Indicators for Sustainable Prosperity?—Challenges and Potentials for Indicator Use in Political Processes

1 Introduction

Indicators are widely perceived to be useful tools for researching and guiding various forms of societal progress, including sustainability (Singh et al., 2012), human rights (Merry, 2011) and wellbeing (Self and Randall, 2013). For example, the Compendium of Sustainability Indicator Initiatives lists 895 sustainability initiatives that either use or develop sustainability indicators (IISD, 2011). Likewise, many national and international sustainability plans have made indicators a key part of their implementation (Lyytimäk, 2012). A high profile case is the United Nations (UN) post-2015 development plan, Agenda 2030. Signed off by 193 countries, Agenda 2030 is centred on 17 ‘Sustainable Development Goals’ with more than 200 indicators. The indicators are described as essential “to help with the measurement of progress”, “ensure that no one is left behind”, and “key to decision making” (UN, 2015, p. 12). In this paper we explore the extent to which such claims about the role of indicators are justified.

1.1 The Pros and Cons of Indicator Use

Indicators are central to many sustainability and other socio-cultural projects because they are a useful way to generate knowledge of, and communicate about, complex issues. Indicators break complex issues into more readily understood chunks of information thus allowing communication between experts and non-experts (Merry, 2011, Morse, 2016). Likewise, through selecting and measuring a finite set of quantified indicators that approximate the essential elements of a concept, experts can ‘measure’ an otherwise immeasurable entity (Turnhout et al., 2007). There is a long history of using indicators in this way in the biological sciences, particularly ecology (Bell and Morse, 2008) and indicators are applied similarly in sustainability research (for example, Mair et al., 2016)). In such cases, indicators can function as an analytical structure, mediating between the nuanced, complex, and difficult to interrogate concept of sustainability and the blunt analytical tools with which complex systems can be investigated. Additionally, indicators are necessary inputs for the investigation of complex concepts by other research tools such as models.

However, indicators have been widely critiqued. Indicators are reductionist analytical tools and their use risks oversimplification, particularly in highly complex and contested contexts (Morse and Bell, 2008; Merry, 2011) where their use can often hide the complexity and interrelations of the underlying system. This can be especially problematic because of how users interact with indicators. A selected set of indicators to measure a certain concept, such as sustainable prosperity, is often assumed to be objective and a complete description of the concept it measures. However, in reality the choice of particular indicators is often value-laden and incomplete (Merry, 2011, Porter, 1995). Additionally, indicators can be biased by the specific indicator construct (conceptualization of the indicator), the determination of the representative sample used to gather data or the choice of statistical methods for the data aggregation. Moreover, indicators help us to construct knowledge and guide decision making. Consequently, where they inadequately describe a contested concept, that concept may even become re-defined in terms of its indicators (Bell and Morse, 2008, Espeland and Sauder, 2007). This can lead to policies and strategies that focus on what is measurable rather than addressing less tangible or measurable issues. For example, the use of GDP as an indicator of societal progress has led to a reframing of societal progress as predominantly about increased productive capacity of the economy, creating a ‘growth imperative’ (Jackson, 2016).

In fact, GDP is a particularly pertinent example of the dangers of indicators. First it is an inadequate measure of societal progress because it misses important factors that contribute to broader conceptions of progress (Prescott-Allen, 2001, Stiglitz et al., 2009, Anderson, 2014). Furthermore, GDP growth is strongly correlated with negative environmental impacts (Antal, 2014) and the extent to which it can solve social problems is questionable (Victor, 2007), at least in the case of developed countries. Such criticisms occasionally find an echo in political discourse. Robert Kennedy (1968) famously disparaged economic growth as a measure of progress:

“Gross National Product counts air pollution and cigarette advertising, and ambulances to clear our highways of carnage… Yet the gross national product not does not allow for the health of our children, the quality of their education or the joy of their play… it can tell us everything about America except why we are proud that we are Americans.”

Likewise, Tony Blair (DETR, 1999) argued:

“there is a growing realisation that real progress cannot be measured by money alone…But in the past, governments have seemed to forget this. Success has been measured by economic growth – GDP – alone. We have failed to see how our economy, our environment and our society are all one. And that delivering the best possible quality of life for us all means more than concentrating solely on economic growth.”

