Political theory in social progress indices: the mutual benefits of theory and practice

ABSTRACT

Recently political theorists have voiced a desire for political theory to be more involved in and relevant to political science and practice. Turning to a concrete instance, the case of social progress indices, we observe that political theory is already to some extent involved in both. Political theories in some instances inspire well-reasoned and systematic construction of measurements, although the conception of social progress in an index is not always overt. At the same time, social progress indicators and conceptions provide a substantial context to aid the self-reflections of political theorists. This note articulates the bidirectional actual and potential processes at play, with an emphasis on the cases when the conceptions represented by measurement are value-laden and contested. It further differentiates manners by which indicators may contribute to the internal and external assessments of concepts and theories. The study highlights the significance of recognizing and enhancing these multi-level and bidirectional processes. Such improvements would be achieved when political theory, political science and practical consideration come together.

KEYWORDS:

• Political Theory
• indicators
• social progress
• concepts; measurements

Introduction

Political Theory has undergone a phase of self-reflection as concerns status, role, and methodology (Blau 2017; Dowding 2016; List and Valentini 2016). Some call for theory to come ‘down to earth’, be more empirical and behavioural, more attached to political science (Floyd 2018; Perez 2022), less abstract, more pragmatic, and more engaged with political realities (Wolff 2019). Others question the relationship between theory and practice (Erman and Möller 2018), and some of these advocate a ‘methodological contextualism’ (Carens 2004; Lægaard 2016) that tests political theory against concrete real-world contexts. These advocates contend that a political theory more deeply involved in real-world cases would not only contribute to a better understanding of the real political world (i.e. would contribute to political science) but also, thereby, be challenged to improve itself.

The present research note examines the interaction between political theories and practices in the case of Social Progress Indices. While many alternative indices have been developed to represent different conceptions of ‘social progress’ or ‘development’ (these terms are used here interchangeably), these same conceptions are often associated with key theories and principles in political theory. This linkage is due to the fact that political theorists and philosophers have long pondered the basic questions: ‘What is a developed/progressed society?’ ‘What is the meaning of development and progress?’ And so forth.

Examining this specific context provides a ready opportunity to examine how political theories and their derived principles contribute constructively, through linkage to conceptualization, to shaping and understanding real-world social progress measurement practices,¹ thus advancing the field of political science. The theory and conceptions provide criteria for what we ‘actually intend’ to represent by measurement (Cartwright and Runhardt 2014, 279). In addition, and as highlighted below, it is an example of the manner by which interactions with real-world practices may contribute, in return, to the refinement and development of principles and theories.

The structure of the research note reflects this reciprocal tendency: Section ‘Linking political theory to social progress conceptions and indicators’ demonstrates ways in which political theories have inspired and informed discussions of social progress conceptions and indicators. Subsequently, it suggests an explanation of the significance of the binding of indicators with conceptions, principles and theories. Section ‘Indicators feedback to theories’ turns to the less obvious opposite direction, looking at the process whereby indicators and scores, in turn, opens up possibilities to re-examine conceptions, principles and theories. The paper highlights several ways through which such refinements and improvements are enabled: in particular it highlights the manner by which such interaction can contribute to internal and external assessments of principles (List and Valentini 2016) and to the finetuning of conceptual ‘connections’, ‘design’ and ‘structure’ (Goertz 2020; Olsthoorn 2017). Ideas in political theory, as a result of this interaction, may advance through being exposed to new kinds of reflections and challenges. Section ‘Two clarifications and conclusion’ concludes and emphasizes the advantages of recognizing and enhancing the interactions between theories and practices. These interactions may occur naturally in some cases, while in others, they can be encouraged, guided and improved.

One way to understand the intended contribution is this: While there is much discussion in the literature of political theories and concepts (as in Dowding 2016; Gaus 2000; List and Valentini 2016; Olsthoorn 2017), and a literature in political science that discusses concepts and measurements (Gerring 2012; Goertz 2020), there is little research that integrates and examines the whole series of connections from political theory to concepts to measurements, and back. To discuss this multi-level process (often implicit and unsystematic, and bridging political theory and political science), it is helpful to focus on a particular case/context. Social progress indicators provide a suitable context to work with.

