Branko Milanović – The Mistake of Using the Kantian Criterion in Ordinary Economic Life

Faced with the imperative of Degrowth, this is one of the most important issues that needs to be addressed.

Branko Milanović is an economist specialised in development and inequality. His newest  book is “Capitalism, Alone: The Future of the System That Rules the World”

Cross-posted from Branko’s blog

File:Covid-19 San Salvatore 09.jpg - Wikimedia Commons

In the last section of his justly celebrated “On economic inequality” Amartya Sen discusses the role of incentives in work motivation in socialist societies. He goes over Marx’s distinction between the role that material incentives would play in socialism (“a society…still stamped with the birthmarks of the old society from whose womb it emerges”) and where such incentives would remain important, and communism where material incentives would not matter and where everybody would contribute according to their abilities and would be rewarded according to their needs.

Sen then discusses the Chinese attempt to move toward the communistic reward structure during the Great Leap Forward. Sen notices that it proved to be a fiasco even if “it is difficult to dissociate the difficulties generated by the use of non-material incentives from those caused by other features of the Leap Forward” (p. 96). The fiasco was due, among other things, to the idea that individuals would not follow their self-interest, namely that they will be indifferent to the relationship between their effort and reward. Since reward was independent of effort people minimized their own effort which, when done by everybody, led to the reduction of output and eventually to famines and disaster.

Sen however considers a situation where the behaviour of individuals would be different as would the outcome. He assumes a Prisoner’s Dilemma two-person game. As is well known, the dominant strategy for each person is to defect, i.e., in the example when rewards are dissociated from effort to work as little as possible. When everybody follows self-interest only, the collective outcome is low output. This is what, according to Sen, happened during the Great Leap Forward.

But assume now, Sen writes, that the value system of individuals charges, that they do what is best for community, assuming or being assured that the others would do likewise. When all do so, the outcome of the Prisoner’s Dilemma is socially optimal (high output), and is individually preferred even if people’s preferences might have remained individualistic. Thus people had to act as if they held an altruistic value system even if their true preferences were individualistic. The Chinese perhaps were right: if self-abnegation, altruism and reciprocity, and not self-interest, could be inculcated or if people could be made to behave as if they had these values. Big ifs.

The requirement is indeed that people do not behave in a self-interested way, but in what John Roemer recently introduced follow Kantian categorical imperative: that they behave in the way that they would like everybody else to behave.* While this nicely solves the Prisoner’s Dilemma, the requirement is so remote from the way people behave in ordinary life that it does violence to any usual and real norms of behaviour.

This does not mean that under exceptional circumstances people may not behave without much regard for the balance between rewards and effort: if a ship is sinking, everybody will probably work at much as they can without looking around to check whether others are pulling their weight too. There are many other exceptional situations where the same principle might hold. Even in elections, like the ones in today’s United States, such a behaviour can be seen in people’s willingness to stand in long lines in order to cast their vote—the value of which is close to nil. (And is even less so in states that are overwhelmingly Republican or Democratic.)

People might behave altruistically in their interactions with family and close friends. There again, the balance of gain and loss, is often overlooked. But in what Alfred Marshall called “the ordinary business of life” our usual and correct assumption is that of self-interested behaviour. Moreover, that assumption has never been as justified as it is nowadays when significant areas of our ordinary life (the ones where altruism might have prevailed in the past) are “invaded” by commercialisation. What commercialisation or commodification do is to place shadow prices on many activities that in the past were subtracted from the market. Once such shadow prices exist, it is only a matter of time before more people, including those who never thought of selling and buying certain activities, begin to participate. The mass participation introduces normalcy and converts hitherto non-market goods and activities into objects of market transactions, not different in any way from the usual areas where market reigns, from purchase of food and cars, to the supply of labour or investment of capital.

The expansion of commodification into our leisure time where insensibly our pleasure in playing computer games or browsing the Internet, becomes a potentially lucrative activity (hosting adds on one’s Website or becoming an influencer), or into our family life (pre-nuptial agreements), or the rest of our lives as in willingness to forego the right to free speech for money (non-disclosure agreements), implies that the sphere of self-interest has expanded and the sphere of altruism has receded.

The realisation that this is so has important implications for at least two types of discussions that are conducted today. One is the extension of what Sen wrote fifty years ago and John Roemer recently: the idea that socialism can be introduced through a change in people’s system of values. This, from all that we have seen in the past, and that we see now is entirely unrealistic.

The same ideas are also present in discussions regarding the climate change. Here too some authors invoke miraculous changes in our behaviour whereby we would no longer care for own income and wealth but would be willing to sacrifice them for the global reduction of CO2 emissions. The argument is exactly the same as what Sen used in his Prisoner’s Dilemma: if everyone were to start behaving in the way he or she would like others to behave, climate change may indeed be controlled.

The problem is here, like elsewhere, that this type of behaviour is incompatible with the system of values propagated by capitalism (and necessary for its survival) which we have accepted and affirm in our ordinary lives: from the moment when we argue about our wage to the time when we draft pre-nuptial agreements. Thus, here too such assumptions of altruism may be useful in modelling various games, but are entirely useless in dealing with the real world and deciding on actual policies, which, if they are to lead to change, must be based on incentives and punishments.

The assumptions about one’s behaviour in ordinary life (and far away from lab conditions where they are often investigated) are crucial if we want to make things better. By assuming altruism and selfless reciprocity in economic life, we are led astray and the results cannot be different from what similar attempts have produced in the past. In fact, they have given free rein to the worst human instincts, as those very far for being so naïve have draped themselves in the sheepskin of altruism to better exploit those who either believed in such fictions or were forced to believe them.

It is only by building “on the granite of self-interest” ** that we can make a change.

* The same idea is present in Plato who in “Gorgias” and “The Republic” argues that the just man never competes against others but does what he believes to be the best thing to do. Competition is a vice.

** George Stigler regarding “The Wealth of Nations”.


The text needs two clarifications.

1) When we assume self-interested behaviour we are at the same time assuming that our interests clash with interests of everybody else. But that does not mean that we do not cooperate with people. We cooperate whenever our interest can be enhanced by cooperation compared to not cooperating. This is why we work in factories with others and not alone at home: output is much greater when we collaborate and hence our self-interest is better served.

2) In the Kantian model, we also assume that people are self-interested. But they adjust their behaviour as to do what they think everybody should be doing. So behaviour is, in some sense, delinked from the underlying individualistic motivation. In addition, the beliefs about how everybody should behave may differ. One person may believe that the socially proper behaviour is A; another person may believe it is B. Thus an additional requirement is to be imposed: that through education or socialization these beliefs should be made the same or similar.

Take the following example. My self-interest may be not to pick up after my dog. But in a Kantian fashion, I might clean up after my dog because I think that everybody should do the same. But another person might believe that not picking up the poop is actually good for the nature, plants etc. Thus although we both follow the Kantian precept we end up by doing two very different things. Hence some cultural “equalization” is necessary. Religion or ethics are supposed to do so—although one may be doubtful how well they could do it.

Branko Milanović

Brave New Europe is beginning its Autumn support campaign. If you have not read their posts, you should. Never as much as now, we need a critical, open, left-wing voice not beholden to millionaires. Even small donations matter.



Be the first to comment

Leave a Reply

Your email address will not be published.