2. How to Monitor an Economy’s Contribution to Meeting Basic Needs

Inclusion in the list of economic outcomes for the purpose of monitoring is subject to specific criteria: as emphasised throughout this book, they result from outputs of the economy, and they enhance the quality of life by meeting basic needs. This accords with the twin requirements, (a) to fit in with the “division of labour” among different types of indicator as depicted in Fig. 1 (Chapter 1), so as to put the monitoring of the economy, society and environment on a clear and rigorous footing, and (b) to be restricted to the economy in the same way that GDP is, i.e. to occupy the same “space”, but change the evaluation criterion from means to ends.

Among the many indicators that already exist, none fulfils these criteria. Some are designed to cover one part of the “division of labour”, as with assets that are measured as stocks, measures of output (GDP) and subjective wellbeing. Others are heterogeneous—either GDP with various additions and subtractions, or heterogeneous aggregates (“composite” indicators). They are discussed in Chapter 3.

The relationships of the economic outcome indicators with upstream causes in the economy and with downstream impacts on health and wellbeing are empirical issues. The main criterion for inclusion as an economic outcome indicator is the empirical demonstration of the causal importance of these upstream and downstream links, and (ideally) their quantitative estimation. The technical aspects of this are discussed in Chapter 5.

The status of economic outcome measures as intermediate outcomes between the economy and health/wellbeing can be illuminated by an analogy from the medical field. High blood pressure has dietary and other causes, and indicates the risk of future cardiovascular disease. It is valuable in medical practice, because treatment reduces the risk. It also has useful measurement properties: it is readily measurable for a whole population, and can therefore be used to monitor the effectiveness of, for example, a population intervention to reduce salt intake. In contrast, adverse health consequences such as strokes apply to only a proportion of the population, making them more difficult to study statistically.

Strokes may also take many years to manifest themselves, implying that monitoring and research on the intermediate outcome can be carried out without requiring a long wait for the diseases to start appearing, thereby providing an early warning system. This also means that preventive intervention can be undertaken on the high blood pressure, rather than waiting for serious disease to occur and then reacting to it.

Monitoring intermediate outcomes has one more important methodological advantage. Another health-related example may help to illustrate this. It is universally considered that the proportion of infants dying around the time of birth—the perinatal mortality rate—should be as low as achievable. An implication is sometimes drawn that it should be monitored. But this only indicates the size of the problem in a particular population, not the reason why it is at that level. It is more useful to monitor the determinants of the perinatal mortality rate, and especially whether the interventions that are known to reduce it are being universally carried out (cf. Dasgupta 2001, Sect. 3.1, in the context of protecting the natural environment). Action can then be taken to ensure that everything possible is being done to minimise perinatal mortality.

The same three advantages apply to the monitoring of the economy, because economic outcomes are intermediate between outputs and the impact on health and subjective wellbeing. First, they apply across the population—e.g. everyone has a housing status that can be specified, from secure ownership to precariousness and outright rough sleeping, whereas (for example) anxiety due to insecure housing is spread unevenly and would be difficult to monitor. Second, the precariousness occurs before the anxiety, so it is an early warning sign, which also implies that intervention can be made at an early stage, i.e. it is preventive, rather than reactive at a late stage when only patching up the damage is possible. And thirdly, knowledge about intermediate outcomes gives valuable information on the parts of the economy that need intervention, enabling a positive policy agenda to be developed.

Economic outcome measures are designed to assess the various ways that the economy affects individuals and households. This has two aspects, or stages: the usefulness of a particular good or service to the recipient, and the wider consequences—the attributable impact—which refers to the difference that this makes to their quality of life (see Fig. 1 in Chapter 1). These are the links that connect economic output with health and wellbeing. (A third aspect, consequential gain, is omitted from Fig. 1, because the primary focus here is on these outcomes as valuable in themselves, rather than as antecedents of something else. It is also deliberately under-emphasised in this book, in order to concentrate on economic outcomes as ends not means.)

