Marcel Mauss: French sociologist/anthropologist
In the field of microeconomics, the notion of sunk costs is commonly invoked to promote rational decision making with regard to decisions that influence the future. The notion of a sunk cost is fairly easy to understand. For example, consider a firm that has invested a significant amount of resources into a new form of technology that they expect to become very profitable. After some time, it becomes apparent that a rival technology supported by a competing firm is filling the same niche and the piece of technology invested in by the first firm is unlikely to become profitable. In microeconomic theory, the money invested in the first piece of technology should be considered a “sunk cost” and the rational firm would be wise to divest, regardless of the size of the original investment.
Of course, we know that humans are not always rational and it can be very difficult to move on from something one has invested in. For the first firm, the original investment established a particular kind of relationality; perhaps even an emotional connection that complicates divestment. This is one reason why it can be so difficult for a gambler to walk away when they have already lost a lot of money at a casino. On a greater scale, it is one reason why it can be so difficult for a nation to abandon a failing war when many lives have already been lost. It is the logic that says we can’t quit now after everything we have invested in this….
In my discipline, sociocultural anthropology, we tend to look with disdain at the practitioners of economics, our sister social science, for their assumption that “rationality” predicts human behavior. We often cite the notion of hau, the Maori word reflected on by sociologist Marcel Mauss almost a century ago, which describes the “spirit” that binds together a giver and receiver. We (anthropologists) know that when people make decisions, far more than rational profitability influences us. Let me defend the economic theorists though and note that the fact that a theory of sunk costs even has to exist reveals that economists know that models of rationality do not always predict human behavior (even if they believe that they should). That is why microeconomists need to have a normative theory to tell us that “no matter what you have invested in something, it is always irrational to allow sunk costs to influence your decision making.”
There is another kind of cost that deserves similar attention, as it is commonly ignored for reasons that microeconomic theory would call irrational. I call these doctrinal costs. Doctrinal costs are the costs that one is willing to incur to allow an action to occur in a way that one finds gratifying. A doctrinal cost is, in essence, a form of opportunity cost, which are simply the the costs that are implicit in making one choice over another. If I make a decision to eat an apple instead of an orange, my opportunity cost in that decision is that I did not get to eat an orange. If I choose to live in a very rural area over an urban one, one opportunity cost I might incur is that I will have to travel a greater distance to shop for groceries.
I choose to give special attention to doctrinal costs because of the magnitude of their expense in our society. Opportunity costs are implicit in all decisions and the majority of them exist in transactions that are completely rational. However, decisions that incur doctrinal costs are, just like decisions based on sunk costs, categorically irrational.
This principle is important primarily within the realm of behavioral economics, and while it can manifest in otherwise seemingly neutral market transactions, it is most at play in decisions impacting policy. Consider the following example:
There are masses of sociological data that demonstrate that violent crime (murder, assault, etc) correlates with poverty.* It would be fair to reason that investing in the elimination of poverty might be an effective way to reduce the overall incidence of violent crime in our society. However, we choose to respond to these phenomena by investing in disciplinary methods, such as a more expansive penal system and heavier policing. While we may profess that our primary goal is the elimination of violence in our society, we do not choose to invest in the methods that are most likely to achieve that result. It is ultimately more important to us that we respond to the problem (violence) in the terms that we prefer (discipline). It simply feels better to just punish wrongdoers. The actual ineffectiveness at reducing violence is the doctrinal cost of our decision to respond to the issue on our own terms, as opposed to a less satisfying method that might be more effective. Economically, it is irrational.
The sociopolitical world of the United States is laden with doctrinal costs, though they are not generally recognized as such. There are several examples:
-insistence on abstinence-only sex education, in spite of evidence that rates of unwanted pregnancy and sexually transmitted diseases are better reduced through pedagogies that teach “safe sex.” The unwanted pregnancies and higher incidence of STD’s are the doctrinal costs of an educational method that prioritizes a set of cultural values over outcomes.
-refusal to negotiate with parties that have been deemed “terrorists,” even when doing so might reduce net suffering.
-violent military responses to security threats that ultimately serve to raise rates of terrorism in the future.
-investing in the aggressive drug testing of welfare recipients, though doing so is more costly than actual welfare fraud.
As is the case with sunk costs, the people who incur doctrinal costs do not generally admit the doctrinal root of their reasoning. For the firm mentioned in my original example, the continued investment in the failing tech will be justified with the argument that the firm believes that the technology will still become successful (even though evidence suggests that it will not). In the example involving reducing violent crime, the argument will be made that increased policing and carceral penalties will work better than other methods, even though evidence suggests otherwise. People will continue to argue that wars reduce terrorism and that abstinence-only education is the most effective manner of reducing rates of unwanted pregnancy and venereal disease, in spite of the fact that analytical evidence contradicts all of these claims. Ultimately, all of the above cases are matters of people choosing to respond to problems (real or imagined) in the manner that they find most gratifying, rather than in the manner best supported by research and data.
As an anthropologist who has been critical of the tendency of economists to fetishize rational behavior, this is not an easy issue to tackle. When we invoke the hau against contemporary economic theory, we do so because we believe that rational market figuring shouldn’t be the only thing people consider when we make decisions. We are complex beings who negotiate multiple arenas of reason. Thus, a hypothetical cultural anthropologist from a different culture who isn’t invested in the project of combating hegemony in the United States, might argue that the economically irrational reasoning in the cases listed above should be considered rational due to sociocultural context. If an economist argues that supporting abstinence-only education is irrational, our anthropologist might argue that the fact that many Americans believe that they are being judged by a deity is important and something that we should respect. For what it’s worth, we anthropologists are better at appreciating those things in cultures other than our own.
I will save that challenging argument for another article. For now, I want to argue that a proper socioeconomic take on the “post-fact” era is one that recognizes that enormous opportunity costs are incurred in the name of schadenfreude and dogma.
* “Crime” and “poverty” are both complicated terms. As any good anthropologist will tell you, “crime” is a flexible category that varies based upon the sociopolitical contexts it occurs within. In the political context of the United States, it is used more to define people and places that deserve to be punished than as a measurement of social suffering.