the system’s resilient behaviour,5 as described in Figure 9.2.

Figure 9.2. Shocks and capacities
Figure 9.2. Shocks and capacities

Source: Manca, A.R., P. Benczur and E. Giovannini (2017), “Building a scientific narrative towards a more resilient EU society – Part 1: A conceptual framework”, JRC Science for Policy Report, No. 106265, Publications Office of the European Union, Luxembourg,

When the time of exposure is not too long and the intensity is not too large, the main characteristic of this ability is the absorptive capacity, which relates to stability and resistance, i.e. a situation where agents absorb the impact of shocks without changing their behaviour. As the time of exposure or its intensity increases, and absorptive capacity is exceeded, adaptive capacity will start playing a role: agents adjust their expectations and aspirations when coping with deteriorating conditions. This requires flexibility, and involves incremental changes that are necessary to allow agents to continue functioning without major qualitative distress in response to disturbances. Agents try to mitigate potential damages and at best turn the adverse situation into an opportunity. Adaptation often takes place on multiple layers, as it is rarely related to a single specific stressor, but rather reflects a broad combination of many.

Ultimately, as the disturbance becomes unbearable (both in terms of its intensity and persistence) and adaptation would lead to a too large change, transformative capacity is the way forward. This transformation can be both the outcome of a deliberate decision and action of agents, like a regime change through a democratic election process, or a change forced by environmental or socio-economic conditions. The main feature of transformative capacity is that it does not only include technical and technological changes, but also cultural changes, behavioural shifts and institutional reforms. Transformative resilience can be defined as the means of learning from past events and engineering changes, ideally towards a better condition given current constraints. Such a shift of the status quo may nevertheless be difficult to achieve.6

In real situations, different agents might experience the two dimensions differently. Moreover, disturbances seldom have a single channel of transmission; instead, they tend to originate from a chain of events and consequences, and trigger multiplicative effects. This means that the three types of capacity often act simultaneously, at multiple levels (individuals, community, regional, country, institutions) and with potentially different intensity at different levels. In other words, they are different perspectives of the same reality rather than opposing or competing components.

In this context,7 a society is resilient if, when facing shocks or persistent structural changes, it keeps its ability to deliver individual and societal well-being in an inter-generationally fair way, i.e. ensuring current well-being without seriously compromising that of future generations. Absorptive and adaptive capacity means that, despite some initial inevitable losses after a shock, a resilient society tends to return to its original level of well-being and functionality, and potentially move to a better one. When the situation becomes unbearable and a transformation is necessary, the original level of well-being and functionality can no longer be sustained; however, these transformations should lead to a new, sustainable path, with acceptable levels of well-being.

This approach establishes a close link between resilience and sustainability, the former being the means to achieve the latter in a dynamic sense. While sustainability in the capital approach is about the quantity and value of the stock of the capital available (which acts as a buffer), a resilience approach focuses on the qualitative side, which in turn depends on many aspects of a “system” (diversity, the flow of assets, inter-connectedness). One way to think of this complementarity is that sustainability is the long-term design phase, while resilience is about reactive capacity, i.e. about managing imbalances and acting to keep or restore sustainability.

In the real world, where reaching a tipping point may determine “breaks” in some parts of the system, sustainability can become impossible because of non-linearities: for example, political institutions can become unsustainable because of a prolonged recession and decline in people’s standards of living. In this case a “revolution” might happen, leading to the collapse (a deep transformation) of the socio-economic system or to deep conflicts (a war, a civil war, etc.).

9.4.1. Layers and inter-dependences

One of the main implications of this approach is that resilience needs to be analysed in the context of sustainability by looking at the entire ecological-social-economic-political system. Such a general approach may have several “sectoral” applications: for example, focussing on resilience of the ecosystems for the benefit of our generation and of the generations to come should be at the centre of any long-term policy, such as the 2030 Agenda, no matter which specific economic or social policies are concerned.

In this perspective, the global system can be visualised as a “doughnut” with different layers: the economic, social and environmental layers, with an indication of the planetary boundaries (and showing where these boundaries have already been crossed), as well as an area of safe and just space for humanity (Figure 9.3). Not only are systems embedded in one another, but there are layers within each of them.

In this perspective, society consists of individuals, communities, regions, nation-states, supranational and international entities and humankind at large. The resilience of individuals should be considered in the context of resilience of communities, which in turn are embedded in regions and nation-states, etc. The concept of resilience goes hand in hand with the situation of a system being hit by disturbances. If the risk materialises, a system can be vulnerable or not, depending on the intensity of the shock and the properties of the system. A vulnerable system can recover with a contained social welfare loss or not.