For example, literacy (in adulthood) is extremely useful: it enables a person to navigate their own locality and other places with relative ease, to communicate using various forms of written communication, to enjoy reading for its own sake, to read to their children, etc. The attributable impacts include the ability to function well in society—perhaps most readily seen in terms of its counterfactual, the substantial problems that adults face when they have little or no ability to read and write. (Literacy also has important consequential gain, including an enormous expansion in employment prospects.)

Similarly, having a home that is secure, safe and sufficiently spacious is a reliable basis for continuity in work and school life, and its attributable impacts include an absence of anxiety, as well as a feeling of belonging. Proximity to one’s family and community is practically useful when a crisis occurs, and in addition it has the attributable impact of the emotional benefits of close social contact.

To provide a clearer idea of what is being suggested, a provisional list of possible indicators for high-income countries is presented in Table 2.​1. Some are listed as negative attributes, e.g. Unemployment and Homelessness, while others appear in positive terms as with Security of livelihood and Adult literacy; it should be obvious which is which. Forty indicators are included, which are tentatively proposed as suitable for rich countries. Data are already collected on most of these, although they may not currently be in the appropriate format.

The choice of items is based not only on the existing literature on human needs, as outlined in Chapter 1, but also on the extensive evidence base on the health impact of the proposed indicators. In this context, “health” refers not to healthcare services nor to health-related behaviour, but to health status as a consequence of exposures in the broadest sense—the social (or wider) determinants of health. A rich literature on this has been developed in recent decades (see, e.g., Marmot 2010; Braveman 2023), and further evidence is available from the economics of health (e.g. Currie 2020). Similar, albeit still relatively sparse, evidence on the determinants of subjective wellbeing is also available (Frijters et al. 2020; Allin 2022; What Works Centre for Wellbeing n.d.), and rapid progress is being made on this topic. More information on these literatures is presented in Chapter 3. It is no coincidence that the items in this list tend to be necessities (Spacey 2017), and are also the issues that repeatedly occur in public discussion, particularly in relation to the various types of deprivation that exist in most societies.

As already explained in Chapter 1, the IEO is intended to replace GDP as a measure of economic success, and therefore only includes outcomes that principally result from outputs of the economy. This excludes such important items as the quality of social relationships, social trust, or feeling respected—although these can be strongly influenced by what is happening in the economy. Similarly, the ability to participate in civic affairs, the absence of corruption and the absence of discrimination are best regarded as in a sphere separate from the economy. All of these can affect people’s health and/or wellbeing, but they are not part of the IEO, which explicitly relates only to the economy. That is the reason why an additional indicator is required, for the important aspects of society that do not belong to the economy, such as governance and personal safety. It is quite possible that a country’s economic outcomes could improve while these other aspects of life are deteriorating.

These distinctions are not clear cut, and could be disputed. For example, arguments could be put forward both for and against the inclusion of the criminal justice system and crime/safety. Personal security (e.g. feeling safe when walking alone at night) and the incarceration and reoffending rates could be regarded as outputs of the economy, and as candidates for inclusion in the IEO. I take the view that these are not part of the economy, even though they are strongly influenced by it—for example, they largely arise from inequalities and deprivation, and from economic outcomes such as lack of literacy.

Comparability across countries and across time is a valuable feature of GDP. It would similarly be desirable for standardisation to be applied to the IEO and its component economic outcome measures. However, different parts of the world have extremely different economies, and their abilities to provide for their residents vary correspondingly. It might be advisable to develop separate sets for groups of countries at different economic levels to reflect this. There could be separate divisions that reflect current economic levels, with the possibility of promotion to a higher division when appropriate. The lists would overlap, e.g. homelessness and job insecurity exist across a wide range of national income levels.

