On ethics and economics

by
Lars Pålsson Syll
My articles My siteAbout meMy booksMy videos
Follow on:Twitter

from Lars Syll

Justice is an ambiguous concept. We use it daily and constantly. But what is justice, really? What should be considered just? How do we measure justice? Is justice the same as equality and impartiality? And is there only one form of justice, or do different notions of justice coexist?

These are important questions to try to shed light on and answer. Otherwise, the concept of justice risks becoming just one of many empty and abstract clichés we use to dress our own interests in prettier clothes or simply to express our opinions. Concepts are like glasses; if they don’t help us see better, they make everything unclear.

Justice is a value-laden concept. In most contexts, it is used to express the belief that a distribution among different individuals of something — goods, well-being, wealth, primary goods, opportunities, capabilities, rights — in some respect is ‘good.’ By good, it is usually meant that the distribution is egalitarian or that there are justified reasons for any differences in the distribution. Examples of such reasons may be needs, merit, freedom, or rights.

All sciences have their specific approaches to analyzing the concept of justice. Although justice is less frequently measured for the economic theory’s mill, here I will try to illuminate — and raise question marks about — how justice is treated in economic science.

In the older — utilitarian — economic welfare theory, the decisive criterion for implementation, choice, or evaluation of alternatives was the total utility generated. However, during the 1900s, the possibility of scientifically comparing the utility between different people as assumed by the utilitarian calculus was heavily questioned. Utility comparisons were labelled as ‘normative’ and ‘subjective.’ The criterion that remained for economists to use was the ‘positive’ and ‘objective’ Pareto efficiency. According to this, a state of society is better than another if it is better (yields higher utility) for at least one individual and not worse (yields lower utility) for anyone else. If both conditions are not met, the criterion cannot rank the alternatives. In simple terms, a state is Pareto efficient if no one can be better off without making someone else worse off.

Pareto efficiency says nothing about how welfare is distributed, and therein lies the heart of darkness! Since theories of justice are precisely about how to distribute something among different individuals, we understand why economists seldom have anything to say about justice.

If A and B are to divide ten dollars, and A gets ten dollars, and B gets zero dollars, this is an efficient distribution because any redistribution would make A worse off. Pareto efficiency can describe states that are as unjust as possible.

For economists, it is evident that efficiency is a necessary condition for choosing a state, and any theory attempting to indicate a just distribution can only mean extensions of efficient states. However, for most others engaged in justice theories, it is obvious that Pareto efficiency is quite useless for making statements about justice. To say something meaningful about how to distribute something among different individuals, one must make comparisons between these individuals. But precisely these comparisons are excluded by the economic concept of efficiency!

The dimensions of justice are often different from those of efficiency. When we, for example, speak of fair housing, we may be more interested in how we distribute housing resources based on needs rather than utility. A reallocation from a situation where everyone is guaranteed housing under equal conditions to a situation where those with lower incomes receive worse housing standards while those with higher incomes receive better housing standards may be Pareto efficient in the utility dimension. This is despite the fact that we may find the distribution unfair or otherwise unreasonable from the perspective of needs, as those with low incomes have their housing needs less satisfied while those with high incomes have their housing needs more satisfied.

The fixation on efficiency, because it only considers the utility dimension, often excludes what is genuinely fair. If the production and consumption of housing mainly concern needs and capabilities rather than utility, the economic theory’s recommendations can be misleading. People often seek to improve their economic situation, but not if it leads to – or is based on – serious injustices. Economy and ethics are connected, and even well-grounded economic analyses can lead to entirely erroneous conclusions from the perspective of justice.

Traditionally, economists have not prioritized questions about distribution and justice. Efficiency considerations push justice issues into the background. The economist’s task is considered to be to indicate how efficiency is achieved and leave it to politicians to take action to correct any unacceptable effects on income distribution and the like. The disinterest in justice issues partly stems from a reluctance to use explicit value judgments and the lack of analytical tools equivalent to Pareto efficiency for justice issues in economic science.