However, for all such misgivings and promises to change focus, policymakers have yet to move on from the use of GDP as a proxy measure of progress. We see this even in the SDGs, which have been widely praised as “holistic” and for moving the development away from GDP alone, but nonetheless have an entire goal (Goal 8) which not only aims to promote sustained economic growth (measured as GDP), but also conflates GDP growth with concerns about decent jobs. GDP growth continues to be the principal objective of most government policy (Victor, 2007), and there is little sign that this is changing.

1.2 Indicators for a Better Future?

As researchers and citizens, the authors of this paper are interested in how indicators may or may not be used to help imagine, explore and create a better future. Specifically, all the authors of this paper are engaged in work and research into ‘sustainable prosperity’. Sustainability and prosperity are both ideas that have been widely used in plans for societal progress (the UN Global Goals, for example, use both). We understand sustainable prosperity as a good future, a world in which people everywhere have the capability to flourish as human beings whilst remaining within the ecological and resource constraints of a finite planet (Jackson et al., 2016). However, sustainable prosperity remains a highly contested concept and, given that indicators can be problematic in the absence of a fully agreed definition, the use of indicators to flesh it out should be approached with care.

Therefore, in this paper we critically engage with indicators, particularly where they have been used in the context of contested and complex phenomena. Based on a review of the literature, we critique the use of indicators as they have been used for various socio-cultural projects, with a view to understanding how they may be used in our work on sustainable prosperity.

The rest of the paper is structured as follows. In the next section we describe sustainable prosperity as a contested concept. In section 3 we highlight that indicators of contested concepts are not neutral, but instead represent a particular perspective on an issue. We then demonstrate the difficulties this raises, illustrated in relation to the United Kingdom (UK), European Union (EU) and United Nations (UN) sustainable development indicator sets (see Appendix). In section 4 we elaborate on this, highlighting how indicators of contested concepts risk oversimplification as they struggle to deal with the complexity of moral and ethical problems. Finally, in section 5 we mount a qualified defence of indicators, arguing that although these two critiques should influence how we use indicators in contested systems, they do not altogether negate their usefulness.

2 Sustainable Prosperity

‘Sustainable Prosperity’ is a highly contested concept. Both sustainability and prosperity relate to issues that are very subjective and politically sensitive. Therefore, sustainable prosperity may be more of an affective-cognitive construct (Davern et al., 2007) than a tightly defined analytical term. In other words, sustainable prosperity as a concept, and how it might be measured, is dependent on each person’s world view. For example, the precise understanding of sustainable prosperity depends on how political and institutional contexts frame each person’s perceptions.

Even the elements of sustainable prosperity that appear to relate to objective physical realities intersect with highly value-laden problems. For example, a common theme in discussing sustainable prosperity is the presence of physical planetary limits and the impetus they provide for us to reduce environmental damage. While planetary limits are grounded in physical science (Rockström et al., 2009, Steffen et al., 2015) they are constructed in such a way as to keep the human economy in a relatively benign environment. For example, a biodiversity ‘limit’ assumes that we value today’s biodiversity more than historic or future biodiversity and does not account for the idea that a new biodiversity could flourish under the conditions of a changed environment. Furthermore, the impact of planetary limits on our ability to live good lives is highly contested and subjective. We take the view that any understanding of sustainable prosperity must be cognizant of these physical limits, thus positioning sustainable prosperity as a ‘strong’ view of sustainability where natural capital and man-made capital are complements rather than substitutes (Daly et al., 1995). But of course this too is contentious. As the name ‘strong’ sustainability suggests, there is an alternative reading of the impact of planetary limits (‘weak’ sustainability) which views them as being of negligible importance in the construction of sustainable policies.

Other elements of sustainable prosperity (indicators) are also contested. For example, central to our understanding of sustainable prosperity is that it is a multi-dimensional concept about more than just the economic health of society. In our view, a prosperous society is one that is concerned not only with income and financial wealth, but also with the health and wellbeing of its citizens, with access to good quality education, and with their prospects for decent and rewarding work. Sustainable prosperity enables basic individual rights and freedoms but also goes beyond this and allows people to flourish. But what is flourishing? What are the capabilities that enable flourishing?

Likewise, for us, inter- and intra-generational justice and equity is a core element of sustainable prosperity, such issues can be measured by indicators such as the Gini coefficient or the more general ‘poverty reduction’. Another element that could be important is civil engagement and democratic inclusiveness, and the autonomy to act, particularly at the community level (Böhmelt et al., 2015, Howard and Wheeler, 2015). In achieving sustainable prosperity, issues concerning finance and the economy are also likely to play an important role, as are issues around diversity of ownership, and investment models (Vickers and Lyon, 2014, Jackson and Victor, 2016).

