Issue 01 / The Ethics of Measuring Uncertainty

The Ethics of Measuring Uncertainty


Krisha Kops

“Whatever we, as prospective participants unaware of our specific features, would desire society to be like is what, morally speaking, we ought to institute.”
- Thomas Pogge

Within ten years from now, poverty should be ended in “all its forms everywhere.” At least, that was the central commitment of governments proclaimed at the 2015 meeting of the United Nations General Assembly in New York when the 2030 Sustainable Development Goals (SDGs) where adopted. The successful implementation of this plan has yet to be seen. Current governments and international agencies largely promote GDP growth while regarding the eradication of hunger and poverty as something that will eventually be realized by means of economic growth. However, as Thomas Pogge has pointed out, few realize that extreme poverty is not a mere economic condition, but a harm structurally inflicted upon the global poor. Pointing to the social component of poverty, he argues that its meaning is as diverse as the different people and contexts upon which it can be inflicted. Close contact between government and the governed is therefore essential for successful development policies. Hence, while econometric and political analyses have for long proceeded as if data were simply to be handed down, asking what poverty actually means could be a first step to effective aid and global justice.

Thomas Pogge is one of the leading voices in debates on health, poverty and human rights. He holds a position as Leitner professor of Philosophy and International Affairs at Yale University, where he is also the founding director of the Global Justice Program. In his work, Pogge is influenced by the thoughts of philosopher John Rawls (1921-2002), who supervised the doctoral dissertation Pogge completed at Harvard University. However, pointing to the difficulty of applying Rawls’ theories to the international domain, Pogge worked on developing his own approach to questions of justice and humanitarianism throughout his career. His publications include Health Rights (2016) and World Poverty and Human Rights: Cosmopolitan Responsibilities and Reforms (2007). In his theories of justice, Pogge especially emphasizes the importance of negative duties (do no harm) over positive duties (provide aid), arguing that the rich must first and foremost refrain from imposing supranational institutional arrangements that foreseeably and avoidably reproduce severe poverty. He co-founded projects such as Academics Stand Against Poverty (ASAP), an international network that aims to enhance the impact of scholars, teachers and students on global poverty, and Incentives for Global Health, which promotes creation of a Health Impact Fund (HIF). This publicly pay-for-performance mechanism constitutes an alternative to patent monopolies by giving incentives to companies to develop pharmaceutics for those diseases that are not covered by market-driven initiatives. While the Health Impact Fund would reward innovators for the health gains they achieve with their registered pharmaceuticals, it would limit the price at which they could be sold to the costs of manufacture and distribution. Dr. Krisha Kops spoke with Thomas Pogge about the issue of measurements in global justice initiatives, Rawls, and questions around expert knowledge.

KK In the 2015 article The Sustainable Development Goals: A plan for building a better world? you openly criticize the SDGs that were adopted by the United Nations earlier that year. What are your main points of criticism?

TP The most important reason for my critique has to do with the shift in terminology from the language of human rights to the language of goals. When you talk about goals the idea is always that we are making progress towards something. We cannot be blamed as long as things are getting better; we might not get there as fast as we had envisaged, but we are getting there.    
The language of rights is quite different. It suggests that when there are unfulfilled rights immediate full action must be taken. For example, if you think about slavery in terms of rights, you will say that the slaves must be freed right now. If you think about it in terms of goals, you may say: “let us see whether we can liberate half of the slaves by 2030 and the other half maybe by 2050.” Of course, it is better to have a poverty-reduction goal than to have none. But given the alternative of the human rights language, I feel that the language of goals is the poorer alternative. It makes us feel comfortable with how far we have come, while the rights language stirs to urgent action against the injustice that remains.

KK In that same article you critically pointed to the way poverty is measured within the framework of the SDGs. Is there a way to obtain measurements correctly when it comes to poverty and other issues?