To provide relevant political and practical guidance, economists must be able to assess situations where efficiency-sanctioned changes are impossible. One possibility to extend the economic welfare theory’s analysis to explicitly include justice issues is to invoke equality as non-envy. If social assets are distributed so that no individual would prefer someone else’s allocation, equality is said to prevail.

This demand for justice, however, can conflict with the requirement of efficiency. Even if everyone would be better off with a redistribution, the equality condition may reject it. If A and B are to divide ten dollars, and the only possible distributions are zero dollars each or six dollars to A and four dollars to B, only the former distribution is envy-free but clearly not efficient. There is an apparent tension between the conditions, and which one should be chosen is an open question.

Another possibility to move forward is to use the concept of reasonableness (fairness). A distribution is said to be reasonable if each actor prefers their share over anyone else’s, i.e., if there is no envy. For example, if A and B are to divide a cake, one can let one person cut and the other choose. This approach ensures that both get a share they find at least equally ‘good’ as the other. If the cake being divided is not homogeneous — or if preferences are not identical — strict reasonableness (fairness) is required to achieve a fair distribution. If A likes chocolate more than oatmeal cookies, and vice versa for B, if she cuts the cake, she will divide it so that the smaller piece contains more oatmeal cookies. Both will then, according to their own preferences, receive more than half of the utility — the distribution is strictly reasonable.

The concept of envy used in these cases means making utility comparisons without making them interpersonal. And that is, in fact, the reason economists have become interested in envy-free distributions. The dream of the ‘positive’ and ‘objective’ criterion of justice can thus be kept alive.

However, reasonableness can come into conflict with the requirement of efficiency, and consequently, we may have an unjust distribution that each individual prefers to a strictly reasonable distribution. Reasonable solutions are also highly sensitive and can be undone by a change perceived as reasonable by all parties. For example, it has been shown that if one starts from an initial distribution that is reasonable, redistributions that are advantageous to all parties can result in an allocation that is non-reasonable for at least one party. Granted, the problem disappears if the reasonableness requirement is replaced with a requirement of super egalitarianism – a distribution that gives each individual a utility at least as great as that of an ‘equally divided’ distribution. But the fact remains that if we extend the reasoning to include the provision of utilities, there are no reasonable solutions. The same applies if we include ‘personal goods’ — goods that cannot be transferred from one individual to another, like health and intelligence. The existence of such goods means that mutual envy between individuals can occur, and it cannot be eliminated.

The conclusion we must draw is that these extensions of the economic welfare theory do not take us very far in the discussion of justice. At least not as long as they adhere to the requirement of not making comparisons between different individuals. Non-envy does not exclude significant welfare differences. Even if the non-envy requirement is met, an individual may still be significantly worse off than others. And such a requirement of justice is hardly fair!

Justice and efficiency are not necessarily inherently incompatible. Conflicts arise only when the concepts are applied, and we must clarify what kind of justice and efficiency we are talking about. But both efficiency and justice are in some sense about goal fulfilment, and if the goal dimensions are sufficiently well-specified, our analyses should be able to accommodate both.

The advantage of using the concept of efficiency is often said to be that it is uncontroversial and value-neutral. But is it really? Statements about efficiency are based not only on facts but also on values. Efficiency must be related to objectives, and these are inevitably of a subjective nature.

Putting forward demands for efficient municipal housing policies is value-laden. Here, the word ‘efficient’ functions as a positive (negative) amplifier for something that we perceive as positive (negative), namely municipal housing policies.

Thus, to say that something should be efficient is, in itself, a value judgment. As long as we only assert that efficiency is a necessary condition for a fair distribution, the judgment is often of a more modest nature. But if we also claim that efficiency is a sufficient condition for justice, the situation is different. Demanding efficient municipal housing production may be compatible with both small and large production. However, for most people, the two cases are not equally acceptable or ‘just’ — even if they happen to be efficient. A society where a small group of wealthy people lives in castles while others live in hovels may be efficient, even if it would not require significant sacrifices from the former group to allow the latter to move into housing of acceptable quality. The fact that a distribution is efficient does not mean that it cannot be improved.