The indicators we choose are influenced by the vision of prosperity that we have, and choosing the wrong indicator will lead to the ‘wrong’ prosperity. We explore this difficulty in the next section.

3 Indicators are limited interpretations, not objective descriptions.

Contested concepts defy the naive understanding of indicators as readily digestible representations of the essential components of a larger system. In this theory researchers construct new information or communicate about the system as a whole (Bell and Morse, 2008, Figure 1) by combining and interpreting multiple indicators of the system. Although this understanding is applicable in perfectly objective and well understood systems, contested concepts, conversely, are characterised by multiple and conflicting ideas about how any given system works. As a result, any given indicator set is only able to represent a subset of these understandings and the differences in these understandings result in different indicators (Meadows, 1998, Davis et al., 2015). Therefore, an indicator of a contested system should not be understood as a piece of information about a system, but a piece of information reflecting how an individual or group conceptualises that system (Figure 2).

3.1 Three different understandings of Sustainable Development

To illustrate how the ability to represent only a limited perspective creates difficulties for those who would use indicators in contested systems, we compare three indicator frameworks that aim to measure and implement sustainable development. The UK (Lofts and Macrory, 2015), EU (Eurostat, 2015) and United Nations (through the Sustainable Development Goals – SDGs) have each developed a set of sustainability indicators. Comparing these three indicator sets is instructive because they have substantial differences, despite all being focused on the same concept and being primarily designed for the same users (nation states).

It is not just that the processes leading to these sets of indicators could not agree on common statistical metrics – they fundamentally could not agree on how to frame the indicator sets. At the most basic level, each set contains a different number of indicators or targets (see Appendix):

• SDGs – 17 goals, 169 targets, 230 Indicators
• EU sustainable development indicators (SDIs) – 10 thematic areas (with headline indicator), 132 indicators
• UK SDIs – 3 thematic areas, 66 indicators

Looking deeper, there are only three indicators that use the same statistical measure across the three frameworks: 1) GDP per capita, 2) Greenhouse gas emissions, 3) Share of renewable energy. Likewise, there are only six further indicators that have a common outcome even if the statistical measure is not exactly the same: 1) Increase research and development, 2) Reduce air pollution (or the impact of air pollution), 3) Increase water efficiency, 4) Increase river quality, 5) Regulation of fishing, 6) Protection of forests. Some of these differences are understandable: it is natural that a national framework to measure progress towards sustainability may have different targets from an international framework (there is little reason for the UK to have a national target relating to rainforests, for example). However, many of the differences are less intuitive.

Several key aspects of the frameworks appear common in nature but offer very different indicator sets. For example, ‘poverty’ appears in all three indicator sets but with a number of different statistical measures. The UN’s Global Goals’ (SDGs) poverty indicators focus on the proportion of a national population suffering from various dimensions of poverty. Income and monetary aspects of poverty are captured through measures of people living below national and international poverty lines, while more social dimensions of poverty are captured by measuring the proportion of the population unable to access social protection systems and lacking access to basic services. In contrast, the UK’s poverty SDIs focus predominantly on the proportion of children living in low income households, where income is defined both in absolute and relative terms.

Indicators both reflect and help construct theoretical perspectives and problem conceptualisations (Merry, 2011): the differences in the UK SDI and the SDG poverty indicators represent alternative understandings of poverty. The UK SDIs emerge from a conceptualisation of poverty as a primarily monetary problem, albeit with a role for societal norms around income. Further, the UK SDI theory of poverty sees households with children as the most at risk group (either because they are most likely to be affected, or likely to suffer the most). Conversely, the SDG poverty indicators emerge from a conceptualisation of poverty as problem that is broader than income alone (hence the inclusion of indicators on societal safety nets); they are based on a view of poverty as a problem for people of all ages (with indicators focused on a range of demographics); and they do not view societal norms as particularly important. Instead the SDGs focus on the absolute understanding of poverty and have no explicit indicators for relative poverty.