By exemplifying and analyzing the reciprocal benefit for theories and measurements in the case of social progress, the paper advices for political scientists and practitioners to pay close attention to concepts and theories, learning from them to improve indicators and indices. Simultaneously, it suggests that political theorists should use indicators for reflecting on theories and principles. Making this process more explicit and acknowledged could advance the roles of political theory and political science in the development of well-reasoned indicators. Division of labour and cooperation between political theorists, political scientists and practitioners is therefore welcomed to improve this process constructively.

Linking Political theory to social progress conceptions and Indicators

As is well known, recent decades have seen an on-going development and implementation of various measures/indices of social progress, a reaction to the longstanding hegemony of the indicator GDP in representing social progress. Social scientists have developed diverse alternative indicators and indices, such as GPI (Genuine Progress Index), HDI (Human Development Index), SPI (Social Progress Index), SWB (Subjective well-being Indicators) and so forth (Bleys 2012; Fleurbaey and Blanchet 2013; Greve 2017). Historians and political scientists have analyzed the historical circumstances in which different indices have gained momentum (Bache and Readon 2016; Offer 2000).

On a different front, political scientists have provided a general analysis of the use of indicators in political science and explained just how, in order for an index or indicator to represent not merely a particular societal attribute but a comprehensive ‘idea’, it is necessary to underpin the indicator by a ‘systemized concept’ (Adcock and Collier 2001, 531). Such analysis is useful for our case of social progress indicators. The systemized concept, in Adcock and Collier’s terms, is ‘a specific formulation of a concept used by a given scholar or group of scholars’, while ‘indicators’ are ‘the operational definitions employed in classifying cases’. A good indicator is supposed to represent by measurement the systemized concept. In addition to these two ‘levels and tasks’ Adcock and Collier depict a level of the ‘background concept’ – a meta level of ‘a broad constellation of meanings and understandings’ to which a systemized concept addresses, and a down to earth level/task of ‘scores and cases’ generated by a particular indicator.

These four levels and tasks, according to Adcock and Collier, demonstrate the process of operationalization – from a background concept (in our case, ‘social progress’) to indicators and scores. But the tasks at the front also feedback to the background levels.²

Our background concept, ‘Social progress’, has been ‘systemized’ or ‘conceptualised’ in various ways. Political theory is relevant in this respect because it has provided alternative rationales for systemizing the background concept. As pointed out by Gaus: ‘To a large extent the study of political theory is an exploration of different ways of understanding core political concepts' (Gaus 2000, 3). When political theorists pondered the question of ‘What is a progressed society?’ they contributed to systemizing the concept by posing normative arguments. To be precise, what they offered were alternative ‘conceptions’ of social progress, where a conception is a less abstract and more concrete and well specified and systemized version of a concept (Dowding 2016; List and Valentini 2016). As emphasized by Olsthoorn, such conceptions often consist of normative propositions together with a particular (mere linguistic) interpretation of a concept, as they incorporate particular principles and standards that the concept demands (Olsthoorn 2017, 160).

In the context of social progress, prominent political theories together with the conceptions they entail were explicitly adopted by social scientists in order to bolster conventional indices.

Hence, the longstanding theory of utilitarianism and its conception of social progress were taken, explicitly, as inspirations and a source of justification by a variety of contemporary indicators. For example, advocates of subjective well-being indicators of social progress, such as Life Satisfaction and Happiness measurements,³ conventionally base their presentation on utilitarian conception of social progress (Frijters et al. 2020; Kahneman, Wakker, and Sarin 1997; Layard 2005; Veenhoven 2010). In these studies, utilitarianism is broadly understood as the doctrine that considers social progress as the sum of the ‘utilities’ of individuals in a given society.