Figure 9.3. A safe and just space for humanity
Figure 9.3. A safe and just space for humanity

Source: Raworth, K. (2012), “A safe and just space for humanity. Can we live within the doughnut?”, Oxfam Discussion Paper, February,, and Rockström, J. et al. (2009), “Planetary boundaries: Exploring the safe operating space for humanity”, Ecology and Society, Vol. 14(2), p. 32,

Resilience of systems should also be seen as inter-dependent with the people within those systems, as one might think of micro-, meso- and macro-economies. While at a macro-level, a country’s economy might be resilient to economic shocks, not all groups of people within the country might be resilient. So the analysis of macro-measures, such as GDP, might be misleading in the analysis of resilience if not accompanied with other socio-economic indicators and by in-depth analysis of vulnerable groups.

Improved measurement should be produced at each layer in order to understand their vulnerability and risks, but the links and interactions between all levels also need to be examined. The systems approach allows us to create different scenarios and estimate and demonstrate the related effects (similarly to stress-tests). The challenge consists in increasing our capacities to distinguish between dangerous situations and sustainable pathways in an uncertain context. This approach could also help to generate a baseline against which to estimate the cost of different types of shocks and the risks associated with them, as well as estimates of investments to be made to make the systems more resilient.

While recognising the limits of scenarios and forecasting, model results are important inputs in the design and implementation of policies and programmes for risk reduction and increasing resilience. These results could provide a framework for a public discourse about choices that have to be made as society moves towards sustainability, choices which might include trade-offs between the “now” and “tomorrow” as well as between the “here” and “elsewhere” dimensions of sustainability.

A practical example of a systems approach is given by Figure 9.4, which describes the impact of changes in water quantity and quality on different system layers.

Figure 9.4. The vulnerability to variations in water availability and quality
Figure 9.4. The vulnerability to variations in water availability and quality

Source: Sosa-Rodríguez, F.S. (2016), “An alternative framework for analyzing the vulnerability of socio-ecological systems”, Realidad, Datos y Espacio. Revista Internacional de Estadística y Geografía, Vol. 7(1), INEGI.

From an overall perspective, a starting point to understand how shocks spread among the different segments of the whole system, and where to intervene, is provided by the materially closed Earth system (Figure 9.5). Its three main ingredients are the inputs (the four types of capital stocks), the outputs (well-being and its determinants) and the engine (the overall “assembly” system) which translates inputs into outcomes and outputs. The final results of a system are ultimately determined by the outcomes, i.e. societal and individual well-being, while shocks typically affect the inputs (capital stocks), and then the effects interact in the assembly system. In some cases, the engine might be in distress, and the place where most of the policy interventions should occur.

Figure 9.5. Ingredients of resilience in the materially closed Earth system
Figure 9.5. Ingredients of resilience in the materially closed Earth system

Source: Manca, A.R., P. Benczur and E. Giovannini (2017), “Building a scientific narrative towards a more resilient EU society – Part 1: A conceptual framework”, JRC Science for Policy Report, No. 106265, Publications Office of the European Union, Luxembourg,, based on Costanza, R. et al. (1997), “The value of the world’s ecosystem services and natural capital”, Nature, Vol. 387, pp. 253-260.

With respect to measurement, this approach implies that we should concentrate on the three aspects:

  1. 1. Resilience of assets, to be measured in the context of the capital approach.

  2. 2. Resilience of the engine, referring to eco- and social system services and to institutions, production processes and their complex interactions. Measurement here is highly problematic, since it refers to the power of institutions to shape the production process, in a broad sense.

  3. 3. Resilience of outcomes/output, in terms of investment, consumption of goods and services, well-being and negative externalities such as pollution, social marginalisation or poverty.

9.4.2. Metrics for resilience, risks and uncertainties

A macro-prudential, system-wide approach in the sense described above does not yet exist either in policy or in statistical terms. Even in the SNA, and its extensions by the SEEA, a classical aggregation concept is used, rooted in the inventory and valuation of single capital goods. Nevertheless, it is possible to broadly outline the main conceptual components and procedural steps that would be necessary to explore and develop in detail a complementary way of accounting for a system’s dynamics and resilience:

  • Scope and dimensions: Available knowledge in various scientific disciplines should be used to evaluate and quantify risks, i.e. threats for the resilience and the sustainability of economic, social and environmental systems. Priority should be given to the risks that are most relevant for sustainability, e.g. those that are pushing systems close to planetary boundaries, as defined by the scientific community. While micro-level accounting tends to undervalue natural and social capital, macro-level accounting can capture systemic interactions between environment, society and economy.