Beyond poverty, there are other examples of differences in the three indicator sets that reflect differences in theoretical understandings. For example, public health, where the SDGs and the EU SDIs focus on specific measurable health outcomes (such as maternal mortality, mental health, suicide rates or access to health care), the UK SDIs take a broader perspective that also encompasses more indirect influences on mental health, such as civic participation, whether people perceive that their neighbours can be trusted and if they have close relationships. In the main, these concepts are missing from the SDGs and EU SDIs. On the other hand, absent at the UK level are measures of trust in governance. The promotion of the rule of law, reduction in corruption, access to information, voter turnout and confidence in government are measured at both SDG and EU SDI levels (in different ways).

3.2 Indicators shape how we view the world

The differences in the three indicator sets may also drive outcomes that undermine each other. Indicators refocus attention on to the elements of a concept that they measure and away from the elements that they do not. In this way, indicators direct how their users think about and attempt to deal with the concept itself (Espeland and Sauder, 2007, Merry, 2011). For example, the UK SDIs include a measure of the origin of food consumed in the UK, while the EU SDIs include a measure of imports from developing countries by group of products. Therefore, at the UK level there is a target to reduce food imports while at the EU level there is a target to increase them (albeit from a specific set of countries). Likewise it seems reasonable to suggest that that the UK SDI poverty indicators will result in different policies and outcomes than the SDG poverty indicators. For example, if the UK reduced the coverage of its social protection systems but simultaneously increased the average income of households with children, poverty could get both worse (according to the SDG definition) and better (according to the UK SDI definition)!

4 Indicators struggle with unquantifiable, moral issues.

Indicators are further challenged by the difficulty of codifying, quantifying and linking important elements of contested concepts. The chief utility of indicators comes from their simplification of complex issues, making the ideas more manageable. By definition, this requires losing information. Often, this information is the contextual and qualitative, discarded because it is hard to quantify or otherwise codify, rather than because it is unimportant (Porter, 1995). Moreover, indicator sets have typically been developed without explicit consideration of their ethical basis and the moral assumptions embedded in the choice and content of particular indexes. Fredericks (2014, p.6) argues that ‘there is a widespread assumption in modern Western society that technical and ethical assessments are, and should be, completely separable’. This needs to be rejected, in her view, on the grounds that ‘developing indexes without explicit attention to ethics runs the risk of ineffective indexes, or even worse, indexes which drive people away from their vision of sustainability’ (Fredericks, 2014, p. 9). This crucial point about the ethical dimension of indicators can be made also in relation to values concerning aesthetic judgements, cultural goals and ultimate human ends.

4.1 Arts, Culture, and Ultimate Ends

Informative examples of the ways in which indicators struggle with messy and difficult to codify ideas and relationships are found in arts and culture indicator initiatives such as ‘Taking Part’. This is not to say that arts and culture offer a unique challenge when it comes to the use of indicators, as economic indicators also have significant problems associated with their measurement and use however the Taking Part initiative offers a useful case to explore. Taking Part is perhaps the largest and most prominent attempt to provide data on the cultural sector in England. Research using Taking Part data claims to provide “robust evidence” (DCMS, 2014, p. 4) of links between certain kinds of cultural participation and subjective wellbeing (wellbeing as measured through how individuals describe their own feelings). However, establishing causal links between participation/engagement and other outcomes – such as increased wellbeing – proves difficult. Though quantitative analyses from the UK, as well as Canada, Italy and elsewhere, demonstrate a link between engagement in art and culture, and wellbeing, for many, “the challenges of disentangling confounding variables and establishing directions of causality remain” (Crossick and Kaszynska, 2016, p. 38).

Daly’s Pyramid (Meadows, 1998; Figure 3), is a useful framework for understanding why connecting the Taking Part participation indicators to ‘wellbeing’ is difficult.

Daly’s Pyramid frames indicators as falling into one of four categories: at the top of the pyramid are Ultimate Ends – happiness, wellbeing, flourishing. These are the things that we strive for, the high level concepts that together (arguably) constitute prosperity. At the base of the pyramid are the Ultimate Means – the fundamental earth systems without which we could not survive, let alone prosper. In between the two are intermediate means (human labour, tools, processed raw materials etc.) which are used to produce intermediate ends (consumer goods, knowledge, services etc.). Intermediate ends are tools that are necessary to achieve our ultimate ends.

