Identity Politics and Positional Goods

A Tentative Analogy

Economists distinguish different kinds of goods. First-year students generally learn in their microeconomic (or economic principles) class the distinction between normal, inferior (or Giffen), and superior (or Veblen) goods. They also learn about public and common goods. One type of goods that doesn’t fit easily in standard classifications is positional goods. Positional goods are a sort of mix between superior and collective (i.e., non-excludable) goods. As their name indicates, positional goods serve the purpose of positioning individuals who consume them in society. The fact that Ann consumes positional good x signals to others that Ann has particular tastes, is wealthy, and more generally helps Ann achieve a social position that can be regarded as “superior” to most other individuals. The American economist and sociologist Thorstein Veblen is among the first to have provided a systematic study of consumption practices that reflect this kind of positional considerations.[1] Veblen shows how late 19^th-century American capitalism is largely fueled by consumption behavior that cannot be only accounted by the necessity to satisfy wants and needs, but also displays an urge to signal one’s social position and status.

Students generally hear about Veblen precisely when they are introduced to the difference between normal and superior goods (which explains why superior goods are sometimes called “Veblen goods”). To put it simply, when the relative price of a superior good increases with respect to other substitutable goods, its demand will not necessarily decrease but to the contrary can increase. An increased price indeed fulfills the positional function of signaling one’s wealth or status and therefore increases the value of the good for its consumer. This is not however the whole story. The value of a positional good for its consumer is partially (and sometimes fully) derived from the fact that it contributes to improving one’s social position relative to others. That means that when someone else is consuming a positional good, it worsens others’ social positions. In this sense, insofar as individuals’ preferences range over their social position, the consumption of positional goods generates negative externalities. If everything else equals, I prefer to be highly socially ranked and my social ranking is a function of my relative consumption of positional goods compared to others’, then the fact that someone else consumes positional goods has an adverse effect on me.

Hence, positional goods combine two characteristics. First, people tend to be more willing to pay for them as others are consuming them (technically, the marginal substitution rate of positional goods with respect to money increases with their aggregate demand). This makes positional goods a kind of superior goods because then their demand is not (necessarily) a decreasing function of their (relative) price. Second, the consumption of positional goods has a non-excludable dimension related to the negative externalities it generates on the rest of the population. Positional goods are in this sense akin to “public bads,” i.e., goods whose consumption is somehow necessarily joint (the characteristics of the good I consume depends on whether you’re consuming it also or not) but with negative rather than positive external effects.

Combined, these two characteristics indicate that the consumption of positional goods has significant social costs.[2] The non-excludable aspect of the consumption of positional goods already implies that market failures are unavoidable. But, on top of that, the fact that consumer’s willingness to pay for them tends to increase with their aggregate consumption suggests that we are likely to be caught in an arms race. Besides the negative externalities that their consumption generates, positional goods will crowd out consumption of, and investment in socially (and eventually, individually) more useful goods and activities. That’s why there is a solid case for taxing the consumption of positional goods at far higher levels than the consumption of “regular” goods.     

Chat GPT rendering of identity politics

As the title of this post indicates, I see an interesting parallel to be made with what is commonly called “identity politics.” That may sound surprising, but if the analogy works, then we can easily see why engaging in identity politics also has an important social cost. On the standard definition, identity politics is the fact of making political claims and organizing a political strategy based on social identities related to, e.g., race, ethnicity, religion, or gender.[3] Identity politics is therefore the antithesis of what we could call “universalist” or “cosmopolitan” politics, i.e., an approach to politics that doesn’t consider social identities as a ground to make claims about what is right and good. Cosmopolitan politics doesn’t ignore the obvious fact that people’s political views are partly a function of their social identities and positions,[4] but it claims that such considerations cannot justify by themselves any principle, rule, or policy.

Suppose that some individuals in a given society start to make claims C based on their supposed social identity S. These claims may consist for instance in demanding some advantages or exemptions to rules that otherwise apply to everybody else. If C is granted to individuals who are categorized as S, that puts non-S individuals at a disadvantage compared to the status quo. For instance, it may become more difficult for non-S individuals to be admitted to elite universities because some spots are reserved for S-individuals. This is independent of the question of whether the advantage given to S-individuals is fair or just. We have here an analog to the first characteristic of positional goods. Granting a claim based on a social identity has a negative external effect on those who don’t have this social identity.

Once the Pandora’s box of identity politics is opened, it is easy to see that this is likely to trigger a kind of arms race. Now that it is admitted that social identities provide a ground to make justified claims for an exemption or an advantage and that politics can be framed in terms of competing social identities, everybody is likely to have an interest in engaging in a similar political strategy. This leads to a backlash effect where people, seeing others making (eventually successful) claims based on their social identity, do the same to stay competitive in the race started by identity politics. Even if initially you didn’t care too much about your social identity, the fact that others have started to use it as a political resource makes it politically valuable for you to do so too. You’re more willing to engage in identity politics the more others are already engaged in it.

As with all analogies, this one is imperfect. The social cost of positional goods consumption is foremost triggered by the fact that individuals have social preferences that value one’s relative social position. If these preferences could be changed, most of the negative externalities and the arms race would be averted. There are however cases where the positional aspect is not purely subjective, as for instance in the case of the competition between individuals to study in an elite university.  The positional nature of identity politics is, at the basis at least, more related to the framing of the political landscape. Again, people may not particularly value their social identities – contrary to consumers who tend to value their relative social position. However, once the social identity framing starts to be acknowledged and effectively used, it becomes very difficult to ignore it and stick to cosmopolitan politics. To do it risks ending up putting you at a serious disadvantage.

Consider the effects that identity politics has within a standard model of political competition. In such a model, we may assume that the median voter theorem applies. Political platforms should therefore converge toward the median voter. But if politics is framed in terms of social identities, both voters and platforms will tend to put more emphasis on social identity claims. Political platforms that don’t do it risk losing voters who feel unacknowledged. Voters who see their social identities as secondary may start to feel estranged. They will be pulled within the identity political arms race even if they don’t want it. The likely result is a polarization of the political landscape organizing a relentless competition between social identities with significant adverse effects.

This is indeed one last feature of the analogy that is relevant. Positional goods tend to offset the mutually advantageous effects of market competition. They transform market relationships into a zero-sum game that deflects a significant amount of resources toward wasteful activities. Identity politics transforms what should initially be conceived as a cooperative (even if also competitive) venture to promote everyone’s interests and freedom into a conflictual battle for recognition. This battle turns politics into a wasteful and damaging war. The world would surely be a better place without positional goods and identity politics. It’s difficult to get rid of the former but there are ways to lessen the social costs they generate (e.g., a progressive consumption tax). There is unfortunately no clear perspective to escape the latter in the current political context.  

---

[1] Thorstein Veblen, The Theory of the Leisure Class, 1899.

[2] Among contemporary economists, Robert Frank has provided the most systematic study of the social costs produced by the consumption of positional goods. See Robert H. Frank, The Darwin Economy: Liberty, Competition, and the Common Good (Princeton N.J.: Princeton University Press, 2011).

[3] See for instance the entry of the SEP, Cressida Heyes, “Identity Politics,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta and Uri Nodelman, Winter 2024 (Metaphysics Research Lab, Stanford University, 2024), https://plato.stanford.edu/archives/win2024/entries/identity-politics/.

[4] For instance, the way people vote can be statistically accounted for by social characteristics, such as income, education, or gender.