Katharina Lima de Miranda and Dennis Snower (hereafter KLMDS) make a landmark contribution with their proposal for new indicators that capture dimensions of empowerment and solidarity that are usually missed in the conventional indicators. In these brief comments, I would like to critically reflect on their approach while making the following points:

  • Agency and solidarity, as they construe these notions, are indeed very important to get to the core of what really matters to human development and the core of the factors that currently feed the social and political crisis unfolding in many countries.

  • As KLMDS argue, each of these dimensions, in addition to economic development and the state of the environment, has to be monitored separately in a dashboard and should not simply be amalgamated into a composite indicator.

  • It is worth relating and comparing the KLMDS proposal to the three main integrated and theory-grounded approaches that have been recently competing against the prominence of GDP as an indicator of performance—namely, the capabilities approach, the happiness approach, and the equivalent income approach.

The Importance of Agency and Solidarity

Agency, as KLMDS present it, is about “empowerment” and about “mastery of the environment, personal growth, and attaining personal goals”; and it “involves people’s need to influence their fate through their own efforts.” After a long domination, in analyses of social performance, of economic development and a consumerist approach to life success, it is indeed being rediscovered that human beings cannot be fully satisfied with the distractions of material comfort. Power and control are at least as important as income and wealth, and economic resources are actually often used as channels not just to material comfort but to various forms of power and control. The power dimension takes different forms, sometimes with cultural variations. Everywhere, control over one’s own life is associated with a strong sense of dignity.

The current crisis of populism has a lot to do with a feeling of disenfranchisement fostered by economic instability but also by the concentration of economic and political decision centers far away from local populations. Giving the “power back to the people” is a very successful slogan in this context, even if it is disingenuous in the discourse of demagogues. One should be careful to develop the KLMDS approach regarding the agency dimension in a way that fully captures the pure power dimension and does not only focus on the more economic aspect of agency associated with productivity and security through health, education, and access to employment. Control over one’s life includes the possibility to participate and have one’s voice heard in policy-making. The fact that policy-making appears determined much more by the wishes of privileged interests, via lobbying, political funding, or mere connivance among political and economic elites, than by the will of the majority and the urgent needs of the more disadvantaged populations, is a crucial factor in the current boiling resentment against the establishment. Agency and empowerment encompass the effectiveness, or lack thereof, of democratic institutions, in the political system as well as in economic institutions. Unionization, for instance, can be an important component of agency, along with similar forms of civil society organization.

Solidarity is presented by KLMDS as covering “the needs of humans as social creatures, living in societies that generate a sense of social belonging.” Belonging is indeed a most important aspect of human flourishing, and they accurately note that it can develop into a positive, open form of solidarity but also into a negative, “us vs. them” form that builds inward cohesion on the basis of excluding others. There is a dark side of solidarity, which the populist demagogues are experts in exploiting, trumpeting that there is a tension between helping our fellow citizens in need and the rest of humanity. It is therefore very interesting to try to develop a measure of belonging that tracks both the strength of the in-group cohesion and the openness to larger connections. One could even view attitudes toward nature and other species as the ultimate expansion of the circle of inclusion, so that sustainability has to do not just with caring for our descendants but also with open attitudes to the “other.”

In summary, it is hard to overstate the importance of agency and solidarity for human flourishing in general, and for addressing the current social and political crisis. The “left behind” feel both a sense of disenfranchisement and a sense of exclusion and indifference, if not outright contempt, when they look up to the successful social groups who navigate globalization, intercultural mixing, gender fluidity, and the secularization and liberalization of values with great ease. Their expressions of backlash may take ugly forms, but their resentment is justified because their dignity is undermined by the collapse of their agency and solidarity bases. What seems to me important is to give to agency and solidarity their full meaning and scope. Empowerment is not just about having a job and controlling one’s immediate conditions; it also involves enjoying a healthy set of democratic and participatory institutions at all levels and in all spheres. Likewise, solidarity is not just about having someone to trust in one’s circle of relatives but also about fostering forms of solidarity that are maximally inclusive. The failures of democratic institutions and of transnational solidarity systems such as the European Union are directly relevant to individual measures of agency and solidarity.