Frijters et al., for example, in their recent work ‘A happy choice: wellbeing as the goal of government' (2020, 130), consider nineteenth century utilitarians like Jeremy Bentham as their predecessors. The authors refer to their methodology, which builds on life satisfaction indicators to reflect social progress, as a modern implementation of the doctrine, where people self-reports of life satisfaction represent their well-being or ‘utility’.

Interestingly, the very same broad theory of utilitarianism inspired the development of alternative indicators (see also Dowding 2009). Thus, measures of ‘experienced utility’ like Kahneman’s moment-based happiness reports, in which people report their emotional/hedonic states on a moment-by-moment basis (Kahneman et al. 2004), are regarded, likewise, as interpretations of utilitarianism (‘back to Bentham’, as in Kahneman, Wakker, and Sarin 1997). Economic aggregate indicators, likewise, have been traditionally regarded as proxies representing ‘total utility’ (Angner 2009). So, moving from the theory and the specific concept of social progress (in this case, utilitarianism – social progress as the sum-total utilities of the individuals in a given society) to the operational definitions opens paths to alternative interpretations (through different concrete conceptions) of the very same broad theory. Analytically, the linkage between all of the above various measurements and conceptions of the utilitarian theory is based on three common features: consequentialism, welfarism, and sum-ranking (Angner 2009; Dowding 2009; Sen 1999).⁴

Another example par excellence of a theory explicitly recognized as the underpinning rationale for a set of indicators is Amartya Sen’s theory of functionings and capabilities, from which the UN Human Development Index took inspiration (Anand and Sen 1994; Anand and Sen 2000b; Haq 1995). Famously, the ultimate justification for using the three particular components of the index – national income, average life expectancy and education (each of the three represented by specified indicators) – is the presumption that they are suitable for representing ‘capabilities’, i.e. ‘the ability of people to lead the kind of life they value or have reason to value’ (Anand and Sen 2000b, 84). Broadly speaking, this conception has been presented by Sen as an alternative theory to that of utilitarianism (Sen 1999).

As in the case of a utilitarian conception, here too, the HDI is but one particular interpretation and implementation of the theory and conception. This has been emphasized by Sen himself, who insisted that the capabilities theory be left open for particular context-dependent interpretations. Robeyns (2006) provides a comprehensive survey of the different specifications and decisions necessary when moving from the theory of capabilities to the level of operational indicators.

Among these, Robeyns highlights three kinds of decisions. First, whether we should focus on ‘functionings’ or on ‘capabilities’? (Capabilities are ‘effective opportunities’ rather than actual fulfilments and accomplishment of functionings). Second, what would be the particular set of capabilities? Regarding the selection of the set of capabilities, or the components of the concept, social scientists and philosophers have come up with alternative ideas and methods (see for example Anand et al. 2009; Alkire 2002; Nussbaum 2000; Robeyns 2005). Some methods are top-down, some bottom-up, in the sense that it is for people in the society, and not necessarily for experts only to participate in the decision about the relevant set of capabilities. Nonetheless, capability theory provides the framework for either experts or the people to work with. Third, how exactly should the different components be weighted and aggregated? Theoretical consideration could help in directing all of these decisions.

When moving from theories to conceptions to indicators, there is thus a substantial indeterminacy, requiring various steps of interpretation and evaluation. Nonetheless, it seems that indicators that are ultimately backed by solid conceptions and normative theories – the construction of which involved the systematic move from established theories to the front operational level – are better indicators in some respects. In what sense are they better? What does a solid theoretical and conceptual background contribute to the quality of the resulted indicators?

It seems that we need to break this into two further questions: Why is it important for indicators to lean on concepts? And why is it important for concepts to lean on theories?

Tackling the first question, as recognized for social science and political science measurements, concepts first and foremost provide criteria and coherence, preventing arbitrariness and fuzziness in the indicators level (Goertz 2020). They go so far as to contribute to the scientific process of ‘measurement validity’. In the social sciences, ‘validation’ of indicators is stronger as far as consistency with systemized concepts is achieved (Adcock and Collier 2001). The fitness of indicators to the content of the intended concept indicates ‘content validity’.