If we are interested in saying something about justice, we cannot avoid making comparisons between different individuals. Otherwise, we can only speak about cases where everyone is better off. However, this becomes quite uninteresting, as we know that in most real situations, someone wins and someone loses.

Among moral philosophers, it is widely agreed today that a person’s well-being cannot be measured solely based on the consequences of goal fulfilment but that the person’s real opportunities, freedoms, and rights also play an independent and important role.

However, economic welfare theory attempts to reduce the real pluralism and diversity of utilities, information, and considerations to a single homogeneous descriptive and analytical category — utility. Admittedly, this is necessary to fully rank different utilities or states in a preference order that can serve as the basis for a Pareto-efficient choice. But wouldn’t it be better to accept that sometimes we can do nothing but incompletely rank different states? In the choice between different forms of housing, we may only be able to observe that they all have different advantages, but it is not possible to unambiguously rank one as better than the other.

Economic theory has largely been built on the very narrow assumption that individuals are guided by self-interest. In many choice situations, this may be unproblematic, but in other contexts (e.g., when designing municipal housing policies), the assumption of self-interest is directly misleading, as the individual’s utility in this context typically also includes consideration for other people.

When we choose between different goods in a store, our choice clearly reveals our preferences. However, when we, for example, consider the extent of municipal housing construction, it may no longer be the case. Self-utility does not necessarily determine individual choice. An individual may prefer a high and uneven housing standard but still decide on a lower but more uniform housing standard. Choices and preferences do not need to coincide.

One of the deficiencies in traditional economic welfare analysis is that it assumes unspecified utilities. Several studies, however, have shown that our perception of what is fair is highly influenced by what the utilities consist of. People react differently when the utility is based on needs or merely ‘taste and opinions.’ The person with greater housing needs, due to, for example, a disability, should be compensated with a larger allocation, whereas differences in taste and opinions should not be compensated in the same way. This indicates that the utility information we typically get in economic theory is insufficient to make statements about what we consider a fair distribution.

A more promising approach is perhaps — as Amartya Sen has suggested — to give greater weight to positive freedom (the ability to achieve desired goals) rather than just negative freedom (the absence of external constraints) in welfare evaluations. Welfare assessment must go beyond traditional welfare measures that only consider commodity possession and real incomes. It is not commodity possession or perceived satisfaction that primarily provides measures of well-being, but our ability to function and utilize our assets. For someone, a house may be just a place to sleep, a necessary evil. For others, it may be a fortress around which their entire life revolves. Because people are different, the ability to transform resources into capabilities for action is also different.

There are no shortcuts to justice. But if we want to say something of interest about how our assets should be distributed, we must accept making comparisons between individuals and groups – even if doing so means relinquishing the dogmatic certainty possessed by Pareto-fixated economists. The gain in terms of increased usability outweighs this.

In economic theory, instrumental rationality has long been dominant. This has, among other things, led it to strive to establish itself as an independent science, free from other social sciences and philosophy. In this process, a deep divide has arisen between economic theory and ethics. By not explicitly considering people’s considerations of right and justice – and how these affect our behaviours and actions – the theory has increasingly distanced itself from reality. The narrow assumption of self-interest and utility maximization as the sole driving force of individual behaviour has meant that economists have had little interest in discussing rights and justice in our societies. But economists’ efforts to build a firewall between science and ethics have only led to ethics being rendered invisible. It is still there, albeit unaccounted for and often unconscious.

Nevertheless, economic science does not need to hide value judgments and subjectivity behind presumed objectivity. By bridging the gap between economic theory and ethics, it is possible to create a more relevant economics that does not regard humans as selfish receptacles for utility but as exhibiting all that we usually associate with the concept of human beings — including notions of justice.