Taking Part measures intermediate ends, gathering information about participation and engagement in the arts, museums and galleries, archives, libraries, heritage, and sport. The survey includes data on frequency of participation. This is important as it draws a distinction between participation, a binary state whereby an individual participates or not, and engagement, which it suggests is related to the frequency of participation. Further questions aim to uncover drivers and barriers to participation. For example, respondents are asked about service provision in their area (e.g. new or closing facilities), and whether they have experienced a range of life events (such as moving house or illness) that could affect participation. It also captures socio-demographic data, including each respondent’s education level, income, occupation, marital status and health. These data are released at 6 monthly intervals and provides headline figures on participation and engagement, broken down by, for example, age, ethnicity or region (DCMS, 2015).

However, ‘wellbeing’ is an ultimate end, not an intermediate end, and it is in connecting the two that problems arise. The data collected in Taking Part allows researchers to produce models that control for other factors (such as income) and provide a statistical evidence-base for claims about the positive impacts of cultural activity that, importantly, can speak to government objectives premised on public utility and notions of wellbeing (Walmsley, 2012). But, as Daly’s Pyramid makes explicit, ultimate ends indicators are the outputs of ethical and theological frameworks. In other words, they emerge from a process of highly personal interpretation informed not only by quantifiable measures but by emotional and moral reasoning. As Walmsley (2012, p. 329) argues, many of the ways that culture and art influence wellbeing are personal and intrinsic, taking us “into the incommensurable realms of spirituality and emotion”.?

Complexity and immeasurability are not specific to wellbeing and the arts but also to other ‘ultimate ends’: happiness, harmony, community. How can psychological and personal growth, helping others or creating something new be measured? There may be some proxies for wider social benefits, such as health or education outcomes, but there is also a risk that inputs which are easier to measure, such as spending on health or education systems, do not capture the desired outcomes and, if captured as proxy indicators, become the desired outcomes in themselves. Rather than simply linking subjective wellbeing or prosperity to cultural or community activity, an approach to understanding how human and social capital (the ‘intermediate ends’) promotes wellbeing (the ‘ultimate ends’) through the role of individual capabilities, using indicators such as the ability to exercise creativity, and imagination is needed.

As noted in the introduction of this paper, perhaps the clearest example of a proxy outcome becoming a desired outcome is GDP. GDP is properly understood as an intermediate means on Daly’s pyramid: GDP measures the productive capacity of the economy, which we can use to produce intermediate ends and then ultimate ends. Viewed in this way it is unsurprising that increases in GDP often have a weak (or in some cases negative) relationship with ultimate ends such as happiness and health (REFS). GDP is a tool we can use to achieve our ultimate ends, but it will not always be an appropriate tool. But because GDP is much easier to measure than the quality of jobs (SDG 8) or how well the economy meets our needs (EU/UK SDIs), it has become the proverbial hammer and we view all our problems as nails.

Finally, it is worth noting here that the lack of clarity and difficulty of quantification is not confined to the moral aspects of ultimate ends. It takes, as Porter (1995, p. 41) puts it, an enormous amount of effort “to arrange an unruly humanity into uncomfortable categories”. Consequently, arbitrary exclusion and subjective categorisation are apparent even in the more mundane aspects of intermediate and ultimate ends indicators. Efforts to produce indicators for cultural work, for instance, are hampered by unclear boundaries and distinctions that make even counting the number of cultural workers difficult. Taking a sectoral approach to these labour markets includes large numbers of individuals in non-‘creative’ roles while excluding cultural/creative workers in non-‘creative’ industries, while approaches that seek to utilise ‘creative intensity’ ultimately include consultancy and management roles that have little cultural output (Bakhshi, Freeman & Higgs, 2013; O’Brien & Oakley, 2015: 12-13).

4.2 Indicators that ignore essential elements risk undermining the concept they purport to measure.

These issues are particularly problematic because of the power of indicators to shape the thoughts and actions of researchers and decision makers: through indicator use, the more complex and qualitative aspects of ultimate ends risk being lost or ignored. As discussed in 3.2, indicators can come to redefine concepts by directing attention only to those dimensions captured by the indicator (Merry, 2011). The act of measuring is not passive; rather, it shapes and defines what it is we are measuring, highlighting aspects to be important and, by omission, defining those aspects that are not important. As a result, indicators that ignore important elements of a concept may lead to policies that either overlook or actively conflict with the original concept as it is more broadly understood. In the arts, for example, Oakley et al. (2013, p. 24) point out that ‘well-being-friendly’ cultural policy may exacerbate current wellbeing inequalities, while also stemming the production of new work that can be viewed as “difficult, upsetting, challenging, or simply solitary”.