For instance, in what sense exactly does the income component in the HDI index represent (the concept of) ‘capabilities’? This question prompted Anand and Sen to revise the income component in the HDI index, and to favour some measurement improvements – bringing it closer to representing ‘capabilities’ (Anand and Sen 2000b).⁵ This was actually an act of strengthening the index’s ‘content validity’ (its fitness for representing the concept of capabilities).⁶

Now, regarding the second question, in the case of social progress concepts, as we saw, concepts, when being systemized, are packed with evaluations and normative considerations – these require particular justifications that political theories and principles are able to provide. Thus, when concepts are bound with theories, they are systemized in a particularly meticulous (and hence, more justifiable) manner. Gaus has insightfully highlighted this point by discerning ‘theories’ from ‘ideologies’ (Gaus 2000, 36–40): while both systems of thought provide conceptions, ideologies mix appeals to reason and emotion, while political theories seek to justify specific interpretations of political concepts with reasons alone. They are meant, first and foremost to justify as opposed to persuade.

Moreover:

A political theory identifies some considerations as important, and having done that, it is led to seeing some aspects of it … as more important than others. We cannot answer Socrates’s question – ‘What is justice?’ – until we know the things that really matter to us, and the way in which we understand society. Thus, only within political theory can we justify one conception … rather than another. (Gaus 2000, 40)

Importantly, it should be noticed that political theories and principles are necessary to better systemize concepts when conceptions are value-laden and carry political consequences.

In such cases, as explained, there are many decisions to make along the way from background concept to the indicators construction (part of which involve values and normative considerations). Theory provides a well-reasoned ground to base them on. As Sharon Crasnow puts it in discussing the measurement of the concept of ‘democracy’: ‘Normative theories of democracy constrain measures of democracy … by providing a theoretical framework – a starting point and a touchstone' (Crasnow 2021, 1223). ‘The reliability (consistency) of the measure with theory provides a means to construct valid measures – measures which cohere with theory' (Crasnow 2021, 1124; similar views, regarding the concept of ‘well-being’, can be found in Alexandrova 2017, chapter 6).

It might be argued that measurements with no theory and no conceptual frameworks, which hang on mere intuitions or folk views about social progress, might work as well. The history of social indicators puts this under serious doubt. As described by Bache and Readon (2016), the current wave of development of social progress indicators, initiated in the 1990s, was preceded by an earlier wave in the 1960s, when social scientists and statisticians developed ‘social indicators’ to complement the conventional economic indicators. This ‘social indicators movement’ had faded out by the late 1970s. The decline of this first wave is explained in various ways, but one prominent explanation is the lack of conceptual coherence underpinning the indicators. Thus, data had been collected regarding social attributes such as the level of education, health and housing, but any linkage of these indicators to coherent theories and conceptions was missing (Bache and Readon 2016, 42; Bulmer 1989; Scott 2012, 20, Cobb 2000). Therefore, the use of these indicators was sporadic and did not reflect a coherent ‘idea’, and this was one of the reasons for their ineffective nature and fading.

As emphasized in this section, constructing indicators and indices based on systematic concepts and theories can enhance their coherence, validity, and well-reasoned justification, especially when dealing with value-laden concepts, which is often the case in political science. The section also illustrated that in some instances, political and social scientists actively seek such theoretical foundations. The next section explores a less straightforward direction, addressing how and why indicators and measurement of a concept like social progress can enhance the development of conceptions and theories.

Indicators feedback to theories

Political theory methodologists highlight how unpacking theories into real life examples ‘encourages theorists to confront challenges they might otherwise overlook and to think through the implications of their accounts more fully’ (Carens 2004). The ways in which practices inform principles in political theory are much discussed and debated (Erman and Möller 2018; Floyd 2018). Indicators, in general, can function as specific form of ‘practice’ (see footnote 1).