  • Quantification: Geographical Information Systems (GIS), accounting methods and indicator systems (dashboards etc.) should be combined to achieve the best possible and most far-reaching condensed presentation of the major risks (current, emerging).

  • Aggregation, valuation: The actual price system does not work well with complex and/or systemic risks. Actuarial expertise (scientists or practitioners) used to estimate “premiums” necessary to insure the major risks could provide valuable inputs for this exercise.

  • Scenarios: These might be used to show the dynamic evolution of sustainability over time. A good example is old-age pensions. A society may confront a large stock of pension entitlements by only one cohort, obligations which will be costly to meet for some years but after which the system stabilises. A policy action might be needed to deal with short-term problems, but possibly a different one from that implied by a large stock of pension obligations towards all future cohorts. Inter-generational accounting models, which typically focus only on government finances, might be used to show that large fiscal deficits in the future could be met by higher taxes or though other ways of shifting the burden to households, in particular when households have low debt and high assets; private debt is already co-analysed with government debt in the context of the EU Micro-Imbalances Procedure (MIP).

  • Communication: It is important to integrate all societal stakeholders (science, civil society, business, policy) from the very early stages in generating knowledge of this kind. New metrics generated using this procedure should in particular facilitate a democratic dialogue. As a consequence, the processes of measurement and political discourse should be seen as mutually dependent, and influencing each other. In this sense, new metrics, generated through new measurement processes, should be tailored and fit for specific purposes in the policy cycles.

Box 9.12. Vulnerability, poverty and resilience

This chapter has argued that it is important to assess the risk properties of the economic system – i.e. its exposure to risk, its vulnerability and its resilience. Changes in economic policy can have significant effects on risk-performance: increasing exposure to risk; making the economic system more vulnerable; reducing the capacity of individuals or other entities in the system to cope with risks; or making the system as a whole (or the units within it) less resilient. Some reforms may simultaneously improve average economic performance but reduce “risk-performance”. It is important not only to know when this is happening, but also to assess quantitatively the effects. If GDP growth increases but resilience decreases, we would want to know this. In some circumstances, a country might want some measure of resilience in its dashboard of key indicators.

While this is an area in which so far there has been limited progress – and it is an important arena for future research – some promising approaches include the following:

Vulnerability and poverty. When individuals move out of poverty, we would hope that that move is permanent. In reality, many of those who escape poverty often fall back in it again. Even those who have never been poor have a chance of falling into poverty. The threat of falling into poverty can loom large in the life of a person and other family members – it can be a source of anxiety that our national income statistics never pick up. One simple measure of vulnerability is the share of people who are not poor at any one date but may experience at least one year of poverty in the next five years (UNDP, 2014).

Resilience to economic phenomena. Vulnerability is a measure of the possibility of downward mobility. Resilience, by contrast, is a measure of “recovery”, i.e. how quickly (if ever) a family or an economy that experiences a negative shock recovers. There can, of course, be many measures of resilience: how fast it takes for a family that winds up in poverty to move out of poverty; or how fast on average it takes for an economy that experiences a negative shock to return to its pre-crisis level, or to the level that it could have attained in the absence of a crisis. At each level, it is important to know the determinants of resilience, i.e. what makes some families or economies more resilient than others. In the light of the systems approach, it is also important to look at resilience from a broad societal perspective, beyond simple income or output measures.

A striking aspect of the 2008/2009 crisis was that different countries experienced shocks of different magnitudes; by and large, the recovery has also been slower than for previous economic downturns, which is understandable given the magnitude of the shock. In the beginning, some commentators had expected a “V-shaped recovery”, with the economy quickly bouncing back; others thought, however, that the economy was less resilient, and that the recovery would be “U shaped”. The latter perspective proved right, and in the following years the debate was about how long the flat bottom of the U would last. These experiences highlight that an economy can be resilient with respect to small shocks, and not with respect to big shocks.