TP In regard to measurement there are two, maybe three different issues to distinguish. One is the questions of who is measuring. Here integrity is of particular importance. The second issue is related: when do you decide about how you measure? Measurement methods ought not to be changed midstream. In my essay The Hunger Games (2015) I recounted how officials were dissatisfied with stubbornly high counts of the undernourished. In order to report more progress in their final accounting Millennium Development Goals (MDGs), they changed their methodology in 2011 and decided to measure in a different way near the endpoint of the 1990-2015 MDG period. This is of course completely unacceptable. However, this is bound to happen with international agencies in charge, as they are funded by governments and their leaders are appointed by intergovernmental consensus.    
A similar instance happened this past July, when the UN Food and Agriculture Organisation (FAO) announced that it had in recent years overestimated undernourishment in China by a factor of twenty. Measurements that bounce around like this are obviously unreliable. International agencies are often obliged to accept numbers given to them by a government, if only because they need government support to do their work. Therefore, if we want to know what is really happening, we should entrust measurement to independent experts who lay down their methods in advance and then stick to them until the end of the reporting period.

KK Above-outlined examples of independent measuring seem to be objective at a first glance. Yet, how can it be assured that also those who are actually affected are included into the debate?

TP I was involved in the creation of the Individual Deprivation Measure (IDM), a new way of measuring poverty resulting from a large four-year project founded by the Australian Research Council. We started by talking to people in economically underdeveloped regions. For this purpose, eighteen locations were selected; one rural, one urban, and one with a disadvantaged minority from six different countries that included Angola, Fiji, Indonesia, Malawi, Mozambique and the Philippines. We went there very informally and talked to people about what poverty means to them. What are the characteristics of poverty? How do you recognize and rank ‘poor people’? We got a long list of characteristics, developed it into a proto index and then went back to the same people to ask whether it made sense to them. With their feedback, we refined it some more and finally arrived with a fifteen-dimensional metric for measuring poverty.     Interestingly, violence was one of the dimensions. Initially I suspected a translation mistake, but when we thought about it, we realized that the people were right. Being poor means being vulnerable to violence. It means you have no home with a front door you can lock to protect yourself, no real access to police or judiciary. If people beat or rape you, you are essentially without recourse. As others know this you are treated as easy prey.

KK Speaking to the affected group is one way to create more inclusive measures. Are there additional ways to avoid reproducing certain power structures in humanitarian practices and policies?

TP Any serious research on poverty must prominently involve poverty-affected people themselves because they know it first-hand, and because they have most at stake in how poverty is measured and addressed. Still, we should not simply hand over the exercise fully to a group defined as being poor. There are sophisticated methodological requirements for a poverty measure that supports diachronic and international comparability. Left to their own devices, affected groups could provide a vivid account in their local language of what it means to be poor in their community. This is interesting and important input into a poverty measure. Yet, experts trained in social science and statistics are still needed to construct the measure.

We tried to moderate the power and preconceptions of the experts involved through what we called ‘deliberative participatory research.’ Outside experts participated in group discussions with local community members and deliberated with them about how best to clarify and systematize their thinking about poverty. We did not simply record the views community members held, but rather invited them to reflect upon such initial views in conversation with one another and the experts. Such group exercises took four distinct forms, including guided group discussions, group exercises to generate poverty ladders that enable community members to discuss different levels or categories of poverty, group exercises to rank different dimensions of poverty in terms of their relative importance, and household mapping exercises to discuss how poverty may vary by gender and age within the household. These group exercises were followed by a final set of in-depth individual interviews to revisit any key points or themes from the group exercises that required further elaboration.

KK In 2007 you began to elaborate on the Health Impact Fund (HIF) with other researchers, aiming to encourage pharmaceutical firms to develop new remedies against hitherto neglected diseases and provide remedies at affordable prices. How does the HIF reflect upon the question of adequately capturing and measuring ‘health’?

TP I do not think there is such a thing as correct measurement. The HIF assigns value to pharmaceutical treatments, an exercise that has a substantial normative component about which there can be reasonable disagreement. For example, one can argue about the inclusion of contextual and person-specific factors such as profession, lifestyle and location that influence how important a specific health gain is to a particular person. To arrive at a manageable measure of value we concluded that we must leave most of these interpersonal differences out. Instead, we rely on averages by asking how important specific health gains typically are. Here one can differentiate by factors such as age or gender or certain genetic traits when these make a substantial difference, but not by factors such as profession or economic potential. In specifying the details of the metric and the methodology for data collection and processing, the HIF would be able to draw on decades of experience and debate involving different versions of quality-adjusted life years (QALYs) as for example deployed by the Institute for Health Metrics and Evaluation (IHME) in their periodic studies of the global burden of disease. For the specific purposes of the HIF this approach would have to be refined with the help of stakeholders, theorists and experts. The result will be a compromise that few will be completely satisfied with. Yet, we are confident that all will agree that the HIF achieves a vastly better value-reward correlation than the present patent monopoly system.