Political science methodologists, for their part, point to the tendency of indicators and scores (in general) to interact on the level of concepts. Insights about scores and indicators enable them to modify systemized concepts by ‘fine-tuning the systemized concept, or extensively revising it’ (Adcock and Collier 2001. ‘Double fitting’ as it is called by Ragin 1994, 98). The case of social progress indicators provides an opportunity to explore and exemplify these insights.

This section discusses four, non-exhaustive, ways by which indicators could contribute to the revision of conceptions and theories of social progress.

Testify to theories’ coherence and meaningfulness (‘internal criteria’)

As pointed out by Carens: ‘Theories aim to generalize and that inevitably entails abstraction from particularity, but sometimes theories are presented at such a level of generality and abstraction that it is hard to tell what they really mean'. Practical examples therefore help to understand a theory, ‘to have some sense of what it really means' (118, 119). It is natural to think of indicators as a way to ‘cash out’ theoretical ideas, to make them concrete, and hence, to enabling reflection upon them.

By reflecting upon the concepts and theories, equipped with the indicators and scores, the basic attributes of a theory that can be assured would be: their meaningfulness, and their being operationalizable (these echo what Erman and Möller 2018 calls ‘fitness’/coherence and ‘feasibility’ of principles). As emphasized in the last section, a theory can be interpreted in various ways and can result in many kinds of indicators. Nonetheless, each of the interpretations, if conducted reasonably and systematically, is re-assuring and telling: It exemplifies how the theory could be ‘grounded to the earth’ under some specified assumptions. Hence, it testifies that the theory is understandable, coherent and consistent.

For example, using Life Satisfaction indicators to indicate social progress is but one interpretation of a utilitarian theory. It is a systematic interpretation because it follows the three rules of consequentialism, welfarism, and sum-ranking – similarly to a utilitarian social welfare function (see footnote 3). It presumes a particular kind of information for indicating ‘social welfare’ (Adler 2012; Angner 2009; Haybron 2007). But the point is ultimately, not only that it leans on the utilitarian theory, but also that it reveals something about the theory: that it can be operationalized, hence that the theory is meaningful and relevant to the real world.

Political science methodologists see these attributes of concepts – coherence, consistency, and its being understandable and operationalizable – prominent criteria for assessing them (Gerring 2001; 2012). Theories that entail such concepts are, at the very least, useful theories. List and Valentini (2016) distinguish such attributes, which they call ‘internal criteria’ (for the assessment of normative theories), from ‘external criteria’. The latter refer to questions of normative force (some would say, ‘correctness’⁷) of a theory. The former nevertheless guarantees, as explained, more minimal requirements about a theory, regardless of its being normatively persuasive or accepted.

Assisting with ‘external assessment’ of theories

Indicators may help with enabling processes of external assessment of social progress theories. There are various relevant methods for strengthening such external assessments (List and Valentini 2016). One method extensively discussed by political theorists involves searching for a ‘reflective equilibrium’ between one theory and other, ‘considered judgements and beliefs’ (Daniels 1996; Rawls 1973). Thus, judgements and intuitions one holds outside a theory can serve to initiate a move from a theory to them and back to the theory. Along this move, one can decide to re-assess and revise parts of the external judgements or parts of the theories, or both, until an equilibrium is reached, a coherence between theories and external judgements. This method, well known in political theory, is meant to prevent theories from remaining isolated sets of principles in an echo chamber.

Referring to social progress indicators and the theories underpinning them, indicators enable a process of reflective equilibrium because one can hold judgements and intuitions about the fitness of the indicators in representing social progress – regardless of the theory she holds. Indicators can spark independent intuitions and judgements. For instance: one can have a judgement regarding the fitness of Life Satisfaction surveys in indicating social progress, regardless of any acquaintance with utilitarianism.⁸

Referring to the question of whose judgements should serve as an input in a reflective equilibrium, this could be left open. Recent scholarship points to the option that these judgements will not be left to experts (theoreticians and scientists) alone, but that in such processes the involvement of members of the public may be welcomed. de Shalit (2020), for instance, suggests making political theories a shared product between members of the public and the philosopher. Alexandrova and Fabian (2022) address the related question of the production of ‘thick concepts’⁹ and their measurement, and suggest that members of the public should be referred to as ‘stakeholders’ in the process, and should contribute to various phases of it through their ‘lived expertise’.