Money-equivalent measures. This chapter has described the economy as a dynamic sub-system, connected to social and environmental sub-systems. In assessing changes in the economic system, we can measure its overall risk performance in a way similar to the Atkinson and Stiglitz measures for inequality and to the Arrow and Pratt measures for risk: how much society would be willing to pay to avoid the systemic risk that it confronts. Such a measure would compound the risk properties of the system as a whole and the aversion to risk of society.1

The chapter has described the properties of systems which affect the size of systemic risks. For instance, better automatic stabilisers could make the economic system more resilient – it will more quickly recover from an adverse shock. Thus, for a given degree of risk aversion, a more resilient economic system – one that recovers more quickly from an adverse shock – would presumably lower the systemic money-equivalence of the risk. This measure could provide guidance on the value that should be assigned to the risk aspects of various economic reforms.2 For example, a move from a defined benefit to a defined contribution pension scheme could weaken automatic stabilisers since individuals are more exposed to business-cycle risks; in this situation, such a measure might provide some guidance to how much “better” in some other way the defined contribution system has to be compared to the defined benefit to offset the loss is systemic stability.

1. The discussions of inequality and of risk highlighted the importance of money-equivalent measures. These measures ask how much a person would be willing to pay to avoid some risk, or society to avoid of inequality. But in economics, we typically think of matters at the margin, how much we are willing to pay to get rid of small amount of risk or inequality. In evaluating a new policy, we may ask what is the incremental value of the reduction in risk or inequality compared to the status quo baseline. Stiglitz has described such a marginal measure for income inequality (Stiglitz, 2015b).

2. There have been many attempts to measure individual risk aversion by looking at individuals’ behaviour towards risk – how much they are seemingly willing to pay to reduce the risks that they confront.

9.5. Conclusions

Since the Stiglitz, Sen and Fitoussi (2009) report, a substantial amount of work has been carried out on measuring progress towards sustainable development, based on different models and approaches and in different geographical settings. However further work is needed. Directions for future work include the following.

9.5.1. On the capital approach

Economic capital

  1. 1. Distinguish between nominal wealth and the quantity of productive capital – data should be collected and displayed in such a way that the volume of productive capital is not obscured by revaluation effects. Also, the scope of nominal wealth and productive capital may be different and should be distinguished (for instance, nominal wealth includes net foreign financial assets, productive capital does not).

  2. 2. Use a balance sheet approach to help assess economic sustainability for all institutional sectors (e.g. banks, non-financial corporations, households) rather than for the government alone; and focus on both liabilities and assets, recognising that fire-sales of assets in depressed financial markets will worsen net worth.

  3. 3. Improve fiscal modelling to incorporate demographic evolution and more generally engage in research on how much further to take simultaneity into account.

Human capital

  1. 1. Increase efforts to understand and measure human capital stock and its formation process. In particular, skills (cognitive and non-cognitive) and other components of human capital, and approaches to their measurement, need further discussion and analysis.

  2. 2. Develop satellite accounts for human capital using the cost approach together with more detailed non-monetary indicators. In practice, satellite accounts for education and training should be a main building block. Coverage of non-formal education processes (e.g. on-the-job training) is important.

  3. 3. Pursue research on the income-based approach to provide more complete information, in particular on labour earnings from both market and non-market activity.

  4. 4. Increase efforts to understand and measure the non-economic returns to human capital.

  5. 5. Rethink how the SNA treats public and private expenditures in human capital in view of capitalising them.

  6. 6. Further explore the links between human capital and social capital.

Natural capital

  1. 1. Improve measurement of environmental assets, including land and ecosystems (e.g. the extent and condition of different types, the services they provide, etc.).

  2. 2. Calculate and communicate at least annually how much carbon space is left before reaching potential “tipping points”.

  3. 3. National Statistical Offices should apply the SEEA, and produce timely estimates of “resource productivity” and the “circular economy”.

  4. 4. Improve the timeliness of indicators and accounts for natural capital applying the same now-casting techniques already used for economic variables like GDP.

9.5.2. On the systems approach

Stretching the boundaries of research

Interest in this complementary way of accounting for systems’ resilience is growing as shown by initiatives taken in the context of the European Commission’s 7th Environmental Action Programme (EAP, European Parliament and Council, 2013), the 2014 United Nations Development Programme (UNDP) framework, the European Environment State and Outlook 2015 (SOER report, European Environment Agency, 2015), or the Joint Communication on “A Strategic Approach to Resilience in the EU’s External Action” (European Commission, 2017).

While the maturity of research in this field is not yet comparable with other fields of research included in this report, multi-disciplinary co-operation should be enhanced. A macro-prudential approach to sustainability policies needs to be underpinned with high-quality information, best fit for this purpose. In this regard, the two workshops on resilience organised by the HLEG can only be seen as starting points. While having been necessary and fruitful for the first collection of ideas and questions, they exemplified the difficulties to overcome “silo” mentalities and cultures of scientific disciplines and to merge expertise from all “camps” in one broader programme.