KK Academic practices and policy making are undercut by specific nodes and strategies of sense-making. In order to obstruct these systems of power, would it be preferable to include other, non-prevalent knowledge systems into existing structures?

TP We should definitely look at them. I doubt we would learn a lot about how to do quantitative social science from other cultures. We have done this for quite a long time and have come up against all sorts of different problems of measurement, comparability, statistic validity, and so on. Perhaps I am prejudiced, but these things are pretty hard and fast. There are no radically different ways of doing statistics. This does not account for qualitative research, which is a whole different thing. I am fascinated to see what goes on in other knowledge systems. There you often have very creative ways of illuminating something. Not with numbers, but with descriptions. Language matters here. It is intriguing to see in what terms other cultures describe a certain problem. Even in our Indo-European languages there are words that exist in one language and not in another. If you have this word, you can understand a certain phenomenon. A good example is the German word Schadenfreude (‘enjoyment from the troubles of others’), which does not exist in other languages. It beautifully captures something you would not capture with the same clarity, if you had no word for it. We live and conceptualize our experience in language. Therefore, language influences how we think and feel. With richer language tools, we get a deeper understanding of what is happening in a social world; and we can really grasp the experience of others only by sharing their understanding of the terms in which they articulate their experience.

KK Your former teacher John Rawls created a theory of justice foundational in debates on global justice up until the present day. However, critics of Rawls point out it is exclusively embedded into the history of ‘Western’ thought. As an alternative, the economist Amartya Sen includes also Indian philosophy in his The Idea of Justice (2009). What is your view on the absence of non-Western perspectives in dominant justice theories?

TP Non-Western philosophy should certainly be included. However, Rawls self-consciously abstained from this. Citing from his work Theory of Justice (1971), he primarily wanted to systematize “one (educated) person’s sense of justice. Here he started with himself, seeking to bring his own settled moral convictions and factual beliefs at all levels of generality into a coherent ‘reflective equilibrium.’ His hope was that other educated people of his time and culture would find this systematization compelling. This hope can reasonably extend to the Anglophone world of the latter half of the twentieth century, but not far beyond that. Rawls would have acknowledged that if people from another era or culture used his method of reflective equilibrium, they would probably arrive at quite a different theory of justice. Even when fully developed, our sense of justice remains bound to the considered convictions of our own culture and era, even though we apply it far beyond this limited scope by making justice judgments about for example feudal France or present-day India.    
Rawls leaves two openings for non-Western perspectives to enter. By seeking a wide reflective equilibrium, he is recommending that we try out several quite different ways of unifying our considered convictions into one theory. Thus, Rawls spent many years trying to work out the most compelling utilitarian theory. He felt he should accept the best specification of his own contractualist approach only after having compared it with the best specification of a utilitarian theory of justice. Developing the best versions of alternative approaches may in some cases change our considered moral convictions, and Rawls was quite open to this. I think he would have agreed to work through non-Western approaches as well, though in fact he did not devote significant time to this effort. As non-Western approaches are more remote from his own moral convictions, he possibly thought that their detailed study would not enrich or influence his theorizing about justice as much as closer-to-home approaches such as utilitarianism.    
The other opening for non-Western perspectives exists in his international theory, The Law of Peoples (1999). Here he argues that an ideal society of peoples should have room for a wide range of differently organized societies. Some of these acceptable societies would be organized according to some version of liberalism, including ones quite different from his own. Others would be ‘decent societies’ that perhaps lack a separation of church and state, full freedom of expression or a democratic political system. Rawls sketches the set of decent societies only in broad stokes, but clearly assesses them from the standpoint of his own liberal values. Although decent societies are to be accepted into the ideal society of peoples as ‘members in good standing,’ Rawls hopes that they eventually reform themselves into liberal ones. For our moral theorizing, he did not think that he or we have much to learn from the moral conceptions that organize the different decent societies.

KK Is this the contribution of your own work, as you attempt to elevate Rawls’ theories to a global level?