Nevertheless, be it a process conducted by a philosopher or a group of researchers, or a process that includes public reflections, the indicators and scores can give all of the above accessible references to work with when reflecting upon social progress conceptions, and when pondering between abstract theories. In the classroom, for instance, it is much easier for students to deliberate on the theory of functionings and capabilities and discuss its normative force, when provided with the example of Human Development Index, or with an alternative measure. Hence, indicators as practices provide an accessible reference to work with when assessing theories.

Refining, re-structuring and re-designing theories and concepts

Related to the process of internal and external validation, the very embodiment of ideas and concepts in the shape of indicators and scores sharpens questions about the accurate meaning of concepts, in particular of their ‘conceptual connections’ (their links to other concepts, Olsthoorn 2017, 161) and their component inter-relations. This is due to operations that are available for connecting numerical data and scores. First of all, the manner by which a combination of indicators within an index is addressed mathematically reflects dilemmas on the conceptual level.

For example, returning to the real case of the development of the Human Development Index, the way by which the three components of the index have been chosen and came into the shape of indicators, as explained in section ‘Linking political theory to social progress conceptions and indicators’, was initiated from the theoretical and conceptual levels. However, the indicators, in turn, highlighted on the conceptual level specific dilemmas that required further theoretical deliberation. This is exemplified, for instance, in a world bank policy paper (Ravallion 2010) in which questions about the particular methods of aggregation of the three components are raised. When there was no need to resolve these questions within the indicators level, the concept ‘structure’ (Goertz 2020) was left under-specified.

Additionally, this holds in the case of a combination between different (competing) conceptions. Interestingly, the embodiment of ideas in the shape of numbers and scores may invite us to ‘re-design’ concepts (Olsthoorn 2017, 170–172). Whereas within theoretical discussions it is more common to define and discuss one conception vis-à-vis another (as in the case of capabilities and subjective utility), in the indicators and indices sphere, there is no need to approve one index by rejecting another. The requirement is different: to identify and clarify the precise relation between different relevant indicators/indices regarding the representation of a shared concept, and the proper ‘division of labour’ between them. Byway, in some cases, a process of concept formation/design is enabled (Gerring 2012).

For instance, we could consider embracing a view that capabilities and total subjective utility are both relevant to indicate social progress. In order to represent this acknowledgement within an index, through the aggregation of the chosen indicators, we should yet ponder two issues. First, how significant are they to the concept? (One in comparison to the other). Second, how far can one compensate for a lack in the other? At the indicators level, these conceptual issues take the form of questions regarding (first) weights and (second) degree of substitution between two variables of one mathematical functional form. Hence, dilemmas on the indexing level – of how to aggregate two indicators – specify dilemmas on the conceptual level about the concept structure (see Goertz 2020 chapter 6).

To sum up, in some cases, the very expression of conceptions in the ‘language’ of scores may lead to eye-opening questions on the theoretical level. For a start, it sharpens the inter-relations between concepts’ components. Secondly, it highlights the option that we should not necessarily choose between competing concepts. Instead, we could accept the view that two, or more, competing conceptions are relevant to a shared background concept of social progress. Philosophers would call this kind of a standpoint ‘pluralism’ regarding goods and values. However, when indicators are involved, the very idea of pluralism of values is forced to take a much more concrete form and requires theorists to specify a manner by which conceptions relate to each other. The operations available to indicators lead to a better understanding of the particular structure of ‘pluralism’ (Cohen Kaminitz 2020).