The Dashboard Approach

KLMDS argue against building a composite indicator that would allow trading off one dimension against another with excessive substitutability. The four dimensions they focus on are economic development (GDP), sustainability (environmental protection), agency, and solidarity. Economic inequalities are also mentioned and could be added here as a fifth dimension. Their plea for a dashboard approach seems very reasonable. It is unacceptable, and even irresponsible, to focus only on one dimension such as GDP and hope that the other dimensions will follow naturally as the economy develops. But it is also insufficient to aggregate all the dimensions into a composite indicator and allow for some of the dimensions to be neglected when the others are doing well. Careful policy-making needs to check that each of the dimensions is on a good track, because a downturn in one dimension reveals that some serious issues need to be addressed in some parts of the population or some sectors of the jurisdiction, and if these issues are not addressed, the problem is likely to spill over into other dimensions and ultimately undermine the general health of the society.

While I fully agree with KLMDS on this prudent stance, they seem to be overly negative about certain aspects of composite indicators. The first, obvious thing to note is that a composite indicator is never proposed to hide the possibly unequal evolution of its components. Typically, composite indicators are accompanied by the dashboard information that explains the evolution of the aggregate. For instance, the Human Development Index (HDI) tables show not only the value of the indicator and the ranking of the countries but also how these are generated by diverse levels of performance in economic, educational, and longevity achievements in the countries. The degree of substitutability between the components is, of course, a matter for careful examination, and again, the HDI provides a good example, with a shift from the arithmetic mean to the geometric mean of the three main components, in recent years, which was motivated by a concern for balanced development and the fact that a low value in one component has much greater power to dampen the aggregate in the geometric mean than in the arithmetic mean.

KLMDS associate their plea for the dashboard approach with a rejection of the “neoclassical” approach that relies on people’s preferences to guide the aggregation over various dimensions. In the theoretical example they provide, people’s preferences are assumed to allow for perfect substitution across the three dimensions of economic success, agency, and solidarity, whereas true well-being is assumed to disallow any trade-off and involves a maximin approach to the aggregation of the three components (i.e., well-being is equal to the lowest value among the three components). Here a paradox is starting to appear. If we seek to empower people, is it promising to disregard their preferences as irrelevant (except to forecast their behavior) and to disparage the approaches that seek to develop indicators of well-being that rely on people’s own attitudes?

It is true that there is a strand of utilitarianism (or, more generally, welfarism) that has been very influential in neoclassical economics and that treats preferences, as they are revealed by people’s choices, as sacrosanct and as the only source of information about their well-being. This approach treats people’s choices in terms of tastes to be respected and not disputed (de gustibus non est disputandum). Behavioral studies have convincingly shown that ordinary tastes and spontaneous choices are actually plagued with multiple problems of inconsistency and instability. Tastes should be disputed. But the implication is not, I believe, that people’s own attitudes are to be disregarded and that the neoclassical theory of preferences can be jettisoned as obsolete. People can still reason, and their values in life deserve respect, once they have been formed in suitable deliberative conditions. It should precisely be part of the empowerment process to assist people in the formation and expression of their authentic values. For instance, the relative importance of health and physical security, compared to other dimensions of comfort and achievement, is very difficult to determine without asking people about the trade-offs they would be willing to make. Ordinary choices about risk-taking and lifestyles may not adequately reveal people’s authentic values, but this is not a reason to ignore their views totally.1

In this process, the theory of preferences remains valuable because it provides a simple but effective way of analyzing trade-offs in the multiple dimensions of life through which people have to navigate. It provides us with concepts to understand the substitutability and complementarity of different dimensions in people’s life goals, and the relative importance of the dimensions for people in different situations. Most importantly, there is nothing in the theory of preferences that says that it can be applied only to immediate tastes. On the contrary, this theory is ideally suited to deal with well-defined values in life and to represent people’s authentic goals among whatever dimensions of life truly matter to them.

It is frustrating that empirical methods to elicit people’s authentic values are still very much debated and not well developed, because all the available data collection methods suffer from imperfections. Three main methods can be mentioned to illustrate this point. First, ordinary choices (used in the “revealed preference” approach targeted in KLMDS’ criticism) rely on spontaneous preferences, which do not track deeper values and are limited to dimensions over which people can exercise control. Second, surveys of stated preferences suffer from cognitive difficulties, even if they offer the most natural setting to experiment with deliberative procedures that would help people ponder and refine their immediate attitudes. Third, surveys of life satisfaction have also been used to retrieve the relative importance of various dimensions; they benefit from the fact that they reflect people’s experience after the fact (rather than their subjective expectations), but they involve unobserved variables and multiple causation issues that are likely to bias the estimations.