TP I follow Rawls insofar as I address people in the West, arguing that if we want to live up to our values we must stop designing and imposing supranational institutional arrangements that reproduce severe foreseeably and avoidably deprivations on a massive scale. But I am also going beyond Rawls by seeking supranational institutional reforms that can be justified in terms of the values of diverse cultures. Some of these reforms are the following: a Global Resources Dividend (GRD) would ensure that all human beings partake in the value of the planetary resources we consume or degrade, like metals and minerals, water and air. The HIF would reward new medicines and vaccines according to the benefits they bring to human beings, regardless of their socio-economic position, thereby incentivizing research into the diseases of poverty, resulting in new pharmaceuticals priced near cost of production. And a ban on ‘national nepotism’ would commit participants in supranational rulemaking to reason and decide impartially in light of justice and the common good, and without special regard for their home country. In a similar vein are national politicians expected to act for the good of the whole country without special regard for their home province, hometown, or family ties. These reforms can be justified from within diverse cultures and can begin humanity’s advance toward a world-order based on shared values; a transformation urgently need to assure human survival in the face of the dangers of war and environmental catastrophe.

KK Who is responsible to bring these different theories and approaches into dialogue?

TP Academics and foreign-policy officials are best prepared to make a start. They can gain a rich understanding of the values and perspective of another culture. Such an understanding is required to appreciate members of another culture as moral persons. Not moral in the sense that they share our morality, but moral in the sense that they have a sincerely held morality that they are committed to honour even at substantial cost. Once we know this of each other across a cultural boundary, we can explore together what moral principles and institutional arrangements we can jointly commit to from our diverse respective moral conceptions. The key here is that each side understands in some detail where the other is coming from. Today, all too often, each side understands only its own values and happily uses them to condemn the other side for all their supposed crimes and failures. No firm moral common ground can be established in this way and all we then have are the ever-shifting international arrangements based on bargaining power and threat advantage, arrangements that are routinely abandoned or renegotiated with every major shift in the perceived power or interests of major states. Such a modus vivendi regime affords no long-term security for societies and their values and entails a perennial danger of major military confrontation.

KK We do not only face differences when it comes to cultures, but also concerning gender and ethnicity. Something Rawls ignores in his ‘ideal theory,’ which stands for a perfect societal structure based on idealised presumptions.

TP Rawls calls for building a social world in which his principles are fully satisfied. He thought of this as a two-stage process. First philosophers design the ideal, then some non-philosophers worry about how to get us there. This seems far too ahistorical. An ideal should be constructed in light of historical possibilities. Advocacy of an ideal is itself an historical event that depending on context may advance or impede progress toward this ideal. By developing a theory that said little about the monumental injustices associated with race and gender. It is quite possible that Rawls greatly diminished the historical impact of his theorizing. To be sure, a theory of justice is not a tool for use in everyday political contests. Still, a political theorist should reflect on the political and historical role of her or his own theorizing.

KK Understanding processes of domination and exploitation as the key forces that structure societies, some scholars suggest that we have to start ethical theory itself by analyzing these very dynamics. Hence, Marxist and other scholars would argue that your thoughts still do not go far enough.

TP The two concepts of domination and exploitation are quite difficult and weighed down by historical baggage. I rarely use them, but I do analyse in both explanatory and moral terms the phenomena they use to describe. For example, I analyse unequal opportunities to participate in collective decision making and to exert political influence; and I critique economic arrangements that distribute the collective social product in a way that foreseeably and avoidably leaves some participants unable to meet their basic needs.

KK John Rawls prefers in his justice theory the ‘maximin rule,’ meaning we have to make those decisions which optimise the outcome while taking the worst-case scenario into account. Do you agree with this approach, that aims to be of the greatest benefit to the least-advantaged members of society?

TP The maximin rule designates both a principle of distributive justice and a maxim for decision making under uncertainty. In the former meaning, the maximin principle requires that the distribution of benefits and burdens of a cooperative scheme be organized to maximize the value of the minimum share. How plausible this is depends in good part of how ‘shares’ are conceived. For example, do they include just those benefits and burdens of cooperation or also such personal endowments as talents, looks, health and cheerfulness? The details matter greatly and neither Rawls nor his followers have provided an account of relevant shares on which the maximin principle would be compelling.
    In the latter meaning, the maximin rule requires that one chooses the option with the best (or least bad) outcome. As Rawls recognizes, such extreme risk aversion is clearly insane in most cases. But it may well be appropriate for a narrow set of cases where the rule rejects options that involve at least one very bad outcome whose probability is highly uncertain. It may also be appropriate in situations where the decision maker stands in a trusteeship relation to the person who will bear the decision’s consequences. Imagine you are a trustee in charge of managing an orphan’s inheritance. Arguably, in this situation you ought to be highly risk averse to ensure preservation of capital. Rawls draws on both these intuitions when he argues that the parties in his original position ought to choose according to the maximin rule. Doing so, they would then be drawn to agree on a maximin principle of distribution, at least if they can formulate one that would ensure political stability.