It is convenient to group the indicators according to the sector that they primarily relate to, as has been done in Table 2.1. However, it is important to recognise that they generally depend on factors outside, as well as within, that sector. The economy is not a unicausal system; multiple causes are operative for most phenomena. Thus, the grouping of outcomes under the heading of a particular sector should not be taken (or criticised) as suggesting that other influences are unimportant. For example, Access to good-quality early years education and childcare could be in the “Education” or in the “Caring” categories with equal validity. And Commuting time is listed in the Transport sector, but is affected by the proximity of employment and residence, not only the availability of transport options. Homelessness is particularly complicated in this respect; it is listed in the “Housing/shelter” category, but as is well known, is often associated with mental health and other issues. This reinforces the inter-connectedness of these various outcomes. One advantage of the wider IEO perspective over the current fragmented system of silos relating to specific policy areas is that a shared target might help to overcome the notorious problem of coordinated action across policy areas.

In addition, some outcomes are not sector specific, such as the availability, source and stability of one’s livelihood, or the ability to find some extra money in an emergency. These are grouped under the headings “Work” and “Livelihood/income” in the table. Many of these correspond to economic indicators that are already widely used at the national level, e.g. the unemployment rate and the extent of household debt.

Many of the indicators are expressed in terms of “access”, for example healthcare access. This is deliberately a combined assessment of availability and affordability. The focus is on whether the outcome is achieved or not, rather than how it comes to be achieved. This could be through state provision, or privately provided in which case the individual would have to be able to afford it. In any case, the relevant facilities would have to be physically present; not inaccessible because of disability, discrimination, ineligibility or distance; and knowledge of their existence would have to be universally available.

This multiple sourceability is important from a policy viewpoint, because it means that an outcome can be achieved by attention to affordability (including by raising income, e.g. through transfers) and/or to the way that provision is organised and funded. It also relates to measures that are nowadays taken by public authorities, to provide documents in large print for people with impaired vision or in multiple languages, and to make public spaces accessible to wheelchair users.

In some cases, “access” is somewhat more complicated. For example, the range of benefits listed under “Communication, information and entertainment” is obtainable via a smartphone. But this requires not only its affordability and physical availability, but also the ability to use it, sufficiently good vision and a reliable signal.

An implication of this focus is that income and prices are not seen as outcomes, and therefore are not included in the table. Income is a means to an end, an output not an outcome—the macro-equivalent is GDP as the sum of all incomes. This is not to say that it is of minor or subsidiary importance; it plays a large, often dominant, role in many of the items, because they depend on affordability. For example, most of the entries in the “Housing/shelter and energy” category depend on the ability to pay, as well as on other factors such as the availability of accommodation of different types in relation to the location of employment, etc. And a change in personal economic circumstances would not be invisible in the IEO: it would show up as a corresponding change in the range of outcomes that are able to be accessed. This also means that fundamentally monetary concepts such as fuel poverty (relating to the proportion of total expenditure that is on fuel) do not appear in the list of outcomes.

Similarly, time is a cross-cutting issue: the impact on economic outcomes of having access to a washing machine arises from the time it releases. This allows other outcomes to be achieved, for example by making paid employment possible.

Affordability and time are not the only cross-cutting issues. There is abundant and growing evidence for the harmful effect of insecurity on people’s quality of life (Stone and Krueger 2018, Box 7.7; Kapteyn 2020, p. 195 and the references therein). For example, job insecurity has as strong an impact as actual unemployment on mental and physical health as well as on subjective wellbeing (De Witte et al. 2016; Giunchi et al. 2019). Food and housing insecurity are among the best predictors of future healthcare expenditure (Fitzpatrick et al. 2015). Case and Deaton (2020, Chapter 11) attribute much of the recent rise in “deaths of despair” among less-educated white Americans to worsening job security, the replacement of good jobs in manufacturing by unsatisfying casual work. Irregularity of work (and therefore of income) also has indirect effects, e.g. via the possibility of obtaining a mortgage. Insecure housing tenure similarly has indirect as well as direct effects, undermining the stability of employment and schooling, etc. The importance of insecurity has also been recognised in the context of metrics for societal monitoring (Stiglitz et al. 2018).

More broadly, once relieved from insecurity, people are better able to focus on social relationships, enabling stronger social bonds to be created and maintained. And they are more able to enjoy work and other activities, and to be creative and innovative (Diener and Seligman 2004).