Furthermore, there is a risk of concepts being redefined at all levels of society. Quantitative measures give the appearance of objectivity and neutrality: numbers often hide the complexity and value-laden nature of the judgements used in their construction (Porter, 1995). Where these constructions are very complex it is difficult for non-experts to challenge the indicator (Merry, 2011). Even where experts are willing to challenge the indicator, if the only dissenting voices have little political power it is easy for the indicator to remain neutral in appearance, and the concept to be re-defined (Espeland and Sauder, 2007).

Arts and culture once again provides a useful example of such risks. Neither art nor culture are included in the UK Office for National Statistics assessment of wellbeing, perhaps because they are too difficult to measure (Walmsley, 2012). Likewise there are no arts or culture based indicators under SDG goal 3 “Ensure healthy lives and promote well-being for all at all ages”. Instead this goal opts for indicators that are based on much more easily quantified aspects of health such as maternal death rates and rates of new HIV infections. By ignoring arts and culture, wellbeing indicators may reduce actions that promote cultural and artistic dimensions of wellbeing; potentially leading to reduced wellbeing.

5 A Qualified Defence of Indicators

Despite the problems of indicators described in the above sections, we still believe that indicators have a useful, if limited, role to play in implementing sustainable prosperity. This section makes that case. The previous two sections critiqued the use of indicators in contested concepts on the grounds that they are only able to represent a small subset of understandings of that concept, and that they struggle to deal with messiness and complexity, often excluding important elements on these grounds. Here we draw on these critiques and begin to outline how we see indicators being usefully applied going forwards.

5.1 Indicators can clarify political views and increase accountability.

While indicators remove contextual information and obscure the process through which this happens, they also force a clarity and rigour that exposes political priorities and beliefs. This is seen clearly in Agenda 2030. The 17 UN Global Goals for Sustainable Development (SGDs) often describe quite broad concepts that are accepted by a majority of global society and are apparently compatible with national sustainability initiatives. However, the SDG indicators reveal very specific perspectives on these problems some of which directly conflict with national perspectives. We have already discussed how indicators reveal very different conceptualisations of ‘poverty’ in the UK SDIs and the SDG indicators, but this is not the only example of this in Agenda 2030.

Goal 8 aims to “promote sustained, inclusive, and sustainable economic growth, full and productive employment and decent work for all”, a statement broad enough that we can say with some confidence most people could agree with it. However, the SDG indicators have been criticised for failing to fully encapsulate the concept of decent work (Frey and MacNaughton, 2016). The SDGs do not, for example, include any measure of trade union coverage, working poverty rates or working time, all of which are ‘main indicators’ for decent work according to the International Labour Organisation (ILO, 2016). Moreover, there are no indicators on job satisfaction or fulfilment, reflecting a very different idea of ‘decent work’ than those for who such ideas are central (see Burchell et al., 2014 for examples). Indeed by conflating decent work with economic growth, and with only 3 of 16 indicators in Goal 8 (average earnings, work place fatalities and labour rights) attempting to measure it (even at a very superficial level), the conceptualisation of decent work at the policy level is shown to be contested at the very least.

This process of making a particular view explicit is, of course, the very same source of difficulty that we discussed in Section 3, here reframed as a strength. That indicators represent only a single perspective is a problem where they are interpreted as neutral fact, but a strength if indicators are understood as a clarification of the worldview. We must be clear here that indicators are no panacea, they do obscure those political judgements made in the construction of the indicator, but they also allow outsiders to see how concepts are being operationalised. So, instead of indicators necessarily re-conceptualising a problem and enforcing a single narrow view, they can also create a platform for debate and critique of a concept.

In part this is related to the public nature of indicators (Porter, 1995). Examining the decision making processes of the European Union (EU) and the Millennium Challenge Corporation (MCC), Dutta (2015) finds that the use of indicators makes those parts of the MCC decision making process that use indicators relatively transparent because,

“external observers can more easily identify the mechanisms by which decisions are supposed to be made. Such legibility makes a contribution to accountability; where observers can easily discern how a decision was supposed to be made, they can more easily identify deviations in how the decision was actually made.” (Dutta, 2015, p. 162)

Similarly, Finnerty (2005), argues that the use of indicators in the Irish National Anti-Poverty Strategy (NAPS) allows the government to be held to account and has formed the basis for much of the critique of the program.