The Stiglitz, Sen and Fitoussi (2009) report has shown that it is possible to breach the boundaries and traditional ways of thinking and thereby achieve essential progress. Similar progress should be made in the field of resilience by inviting researchers to contribute to this major set of questions.


  1. 1. Improve measurement of resilience so as to better understand vulnerability and risk at each level and across all dimensions, while also examining the links and interactions between all levels, and dynamic properties of the system. The international statistical community should establish a taskforce on the measurement of sustainability using the systems approach.

  2. 2. Further explore and document the complementarity of both the capital and the systems approaches, liaising theoretical considerations with empirical information.

  3. 3. Improve estimates and communication of risk and resilience to all stakeholders.

  4. 4. Involve various disciplines and assure horizontal co-operation to lay the basis of a special education path for sustainable development.

  5. 5. Introduce standardised terms and variables, which can serve as “ideal types” (in the sense of Max Weber), so that they fulfil expectations from both theoretical and empirical sides, thus helping to produce statistical metrics with high quality.

A new impetus

The establishment of the 17 SDGs and 169 targets has given a new impetus to the development of common sustainable development indicators, and to the scientific work needed to underpin these indicators, so that progress towards sustainable development could be traced at global level, across countries and regions, in a reliable and timely manner. New measurement initiatives should take into account developments in statistical methodologies such as making use of big data or other approaches as proposed by the Global Conference on a Transformative Agenda for Official Statistics8 and in the report by the Independent Advisory Group to the UN Secretary-General (United Nations, 2014).

However, in a world of complex systems, sustainability cannot be measured in full: there are limits to the measurement and only part of the knowledge is available. According to good democratic principles, both the knowledge available – monetary and physical indicators as well as models such as for considering resilience of systems – and the areas where knowledge is missing should be communicated in a correct way to all stakeholders so as to show possibilities as well as limits for governance.

More investment is also needed to develop analytical models at global scale to evaluate future scenarios and the impact of alternative policies for sustainable development adopting a systems approach. Transboundary effects of policies and the global interactions of economic, social and environmental phenomena can only be addressed by analytical models. Development of these models will require better and more-timely data from the international statistical system. A more continuous and fruitful dialogue, at global scale, between scientists and statisticians has to be established as soon as possible.


Adams, W.M. (2006), The Future of Sustainability – Rethinking Environment and Development in the Twenty-First Century, Report of the IUCN Renowned Thinkers Meeting, 29-31 January 2006, IUCN,

Algan, Y., P. Cahuc and A. Shleifer (2013), “Teaching practices and social capital”, The American Economic Journal: Applied Economics, Vol. 5(3), pp. 189-210.

Beck, U. (1992), Risk Society: Towards a New Modernity, SAGE publications Ltd, New Delhi.

Berkes, F. and Jolly, D. (2001), “Adapting to climate change: Social-ecological resilience in a Canadian western Arctic community”, Conservation Ecology, Vol. 5(2), p. 18,

Berman, R., C. Quinn and J. Paavola (2012), “The role of institutions in the transformation of coping capacity to sustainable adaptive capacity”, Environmental Development, Vol. 2(April), pp. 86-100.

Boarini, R., M. Mira d’Ercole and G. Liu (2012), “Approaches to measuring the stock of human capital: A review of country practices”, OECD Statistics Working Papers, No. 2012/04, OECD Publishing, Paris,

Borio, C. (2009), “Implementing the macroprudential approach to financial regulation and supervision”, The Financial Stability Review, Bank of France, Vol. 13, pp. 31-41.

Brundtland, G.H. (1987), Our Common Future, Oxford University Press, Oxford.

Chambers, R. (1989), “Editorial introduction: Vulnerability, coping and policy”, The IDS Bulletin, Vol. 20(2), pp. 1-7.

Costanza, R. et al. (2014), An Introduction to Ecological Economics, Second Edition, CRC Press.

Costanza, R. et al. (1997), “The value of the world’s ecosystem services and natural capital”, Nature, Vol. 387, pp. 253-260.

Cumming, G. and J. Norberg (2008), “Scale and complex systems”, in Norberg, J. and G. Cumming (eds.), Complexity Theory for a Sustainable Future, Columbia University Press, New York.

Devereux, S. and R. Sabates-Wheeler (2004), “Transformative social protection”, IDS Working Paper, No. 232, Institute of Development Studies, Brighton,

Diewert, W.E. and P. Schreyer (2008), “Capital Measurement”, in Durlauf S.N. and L.E. Blume (eds.), The New Palgrave Dictionary of Economics, Palgrave Macmillan, London.