There are also further benefits. Less preoccupation with the necessaries of life releases psychological resources—“bandwidth”—that then become available for other activities (Schilbach et al. 2016). This allows scope for developing capacities and taking opportunities—in Sen’s terminology, it leads to an increase in functionings (Sen 2001), adding to human capital, and possibly also to social capital in the sense of trust. This would increase individuals’ work potential, with benefit to their households and to wider society, as well as to their own life chances—more consequential gain. Awareness of these prospects allows aspiration to develop. The fulfilment of necessaries therefore forms a foundation for other aspects of life that individuals have reason to value.

I propose that each indicator should be presented in the form of the proportion unfulfilled, i.e. the proportion of the resident population who lack the amenity specified in that indicator. Thus, Inability to find the equivalent of $400 (US) in an emergency would be the proportion who lack this degree of financial security, and Sufficient housing space would be presented in terms of the proportion whose living area is less than a specified threshold (e.g. UK Parker Morris standards). These measures, taken together, can be thought of as a multi-dimensional measure of deprivation (cf. Duclos and Tiberti 2016). Many existing measures take this form, for example the unemployment rate is the proportion of the working-age population who are without work, are available for work and have taken specific steps to find work. In other cases that are not naturally dichotomous, appropriate thresholds would need to be selected.

The proportion unfulfilled corresponds to the widespread value system, across ideological positions, that everyone’s basic needs should be met, as previously described—the commitment to leave no one behind. It directly measures the extent of unmet need. It is also related to inequality in the lower part of the income distribution. And it is readily understood by the general public, journalists, etc., not least because it can readily be brought together with qualitative evidence applying to the experience of actual people and households. It is therefore superior to, for example, the use of the mean or median plus a measure of dispersion. In addition, the proportion unfulfilled as a metric lends itself to aggregation to generate a summary index.

An extension of this approach would be to represent degrees of severity, using multiple thresholds for each item, thereby generating a variable with ordered categories. This finer-grain measure would preserve more information. Thus, it would be possible to have, for example, “degrees” of Security of tenure, and differentiation within Healthcare access—e.g. mental health facilities may be lacking or inadequate, especially for some age groups. This would provide more granular information, although it would make it less easily interpretable by the public, and would also complicate the aggregation procedure.

The economic outcomes would be published in two formats: individually in dashboard form, facilitating their use in public discussion and policy development, and as the aggregate IEO for the purpose of overall economic evaluation. Aggregation would ideally involve each item being weighted by its contribution to health and wellbeing, based on the available evidence. Details of the method of quantification and aggregation are given in Chapter 5. In practice, the evidence base on the health and wellbeing impacts of the economic outcomes is not currently robust enough to support these calculations. Reliance on expert judgement would probably be necessary in the initial stages, with an incremental process to bring the estimate gradually closer to the ideal situation, based on statistical evidence.

One of the problems with GDP is that most people are not clear how it is calculated, or what it really means. The proposed economic outcomes, as well as having intrinsic advantages, have the benefit that they are likely to be readily understandable by the public, journalists, etc. This would enable official statistics to be better integrated into a comprehensive information system for a democratic society (Allin and Hand 2021). It is true that the method of calculation of the aggregate IEO is necessarily highly technical, but its purpose of measuring the satisfaction of basic needs would be clear to all.

A focus on basic needs has the advantage that it incorporates relative disadvantage. This means that a separate indicator for the evaluation of inequalities would not be required, unlike with per capita GDP (and other measures such as wellbeing and life expectancy expressed as averages). Whereas economic growth, or increasing life expectancy, can occur while those at the lower end of the scale are left behind, the IEO increases only when the position of relatively disadvantaged people improves. It is thus sensitive to inequalities among the less fortunate.

On the other hand, it is agnostic about the wider distribution of wealth and income: the complete satisfaction of basic needs in a population is compatible with a wide degree of variation in the degree of inequality higher up the scale. An advantage is that people who positively value a substantial degree of economic inequality are just as able as egalitarians to support the economic outcomes perspective.