5.2. Indicators facilitate new understandings of complex systems.

While indicators may lead us to re-conceptualise issues in ways that somehow lessen or reduce our understanding of an issue, they can also facilitate a helpful re-conceptualisation of knowledge. For example, Porter (1995 p. 37) makes the case that the widespread use of quantified indicators helped to create the idea of society by reframing individual problems as societal:

“Indeed the concept of society was itself a part statistical construct. The regularities of crime and suicide announced in early investigations of ‘moral statistics’ could evidently not be attributed to the individual. So they became properties instead of ‘society’…Similarly, people sometimes found themselves or people they met to be out of work before this had become a statistical phenomenon. The invention of crime rates in the 1830s and unemployment rates around 1900 hinted at a different sort of phenomenon, a condition of society involving collective responsibility rather than an unfortunate or reprehensible condition of individual persons”

By reframing a concept in this way, indicators can help us to consider new options and ways of thinking. As a result, indicators are widely used as tools to highlight problems which then guide a more detailed and contextually-sensitive analysis. For instance, subjective wellbeing indicators show large spread but relatively stable mean values. They also vary widely across geographical areas. Though the measures themselves do not explain underlying causes, they do highlight a potential problem to be explored further (Seaford, 2013).

The actual process of selecting indicators can also help understanding. This paper recognizes that there may be different views of definitions and configurations of indicators. By engaging in participatory processes, different views can be considered (Fraser et al., 2006, Bell and Morse, 2008) and space can be created for the voices of those who might otherwise be excluded. By understanding the contested nature of indicators, those indicators selected can be refined and the limitations of any research identified. Participatory processes for discussion, co-design and co-implementation of indicators became widely used in the wake of Local Agenda 21 in the 1990s, which gave considerable impetus to community-level initiatives for measuring and practising sustainable development (Warburton, 1998; Buckingham and Theobald, 2003). The basic claim made for such processes is that they can, by engaging local insights, expertise and everyday experience, make indicators both more accountable and accepted, and more reflective of local complexity and qualitative as well as quantitative change (Lawrence, 1998; MacGillivray, 1998; Walker et al, 2000; Chambers, 2008). Despite considerable problems of comparability, scaling and integration into national and international indicator sets (Chambers, 2008), participatory systems for indicator development can generate important insights and build up trust and cooperative capacity (MacGillivray, 1998; Walker et al, 2000).

Indicators also enable quantified forms of analysis that can enhance our understanding of highly complex systems. A key example, in our view, is the use of models to understand dynamic non-linear systems. Such systems are difficult to conceptualise and interrogate without models because multiple inter-linkages and feedback mechanisms can result in counter-intuitive and emergent behaviours (Sterman, 2000). Models can be viewed as tools that mediate between theory and reality. In contrast to indicators, system dynamics models show two-way interlinkages between components, including two-way links between intermediate and end goals, and give information about the underlying structure (causal links between components) of a system. In this view, models draw from both theories and the ‘real’ world but retain a level of autonomy (Morgan and Mary, 1999). Therefore, they facilitate learning by allowing users to test and refine the mental models (theories, value judgements and assumptions) that they inevitably bring to research (Sterman, 2000, Epstein, 2008, Meadows, 2008). By manipulating models we are able to see how the outputs of a model diverge (or converge), from the theoretical predictions or ‘real world’ observations and explore why this is the case. Moreover, where models are sufficiently representative of some aspect of the real world they can be considered ‘surrogate’ worlds and we can make qualified inferences from our model world to the real world (Sugden, 2000, Mäki, 2009). Indicators are essential in this process because they provide the mechanism that allows the model world, theory and reality to be compared.

Reframing a concept may also be useful for more strategic reasons, particularly where we believe current concepts are inadequate. Indicators may be pursued by communities as a way to try and embed their conceptualisations within decision making processes (Hezri and Dovers, 2006). As discussed in the introduction, GDP is widely considered an inadequate measure of societal progress. Though GDP itself plays a role in creating an idea of societal progress that is inadequate, it is also true that GDP emerged from, and is reinforced by, an inadequate concept of societal progress that centres on material goods (see, for example, Blair, 1999 and Anderson, 2014). It has been suggested that rival indicators present a useful way to challenge GDP and reframe debates on societal progress in a broader way (Cassiers and Thiry, 2014). Indeed, the SDGs have been called transformative because they represent a much broader and more holistic view of societal progress (Hajer et al., 2015, WWF, 2016).