El Serafy, S. (2013), “Herman Daly Festschrift: Hicksian income, welfare, and the steady state”, The Encyclopedia of Earth, (accessed on 31 May 2018).

European Commission (2017), “A strategic approach to resilience in the EU’s external action”, Joint Communication to the European Parliament and the Council,

European Commission (2016a), “Next steps for a sustainable European future – European action for sustainability”, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2016) 739,

European Commission (2016b), “Proposal for a new European consensus on development our world, our dignity, our future”, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2016) 740,

European Parliament and Council (2013), “Living well, within the limits of our planet”, 7th EAP – The new general union environment action programme to 2020,

European Environment Agency (2015), The European Environment – State and Outlook 2015, SOER synthesis report, Copenhagen,

European Environment Agency (1999), Environmental Indicators: Typology and Overview, Technical Report No. 25/1999, prepared by Smeets E. and R. Weterings, TNO Centre for Strategy, Technology and Policy, The Netherlands.

European Statistical System (2011), Final Report on the Sponsorship Group on Measuring Progress, Well-Being and Sustainable Development,

Eurostat (2016a), “Sustainable Development Indicators”,

Eurostat (2016b), Sustainable Development in the European UnionA Statistical Glance from the Viewpoint of the UN Sustainable Development Goals, Eurostat Statistical Books,

Eurostat (2003), “Household production and consumption: Proposal for a methodology of household satellite accounts”, European Commission Working Papers and Studies,

Fiksel, J. (2006), “Sustainability and resilience: Toward a systems approach”, Sustainability: Science, Practice and Policy, Vol. 2(2), pp. 14-21,

Fleurbaey, M. and D. Blanchet (2013), Beyond GDP: Measuring Welfare and Assessing Sustainability, Oxford University Press, Oxford.

Folke, C. et al. (2002), “Resilience and sustainable development: Building adaptive capacity in a world of transformations”, AMBIO: A Journal of the Human Environment, Vol. 31(5), pp. 437-440.

Galaz, V. et al. (2016), “Planetary boundaries: Governing emerging risks and opportunities”, Solutions, Vol. 7(3), pp. 46-54.

Gallopin, G. (2003), “A systems approach to sustainability and sustainable development”, Publications of the Economic Commission for Latin America and the Caribbean (ECLAC), Santiago de Chile.

Garavito-Bermúdez, D., C. Lundholm and B. Crona (2016), “Linking a conceptual framework on systems thinking with experiential knowledge”, Environmental Education Research, Vol. 22, pp. 89-110.

Hanson S.G., A.K. Kashyap and J.C. Stein (2011), “A macroprudential approach to financial regulation”, The Journal of Economic Perspectives, Vol. 25(1).

Heal, G.M. (ed.) (1974), The Review of Economic Studies: Symposium on the Economics of Exhaustible Resources, The Society for Economic Analysis Limited, Longman, Edinburgh.

Hicks, J.R. (1939), Value and Capital: An Inquiry into Some Fundamental Principles of Economic Theory, Clarendon Press, Oxford.

Hjorth P. and A. Bagheri (2006), “Navigating towards sustainable development: A system dynamics approach”, Futures, Vol. 38, pp. 74-92.

Holling, C.S. (2001), “Understanding the complexity of economic, ecological and social systems”, Ecosystems, Vol. 4, pp. 390-405.

IPCC (2015), IPCC Assessment reports,

Jorgenson, D.W. (1963), “Capital theory and investment behaviour”, The American Economic Review, Vol. 53, pp. 247-259.

Jorgenson, D.W. and B.M. Fraumeni (1992a), “Investment in education and U.S. economic growth”, in The Scandinavian Journal of Economics, Vol. 94, Supplement, pp. 51-70.

Jorgenson, D.W. and B.M. Fraumeni (1992b), “The output of the education sector”, in Zvi Griliches (ed.), Output Measurement in the Service Sectors, The University of Chicago Press, Chicago.

Jorgenson, D.W. and B.M. Fraumeni (1989), “The accumulation of human and nonhuman capital, 1948-1984”, in Lipsey R.E. and H.S. Tice (eds.), The Measurement of Savings, Investment and Wealth Studies, The University of Chicago Press, Chicago.

Jorgenson, D.W. and Z. Griliches (1967), “The explanation of productivity change”, The Review of Economic Studies, Vol. 34(3), pp. 249-83.