Strengthening ‘Nomological validity’

‘The underlying idea of nomological validation is that scores which can validly be claimed to measure a systemized concept should fit well-established expectations derived from causal hypotheses that involve this concept' (Adcock and Collier 2001, 542). When theories take the forms of systemized concepts and then of indicators, they can be allegedly tested against causal hypotheses, or so one might assume. This assumption should be taken with a grain of salt, though.

For example: suppose our hypothesis is that people tend to exit societies that are less progressed (when they are able). With this hypothesis, we could design an empirical test in which we take two competing theories of social progress (say, utilitarianism vs. capabilities), with their corresponding indicators, and using an indicator for representing immigration patterns, we could test which of the two theories produces better predictions regarding immigration (controlling for other variables).

As explained by Adcock and Collier, this kind of exercise is not pointless, because it could add some weight in favour of a particular systemized concept, and therefore contribute to ‘external assessment of theories’; nonetheless it is far from being a true test of a concept or of a theory. This is because the hypothesis itself may conceal a particular conception of social progress in the first place (for example, in this case, it implicitly assumes a subjective conception of social progress – how people perceive the situation rather than an objective conception of social progress). Moreover, the hypothesis itself, as in many cases in political science, is by itself not inherently more ‘epistemologically secure’ than the tested theories of social progress. It is not obvious that people indeed tend to leave less progressed societies when they can.

Thus, this section identified four non-exhaustive ways in which real-world indicator practices, can contribute to specific reflections on concepts and theories. In sections ‘Linking political theory to social progress conceptions and indicators’ and ‘Indicators feedback to theories’ both directions, Theories->conceptions->indicators, and Indicators-> conceptions-> theories, were explained and illustrated. In the final section, we need to consider the significance of highlighting these multi-level, bidirectional processes. How does this contribute constructively?

Two clarifications and conclusion

As highlighted in this research note, and demonstrated through the case of social progress and its measurements, political theory can play a role in refining and systemizing concepts and conceptions, paving the way for the development of well-justified and meaningful indicators. For political theory, this process provides a valuable context, for testing and refining its concepts and theories. Indicators and measurement, thus, can facilitate specific forms of reflection and contribute to the development of concepts and theories.

When considering the interaction between concepts and indicators, theories and practices, two main questions arise regarding the significance of highlighting this process. First, is this a circular interaction where theories and concepts validate their corresponding indicators, which then reinforce those theories and concepts? If so, such circularity might hinder critical assessment. Second, is this process happening spontaneously? If not, should it be encouraged? Is the present research note primarily descriptive or prescriptive? Each of these options implies a different significance in pointing to the process.

Regarding the first question, some cases show that the interaction between concepts and indicators is not necessarily a self-confirming closed circle. Theories and concepts can be discussed and evolve not only through the interaction with indicators, and indicators likewise can evolve not only due to a change in concepts. Hence, the dynamic character of the process can be affected by external inputs, either on the conceptual level or on the indicators level.

For example: a growing acknowledgement of the role of sustainability has led to reassessing the theory and concept of ‘capabilities’ in order to reconcile it with the idea of ‘sustainability’, hence the new concept of ‘Sustainable Human Development’ (Anand and Sen 1994; Anand and Sen 2000a; Pelenc et al. 2013). According to these new understandings, at the theoretical and conceptual level, some changes had to be made also within the Human Development Index. These changes were initiated by a broad new understanding of the idea of social progress that pushed for a re-conceptualization, and not from the interaction between the indicators.

Regarding the second question, significantly, the fact that examples exist for constructive processes of an interaction between theories, concepts and indicators in the measurements of social progress, does not guarantee that this process is always successfully realized. The aim of this discussion is not solely to describe past processes, but to use these examples to identify instances where such interactions are less successful or entirely avoided. Learning from both positive examples and challenges is crucial, particularly at a time when indicators are gaining significant prominence in both political science and the political sphere (Davis et al. 2012).