5.3 Looking forwards: Indicators for Sustainable Prosperity.

The preceding discussion leads us to take a view on the use of indicators in our sustainable prosperity work (and contested concepts more generally). To promote the use of indicators as tools of clarification and to correct the impression that they are objective, the choice of indicators and their conceptualization should be developed in participatory way, and explicitly linked to a narrative description of a specific sustainable prosperity vision. The literature on participatory indicator development emphasises deliberative construction of visions in conjunction with a variety of stakeholders and then identifying indicators that mean something in the context of that vision (e.g. Bell and Morse, 2008). Indeed, the most effective implementation of sustainability frameworks, such as Local Agenda 21, involved the co-creation of indicators by community groups and other local actors to ensure measures had resonance (see, for example, Barrutia and Echebarria, 2012).

For our own work we propose going slightly beyond this and co-creating multiple alternative, and possibly conflicting, visions of sustainable prosperity (within the understanding outlined in this paper) and then identifying multiple conflicting indicator sets that are meaningful and useful to the specific visions. By developing intentionally conflicting indicator sets explicitly tied to specific visions of sustainable prosperity we hope to emphasise the political nature of our indicators. Drawing on the notion that indicators increase levels of accountability and clarity, we hope that this process will also create a space in which we can critically engage with both the indicators and the visions they represent. The alternative indicator sets and visions will be explored using a variety of analytical techniques. This will include, but should not be limited to, different economic models.

6 Conclusion

The use of quantified indicators and targets, such as the Sustainable Development Goals, the EU Sustainable Development Indicators or the UK Sustainable Development Indicators, are now a well established policy tool. However, as shown in this paper there is little commonality in the exact statistical measure that these frameworks use. Indeed while the outcomes of these frameworks appear to have some commonality, at least at the broadest level, the measures used are more often against inputs to the system. While the geographic scale of these frameworks is important, and we acknowledge the place specific nature of measurement (indeed we have highlighted the vital aspect of very local measurement of cultural interventions for example), the variance in the measures of inputs in these three frameworks still surprised us.

The common measures across the three frameworks relate to action on climate change and economic growth. The GDP per capita indicator, while straightforward, dominates the political process and without a structured approach to dialogue around policy development it often drives outcomes which are counter to the other indicators included in the frameworks. As the SDGs develop further, and specific indicator targets are agreed, this wider issue of process and the importance of acknowledging the difference between input (means) and outcome (ends) measures should not be lost.

To develop and use indicators which might measure sustainable prosperity poses numerous challenges. For CUSP the process around the political use and misuse (or disregard) of these indicators is as important as the indicators themselves and this is often not explored in their reporting. This should form an important part of the exploration of prosperity in practice. The limits of indicators, both in regard to what they can measure as well as their historic use to measure things that some perceive as either worthless or indeed driving incorrect behaviours, should also be acknowledged within CUSP’s work.

Given the place of arts and culture within CUSP we need to acknowledge the implications of dialogue with this sector in the context of indicators and it will provide a useful case to explore the use of measures and indicators. For example, indicators of cultural consumption are important in understanding who is (and isn’t) participating. However, these types of indicators need to be sensitive to amateur, vernacular and everyday kinds of activity outside funded, legitimated organisations as well as account for the aversion, or outright hostility, on the part of practitioners and arts organisations towards attempts to measure, evaluate and quantify their activity, and the perceived instrumentalisation of art and culture. Thinking about how to capture social capital in ways that provides a basis for comparison may also be important in understanding how it affects health, wellbeing and other markers of sustainable prosperity. If art and culture, alternative investment models or increasing community development through social enterprise, can promote a kind of sustainable prosperity then we must understand, as a matter of social justice and perhaps environmental necessity, how to make them accessible to all. Robust indicators can play an important part in developing these narratives.

Developing a multi-dimensional view of sustainable prosperity based on trans-disciplinary research and dialogue across sectors requires us to break down many barriers. There is a need for us to debate the meaning of ‘modelling’ and scenario testing alongside measurement and indicators. CUSP, as a team, will reflect on how we work in an interdisciplinary and transdisciplinary way, to explore these issues.