The focus on basic needs has an important implication for data collection. It is imperative that all residents of the population are represented. Some subpopulations are especially likely to experience deprivation of various kinds, especially if they do not have a secure home, and sampling therefore needs to be carefully designed to ensure that they are not excluded. In particular, “hard-to-reach” groups require particular attention, including people who do not live in private households, for example soldiers and prisoners.

The outcomes of economic activity are intrinsically valuable, and contribute to a fulfilling life. It is in addition possible to put a monetary value on them, or on their absence, as valuations of each outcome itself, and/or of its consequential gain. However, if this is done, there is a danger that the pecuniary valuation would come to dominate. The urge to provide a monetary value is understandable as it enables commensurability and assists when giving policy advice, but it implicitly rests on the argument that money is the measure of all things, and this is disputable.

Attention to the meeting of basic needs has featured strongly in the development literature and in international policy debates. The UN Sustainable Development Goals (United Nations 2015; United Nations n.d.) constitute the global benchmark of economic, environmental and societal monitoring (United Nations 2015; United Nations n.d.). They are designed to ensure that, as far as possible, basic needs are met throughout the world, while seeking to minimise environmental damage. Quite rightly, most attention has been focused on societies where lower income levels are prevalent. The Sustainable Development Goals include numerous items that are economic outcomes, listed below, as well as many other elements. The implication is that for economies at the lower income levels, economic outcome indicators would generally correspond with those already established by the Sustainable Development Goals process.

Many of the Sustainable Development Goals are economic outcomes, in the sense used in this book. They include:

The Goals are heterogeneous: as well as economic outcomes they include also assets (e.g. goal 15, Life on land); outputs (goal 8, Economic growth, which also includes sustainability and inclusion); and several rates of mortality and of specific diseases (goal 3); as well as non-economic items, e.g. Peace, justice and strong institutions (#16); action, as with Climate action (#13); and organisation as with Partnerships for the goals (#17).

The Sustainable Development Goals have had little influence on what is measured or publicly discussed in rich countries, where GDP still dominates. Similarly, basic needs do not receive systematic public and policy attention in the rich world; rather, they appear piecemeal in particular policy areas when they happen to become news items, where they have to compete on an agenda largely set by those with vested interests.

In rich economies, which form the focus of this book for the reasons given in Chapter 1, substantial pockets of deprivation exist, so that many people have great difficulty meeting their basic needs. At the same time, a great deal of spending is on excess or luxury consumption, on one’s position relative to others, on getting a transient buzz or on fear of missing out. Having “too much stuff” is widespread, and commercial storage facilities have mushroomed. A large proportion of consumption makes zero contribution to the satisfaction of basic needs, and more broadly has little positive impact on people’s quality of life.

This observation is backed up at the macro level by abundant evidence that as per capita GDP increases, its statistical relationship with measures of health and happiness disappears—there are diminishing returns (see Chapter 3). Further GDP growth would contribute little or nothing to the quality of people’s lives, unless the proceeds were to go to satisfying previously unmet basic needs. This would not be a serious problem were it not for its severe environmental consequences.

In contrast, in countries where material scarcity is widespread, GDP growth—if well distributed—can play an important role in allowing people to escape from grinding poverty. Even there, however, the use of per capita GDP as a measure of economic wellbeing can be misleading, as it is distorted by the affluence of the rich and by economic activities such as natural resource extraction that do not benefit most people. And as is well recognised, average per capita GDP obscures inequality in the distribution of prosperity.

An important feature of the Sustainable Development Goals is that they retain the specificity of the different topics, the same approach as is taken here. The tradition of converting data on basic needs to a single monetary figure, as with the World Bank’s extreme poverty line, loses the opportunity of using this information for public discussion and policy development in specific areas. In this context, UNDP with the University of Oxford has proposed the Global Multidimensional Poverty Index (MPI), measured by ten basic indicators, including adequate housing, child mortality, clean water, sanitation, cooking facilities and an electricity supply (UNDP and OPHI 2022; OPHI 2023). This retains the detail; the World Bank has adopted the MPI but combines it with monetary poverty as the Multidimensional Poverty Measure (MPM) [World Bank n.d.]. Monetary measures are also proposed in other contexts such as at the national level, e.g. Minimum Income Standards in the UK (Loughborough University Policy Unit 2023).