Jorgenson, D.W. and J.S. Landefeld (2007), “Blueprint for expanded and integrated U.S. accounts: Review, assessment, and next steps”, in Jorgenson, D., J.S. Landefeld and W. Nordhaus (eds.), A New Architecture for the U.S. National Account, The University of Chicago Press, Chicago.

Kendrick J.W. (1976), The Formation and Stock of Total Capital, Columbia University Press, New York.

Klenow, P.J. and A. Rodríguez-Clare (1997), “The neoclassical revival in growth economics: Has it gone too far?”, in Bernanke B.S. and J. Rotemberg (eds.), NBER Macroeconomics Annual 1997, MIT Press, Boston, MA,

Liu, G. (2011), “Measuring the stock of human capital for comparative analysis: An application of the lifetime income approach to selected countries”, OECD Statistics Working Papers, No. 2011/06, OECD Publishing, Paris,

Lupton, D. (2013), Risk, Routledge, New York.

Manca, A.R., P. Benczur and E. Giovannini (2017), “Building a scientific narrative towards a more resilient EU society – Part 1: A conceptual framework”, JRC Science for Policy Report, No. 106265, Publications Office of the European Union, Luxembourg,

Nyamu-Musembi C. and A. Cornwall (2004), “What is the “rights-based approach” all about? Perspectives from International development agencies”, IDS Working Paper, No. 234, Institute of Development Studies, Brighton,

OECD (2016a), Education at a Glance 2016: OECD Indicators, OECD Publishing, Paris,

OECD (2016b), OECD and the Sustainable Development Goals: Delivering on universal goals and targets,

OECD (2015), Education at a Glance 2015: OECD Indicators, OECD Publishing, Paris,

OECD (2010), The High Cost of Low Educational Performance: The Long-run Economic Impact of Improving PISA Outcomes, OECD Publishing, Paris,

OECD (2009), Measuring Capital – OECD Manual 2009: Second edition, OECD Publishing, Paris,

OECD (2001), The Well-being of Nations: The Role of Human and Social Capital, OECD Publishing, Paris,

Perrings, C. (2006), “Resilience and sustainable development”, Environment and Development Economics, Vol. 11(4), pp. 417-427.

Piketty, T. and G. Zucman (2013), “Capital is back: Wealth-income ratios in rich countries, 1700-2010”, The Quarterly Journal of Economics, Vol. 129(3), pp. 1255-1310.

Putnam, R. (2000), Bowling Alone: The Collapse and Revival of American Community, Simon and Schuster, New York.

Radermacher, W.J. and A. Steurer (2015), “Do we need natural capital accounts for measuring the performance of societies towards sustainable development, and if so, which ones?”, Eurostat Review on National Accounts and Macroeconomic Indicators, Vol. 2015/1, pp. 7-18.

Rapport, D.J. and A. Friend (1979), Towards a Comprehensive Framework for Environmental Statistics: A Stress-Response Approach, Statistics Canada, Ottawa.

Raworth, K. (2012), “A safe and just space for humanity. Can we live within the doughnut?”, Oxfam Discussion Paper, February,

Rockström, J. et al. (2009), “Planetary boundaries: Exploring the safe operating space for humanity”, Ecology and Society, Vol. 14(2), p. 32,

Rudinow Saetnan, A., H. Mork Lomell and S. Hammer (eds.) (2012), “By the very act of counting – The mutual construction of statistics and society”, in The Mutual Construction of Statistics and Society, Routledge Chapman & Hall, London.

Sachs R. (2014), “Emerging Risk Discussion Paper: Risk transfer and uncertainty”, Discussion paper for the EEA Scientific Committee Seminar on emerging Systemic Risks, Copenhagen, 24 February,

Seccombe, K. (2002), “‘Beating the odds’ versus ‘changing the odds’: Poverty, resilience, and family policy”, The Journal of Marriage and Family, Vol. 64(2), pp. 384-394.

Shaw, R. et al. (1991), “Sustainable development: A systems approach. Final Report”, International Institute for Applied Systems Analysis, Laxenburg, Austria.

Solow, R.M. (1986), “On the intertemporal allocation of natural resources”, The Scandinavian Journal of Economics, Vol. 88, pp. 141-149.

Sosa-Rodríguez, F.S. (2016), “An alternative framework for analyzing the vulnerability of socio-ecological systems”, Realidad, Datos y Espacio. Revista Internacional de Estadística y Geografía, Vol. 7(1), INEGI.