KK How would you then treat risks in these situations? Unlike Ulrich Beck suggested in his Risk Society (1992), is it not the reality that risks are not fairly distributed in most cases?

TP This is right. Risk taking seems most acceptable when the decisionmakers themselves bear the consequences. By raising the speed limit, a society trades some degree of safety for greater convenience. This trade-off becomes problematic when many citizens are cyclists or pedestrians who will get most of the extra risk and none of the extra convenience. In such a case and even if they constitute a majority, the drivers ought to be risk-averse and also compensate the non-drivers through fuel taxes or in other ways.      Asymmetrical distribution of risks and benefits is a feature also of the two most prominent risks of the present era: major war and climate change. Politicians can often increase their power and standing by creating crises, tensions and hostilities. The risks of war they thereby create are borne by millions of people who had no part in political decisions and derived no benefit from it. Similarly, the worst emitters are imposing risks and harms mostly on people other than themselves. Communities in the tropics, for instance, and members of future generations. Such unilateral impositions of risk are deeply wrong paradigm injustices of our time.
KK These examples illustrate that risks evoke ethical questions. Traditionally, there are two dominant ethical approaches to deal with ethical issues in the ‘Western’ tradition. On the one hand utilitarianism, which judges those deeds to be ethical which create the maximal happiness for all affected individuals. On the other hand, deontological ethics focus not on the consequences, but on whether actions are intrinsically good or bad. Are these theories adequate for handling modern-day risks?

TP Kant’s deontological ethics is based on the categorical imperative, which gives little guidance for how to deal with risk and uncertainty. A utilitarian or consequentialist ethics is well equipped to handle probabilities. However, it cannot be applied in practice as we cannot fully foresee all the effects of our decisions, even probabilistically, because these effects get comingled with the effects of the conduct decisions of countless other agents.     As John Rawls recognized, the best way forward is to shift emphasis from ethics to social justice, from the moral assessment of agents and their conduct to the moral assessment of institutional arrangements. In that way are the effects of such rules and procedures of a large social system more easily foreseeable than the effects of individual conduct. For example, we can adjust particular parameters of a country’s tax system and then observe how these changes affect the distribution of income and wealth as well as opportunities for education and employment.

KK Does this not create the danger of people denying personal responsibility, while blaming the supposedly ethical institutions?

TP No. It merely shifts our responsibility toward institutional design. We bear a collective responsibility for the ground rules of our society, because as citizens we participate in shaping and imposing these rules. By living up to this responsibility we are more likely to achieve the desired change.
    Suppose we are appalled by the poor wages and working conditions of coffee farmers. We can resolve to shift our coffee purchases toward fair-trade-certified products. This will benefit some coffee farmers, hurt others, perhaps motivate some entrepreneurs to seek fair-trade certification and some others to start a new free-trade certification business. It is difficult to be confident even ex post that our shift had an overall positive impact commensurate with the cost involved. Alternatively, we might mobilize in favour of a new law that requires coffee importers to monitor their supply chain and to make sure that every bean entering our country was grown by people who are paid a decent wage. Here we could realistically monitor the impact of the new law and be sure of its positive impact. This is the course of action I advocate. Whoever fails to help make our laws more just can blame the laws but remains co-responsible for their moral defects.

Krisha Kops received his Ph.D. in intercultural philosophy from the University of Hildesheim. His dissertation theorizes the modern philosophical receptions of the Bhagavad Gītā in Germany and India. Kops previously obtained degrees in philosophy and international journalism at London and Westminster University and works in Germany and India as a journalist for publications such as Times of India and SZ-Magazin. As a teaching and consulting philosopher, he focuses on questions of global justice, intercultural philosophy and exchange.