These empirical difficulties notwithstanding, it seems to me that one should build on the existing tools and seek empowerment through the reflective development of people’s authentic values in life. Respecting people should build on the (neoclassical, why not?) conceptual tools that can serve to represent how their values rank the possible lives they could have; it should develop processes of deliberation and participation in which they can form and refine their values; and it should convey them to the relevant policy-making bodies. This is still very much research in progress, but it is worth undertaking.

Perspective from the Field of Alternative Indicators

It is interesting to try to locate the KLMDS approach in the landscape of alternative indicators. Their proposal naturally belongs to the family of composite indicators that pick various aggregate statistics and, after a suitable normalization, compute an average score. KLMDS do this separately for each of the dimensions (GDP, environment, agency, solidarity), and instead of computing an overall average, they show how different countries fare in terms of the various dimensions. This provides interesting contrasts and valuable perspectives on the evolution of agency and solidarity.

In the book Beyond GDP (Oxford University Press, 2013) I wrote with Didier Blanchet, we argue against this “composite” approach on the ground, in particular, that it lacks any principle for the selection and aggregation of the various components. This negative judgment may sound somewhat fussy when conventional measures display a glaring lack of interest in certain key aspects of people’s plight, and measures like the one proposed by KLMDS can go a long way toward illuminating urgent social issues. But it may be useful to have a clear conceptual view of what would be a desirable measure before settling for whatever is possible given the limited data that are actually available.

In particular, the three prominent approaches that Blanchet and I consider more interesting all share an important feature. They are individualistic. KLMDS actually disparage what they call individualistic approaches, but this may be because they interpret individualism narrowly in terms of the substance of life. The individualistic approach to the substance of life takes a Robinson Crusoe perspective where the individual’s flourishing depends on personal achievements, especially material success. This is obviously naive and toxic. But the type of individualism that seems important to preserve is a concern for each individual’s flourishing in social evaluation, as opposed to relying on average measures that can ignore the left behind. When KLMDS incorporate a measure of trust in their indicator of solidarity, they take the average score in national surveys, thereby ignoring the wide gap between those who enjoy trust-filled environments and others who live in constant fear. An individualistic measure of well-being is one that seeks to track how each individual is doing, and this is the only type of approach that can record correlations between the various dimensions of life. Disadvantage usually takes the form of an accumulation of ills and injustices, and it can be accurately measured only by statistics that survey individual situations in a comprehensive way that fully captures this “intersectionality,” as social scientists call it. We may lack such comprehensive surveys in many countries, but taking an individualistic approach as the focal method, before compromises are made with the existing data, is of paramount importance.

Incidentally, KLMDS also downplay the importance of measuring inequalities, arguing that inequalities are not always bad because they may correlate with opportunities that convey a sense of agency. However, the literature on inequality has shown that inequalities of opportunities tend to correlate very strongly with inequalities of outcome in the now famous “Gatsby curve.” Unequal societies offer less, not more opportunities. But more importantly, one can agree with KLMDS that inequalities in income and wealth are incomplete measures of social fragmentation. But, far from pushing us away from looking at inequalities, this should encourage us to look at all the dimensions of inequality. The feeling of disenfranchisement and loss of dignity that was mentioned earlier definitely belongs to the domain of inequalities.

The three individualistic approaches that are worth considering (according to Fleurbaey and Blanchet) differ in how they incorporate people’s own attitudes into the measurement. On one extreme, Amartya Sen’s capability approach is exclusively focused on agency, albeit in a very broad sense, because capabilities depict the possibilities, as opposed to final achievements, open to people in all aspects of life. This approach does not make much room for individual values and preferences, because Sen considers that the relative importance of different dimensions of life should be decided collectively on the basis of a deliberation on values, without allowing much variation across individuals and social groups about such values. But this approach is still willing to rely on people’s values, even if that is through a collective deliberation filter.