Stiglitz, J.E. (2015a), “New theoretical perspectives on the distribution of income and wealth among individuals: Part I. The wealth residual”, NBER Working Paper, No. 21189,

Stiglitz, J.E. (2015b), “Inequality and Economic Growth”, The Political Quarterly, 86, pp. 134–155

Stiglitz, J.E. (1979), “The use of prudential measures in the international banking markets”, BISA, Vol. 7.18(15), Papers Lamfalussy, LAM25/F67, pp. 1-2.

Stiglitz, J.E., A. Sen and J.-P. Fitoussi (2009), Report by the Commission on the Measurement of Economic and Social Progress,

UNECE (2017), Guide on Valuing Unpaid Household Service Work, United Nations, Geneva,

UNECE (2016), Guide on Measuring Human Capital, United Nations, Geneva,

UNECE (2014), “Conference of European Statisticians recommendations on measuring sustainable development”, Geneva,

UNESCO (2016), “Gender review – Creating sustainable futures for all”, EFA Global Monitoring Report 2016,

UNESCO (2015), “Teaching and learning – Achieving quality for all”, EFA Global Monitoring Report 2013/4,

UNEP and UN-IHDP (2014), Inclusive Wealth Report 2014: Measuring Progress Towards Sustainability, UN University – International Human Dimensions Programme on Global Environmental Change and UN Environment Programme, Delhi.

UNEP and UN-IHDP (2012), Inclusive Wealth Report 2012: Measuring Progress Towards Sustainability, UN University – International Human Dimensions Programme on Global Environmental Change and UN Environment Programme, Delhi.

United Nations (2015a), “Transforming our world: The 2030 agenda for sustainable development”, Resolution 70/1 of the UN General Assembly,

United Nations (2015b), “The World Women’s 2015 – Trends and Statistics”,

United Nations (2014), A World that Counts. Mobilising the Data Revolution for Sustainable Development, Report prepared by the Independent Expert Advisory Group on a Data Revolution for Sustainable Development (IEAG), New York,

United Nations (2000), “Household accounting: Experience in concepts and compilation, Vol. 2, Household Satellite Extensions”, in Handbook of National Accounting, Studies in Methods, Series F, No. 75/Vol.2,

United Nations et al. (2014a), System of Environmental-Economic Accounting 2012 Central Framework,

United Nations et al. (2014b), System of Environmental-Economic Accounting 2012 Experimental Ecosystem Accounting,

UNDP (2014), Human Development Report 2014 – Sustaining Human Progress: Reducing Vulnerabilities and Building Resilience, United Nations Development Programme, New York,

UN Women (2015), Summary Report: The Beijing Declaration and Platform for Action Turns 20, New York,

Walshe, R. (2014), Concept paper: A new systems approach to resilience”, Climate Exchange, (Accessed 4 June 2018).

World Bank (2011), The Changing Wealth of Nations: Measuring Sustainable Development in the New Millenium, The World Bank, Washington, D.C.,

World Bank (2006), Where is the Wealth of Nations? Measuring Capital for the 21st Century, The World Bank, Washington, D.C.,


← 1. As is the case in the recommendations of the Conference of European Statisticians (UNECE, 2014).

← 2.

← 3. Previous definitions of human capital by the OECD differed in that they referred to economic well-being only. Social capital is considered in the chapter Trust and Social capital of this book.

← 4. However, some of the strong assumptions of the lifetime income approach to calculating human capital are not very appealing. For example, “rates of return to schooling” typically capture only labour market returns, while they should ideally be extended to capture non-monetary returns (such as the longer life expectancies of better educated people, though there is debate over causality in this relationship). However, capturing non-monetary returns requires valuing them, which is frequently done using income. This is problematic, because such an approach might lead to the conclusion, with respect to life expectancy, that lives in poor countries are worth less than lives in rich countries, and by implication that the returns to human capital are lower. This approach is also less useful to explain future growth of GDP and productivity because the (real) service flows from that stock are themselves a function of the income streams expected in the future.

← 5. Adjusting the 3P + T framework of social protection (Devereux and Sabates-Wheeler (2004) to a broader resilience framework.

← 6. For example, the European Union faced a massive flow of migrants and refugees from 2015-16 onwards. This is likely to have a significant impact on the composition of European society in the coming decades. The ability of the EU to adapt to the new situation, and to transform itself through the integration of non-native European citizens, will be key to avoiding a massive socio-economic crisis and instead build a strong cohesive society.

← 7. This interpretation of resilience draws on Manca, Benczur and Giovannini (2017).

← 8.

End of the section – Back to iLibrary publication page