To take one well-known example, the GDP indicator had been regarded for many years (especially in the 1950s and 1960s) as a leading indicator in representing social progress, although its conceptual foundations where never truly linked to such a background concept. It was a different concept that GDP was representing relatively well: a concept of ‘economic production’ (see Hicks 1975; Schreyer 2016). The problem was that the indicator became so popular and gained such a reputation in the social sciences, politics and the public sphere that it was carelessly associated with other conceptions, including social welfare and social progress. Thus, the two waves of ‘beyond GDP’ (Bache and Readon 2016, see section ‘Linking political theory to social progress conceptions and indicators’) can be understood as a reaction to the gap revealed between the theoretical and conceptual level and the indicator level.

Eventually, the indicator was widely recognized as inadequate for representing particular concepts like social development and social welfare, leading political and social scientists to seek and develop alternative indices for representing these concepts. This case is telling because it demonstrates not only the tendency of disruptions between concepts and indicators to be resolved at some point, but also, on a less optimistic note, the tendency for institutionalized and well-established indicators to unjustifiably override conceptual limitations. To prevent such indicator over-dominance (as well as other instances of theoretical over-dominance), the process described in this research note should be acknowledged, continuously debated and improved.

Recognizing the multi-level bidirectional interactions highlights opportunities for enhancing both indicators and theories, and in a broader sense, for advancing the roles of political theory and political science in these processes. These opportunities for improvement can be fully realized when political theorists, political scientists and practitioners acknowledge the significance of such interactions and actively work to enhance them.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by Israel Science Foundation.

Notes

1 ‘A practice’ is straightforwardly understood here as a real- world regularity of behavior, in contrast to theories and principles. This perspective aligns with ongoing debates in political theory regarding the relations between theory and practice (as discussed by Erman and Möller 2018). In this context, indicators are considered part of the category of ‘practices’ because they are tangible, real-world tools. Sociologists refer to indicators and quantifications as a form of ‘knowledge production practice’ (see Bueger 2015). This does not contradict the role of indicators in serving political scientist and practitioners for evaluating and measuring other real-world practices.

2 A parallel analysis is presented in the analysis of ‘Measurement’ by Cartwright and Runhardt 2014. They use the alternative titles: ‘characterization’ (in parallel with the level of systemized concept), ‘representation’ (in parallel with indicators), and ‘procedures’ (in parallel with scores).

3 These indicators are based on peoples’ introspective estimations in response to survey questions. In these surveys, respondents are asked to rate their satisfaction of their life as a whole on a scale (Layard 2005; Veenhoven 2010).

4 ‘Consequentialism’ means that the results generated are what counts (contrary to intentions and other reasons behind these results); ‘welfarism’ restricts the relevant results to the utilities in the respective states (the subjective attitudes); and ‘sum-ranking’ requires that the utilities of different people be simply summed together to derive their aggregate merit (Angner 2009; 152–153 quoting Sen 1999, 59).

5 In this light, Anand and Sen examine the different manners by which the national-income component can better reflect the idea that there are ‘diminishing returns’ from income to capabilities – that an increase in income represents an increase in capabilities in a non-linear way. On the same grounds, distribution corrections were added to the income component (Anand and Sen 2000b).

6 Sabina Alkire expressed a similar idea: ‘It seems for this reason yet more important to bind the practical methodologies that prove feasible together with a fuller conceptual account of what would be desirable, so that the practical methods, however circumstantially valid, remain open to challenge, modification and expansion on the grounds of their inconsistency with their own goal’ (Alkire 2002, 199).

7 The question of whether there is such a thing as a ‘correct’ normative theory is in dispute between moral ‘realists’ and ‘anti-realists’ (List and Valentini 2016, p. 16). I am in the second camp.

8 Of course, rejecting Life Satisfaction indicators does not mean that one should also reject utilitarianism all together (because, as explained, this is only one possible interpretation of the theory). On the other hand, if one tends to approve of these indicators, she should approve of utilitarian theory, as well.

9 Thick concepts ‘describe and evaluate simultaneously’, see Alexandrova and Fabian (2022). In the terminology of this paper they parallel the ‘systemized concept’, which involves both evaluation and description.