On the other extreme, subjective well-being approaches trust the declarations of satisfaction or happiness that people make and propose to take these declarations as a direct measure of people’s well-being. A key difficulty with this approach is that, while these data convey relevant information about how people feel, they may not provide levels of well-being that are directly comparable across people with sufficient precision or with suitable ethical relevance. They depend not just on people’s situations or how much importance they attribute to various aspects of their lives, but also on how they grade their situation against benchmarks that are likely to differ across people (such as previous experience, personal goals and expectations, reference groups). It is quite unlikely that a Nigerian grading her life at 7 is really better off than a European grading hers at 6, for instance.

In the middle between these two extremes, one finds approaches that rely on people’s ordinal preferences but construct interpersonal comparisons in terms of resources or opportunities. The equivalent income measure is a prominent example of such an approach and consists of adjusting people’s incomes for variations in their quality of life (such as health or social status), relying on people’s own willingness to pay to compute the adjustments. This method can accommodate as many extra dimensions of life as one wishes, provided that preferences over these dimensions can be estimated.

Although I have a personal attraction to the last method, I think that all three deserve to be pursued and further developed. For each of them, KLMDS’ invitation to put the spotlight on agency and solidarity is very timely and useful. Once measures of empowerment and belonging can be developed at the individual level, surveys can be administered that explore how these dimensions can be incorporated in a measure of capabilities, how they relate to subjective well-being, and how much importance they have in people’s preferences over life for incorporation in equivalent income computations.


On the basis of these remarks, I would like to propose practical steps toward further developing the KLMDS approach and, in particular, the incorporation of agency and solidarity into measures of collective welfare.

Regarding either measure, it would be interesting to ask people about their sense of control over their lives and their sense of belonging, and, in particular, about the concrete elements that contribute to this sense. For instance, how much depends, in their sense of control, on their personal productivity (education, health), security (e.g., job stability), and freedom (rule of law, control over one’s work and working conditions), as opposed to participation in collective endeavors (civil society) as well as in group and political decisions, and a sense of proximity to decision centers? How much depends, in their sense of belonging, on informal relations of giving, trusting, and receiving (which are emphasized in the KLMDS measure), as opposed to formal institutions of solidarity involving associations and public agencies? Civil society, which is often cited as a key player in a potential renewal of institutions, could probably appear prominently in both empowerment and belonging measures.

Once the concrete elements forming each dimension are identified, one could examine how they correlate, at the individual level, with other measures of achievements such as income, social status, family situation, subjective well-being, or health. Health and education are currently incorporated into KLMDS’ measure of empowerment, and the correlation with other components of empowerment deserves careful scrutiny as well. One can, for instance, worry that people’s sense of disempowerment does not square well with the growing outlook of empowerment in KLMDS’ data. This may be attributable to various issues, such as KLMDS’ focus on averages rather than the lower part of the distribution for each indicator, but it may perhaps also be an issue of selection and weighting of the various components.

Finally, once these measures are documented, it would be most interesting to identify the policy and institutional levers that can be put in motion in order to address the current crisis in these two dimensions. The most exciting and promising aspect of KLMDS’ proposal is the promise to open up a new debate about instruments that goes beyond the traditional welfare state and redistribution recommendations. Putting people back into command of their lives and bringing them back together in a socially cohesive society is a most challenging task, but this is what needs to be done, and it requires imaginative reforms and policies.

Author Biography

Marc Fleurbaey is CNRS senior researcher and professor at Paris School of Economics. Author of Beyond GDP (with Didier Blanchet, CUP 2013), A Theory of Fairness and Social Welfare (with François Maniquet, CUP 2011), and Fairness, Responsibility and Welfare (OUP, 2008), he is one of the initiators of the International Panel on Social Progress, and lead author of its Manifesto for Social Progress (CUP 2018). He was professor at Princeton University from 2011 to 2020.

For more information, visit http://sites.google.com/site/marcfleurbaey/Home.



KLMDS also point out that neoclassical theory posits that well-being comes only from selfish consumption of goods and services. It is a fact that neoclassical theory has primarily been used to analyze market behavior; therefore it focused on preferences over goods and services, by necessity, but I think that there is nothing in the theory that imposes restrictions on people’s ultimate goals. It is unfortunate that a widespread focus on economic transactions and on the benefit people derive from such transactions, both in public discourse and in economic analysis, has made people think that well-being comes primarily from economic activity, or that economists believe so.