The official statistics systems of each country, as part of their public role, would have a leadership role in the design of the data collection and in its practical operation. Hopefully, they would collaborate to ensure that the indicators are internationally comparable, and coordination would be provided by international organisations such as the United Nations and/or the OECD.

It is vital to maintain a scope that is genuinely restricted to outcomes of the outputs of the economy and related policy, to guard against the danger of the list becoming unmanageably long and open ended. There should be relatively few indicators, so that users are not overwhelmed, and so that public debate is facilitated. The principles listed here are intended to provide the criteria that will enable this position to be upheld.

In addition, the margins of error of each measure should be calculated and made available. And provision should be made for regular statistical review, analogous to that involving the composition of a typical basket of commodities for the purpose of assessing inflation.

It is crucial that the selected indicators assess the actual outcomes, rather than being a proxy. This is not only to minimise gaming, as in Principle 8, but also to avoid the confusion of means with ends. A frequently used argument for promoting GDP growth is that ever-increasing consumption is needed in order to boost employment. In a finite world, this argument is dangerous. The solution proposed here is to consider the major aspects of employment—quality as well as quantity—as inherent parts of the desired economic outcomes. The creation of jobs is one of the outputs of the economy, and it has important consequences in terms of outcomes for the employees and from a societal viewpoint (Joffe 2011, 2018). Policy development should seek to improve labour prospects without resorting to unlimited consumption growth. The aim should be to directly improve the experience of working, rather than seeing employment as a side effect of economic growth.

The methods of data collection require development work. Nationally or internationally agreed indicators already exist for most of the suggested items. But in some cases development work, involving collaboration between statistical agencies, would be needed to put indicators into the standard format of assessing the proportion unfulfilled, and to decide on thresholds. This would build on pioneering work in many countries, such as the UK Index of Multiple Deprivation (Abel et al. 2016) and the European Deprivation Index (Launoy et al. 2018), the UK’s Measuring National Well-being Programme (ONS 2023) and especially the OECD’s Better Life Index (OECD n.d.).

The indicative list presented in Table 2.1 is intended to illustrate how the concept of economic outcome indicators could be made operational. It is for the purpose of discussion, and requires further work. This includes collation of the evidence (where available) on the strength of the causal relationships between each of the indicators and their health and/or wellbeing impacts, drawing on the rich literature on the social determinants of health, and the growing evidence on the causes of subjective wellbeing. This would provide information relevant to deciding on the inclusion of particular items, and on weights to inform the construction of the IEO as an aggregation of the individual outcome indicators, as described in Chapter 5.

Alongside this technical procedure, the development work would involve consultation with topic and sectoral experts, especially in the early stages, before a comprehensive evidence base becomes available in the correct format. Some experience in such collaboration has already been gained in the UK in the course of the work that followed the Atkinson Report. This could be supported by Structured Expert Elicitation, which treats expert judgement as scientific data in a methodologically transparent way with the aim of taking account of uncertainty (Hald et al. 2016).

The views of the general population could be sought on what people count as essential components of a good life, e.g. using questionnaires and/or focus groups (as used in constructing the Minimum Income Standard [Centre for Research in Social Policy n.d.]). Of particular interest would be to consult relatively deprived people, who might be expected to have special expertise in prioritising what really matters. The views of a representative population sample could also be valuable, albeit susceptible to the limitation that most people are likely to be relatively insensitive to the impact of low-prevalence but high-severity items such as homelessness, illiteracy and dependence on food banks. There is a precedent for the use of such informal information sources by statistical agencies. For example, “market research companies, trade journals and … press reports” have been used in measuring inflation (ONS 2021).

It is not possible to produce satisfactory measures for all the features that Robert Kennedy (1968) listed, such as “the health of our children, the quality of their education, or the joy of their play” and “the beauty of our poetry or the strength of our marriages”. But I suggest that the proposed economic outcomes make a vital contribution to most of these more impalpable qualities.