Philosophers Annual is a highly prestigious and 45-year-running anthology of the 10 best philosophy papers published each year. The collection for 2025 was recently released and #1 on the list (though I’m not sure there’s any particular order) is Desert and Economic Interdependence, by Evan Behrle.
In the paper, Behrle is disputing a particular justification for economic inequality: the idea that workers deserve to be paid according to their productive contributions.
This argument is popular and intuitive. One worker can justifiably make more than another by taking longer hours, offering higher skills, or enduring worse conditions. Billionaires are widely reviled, but income inequality within the working class, say between lawyers and laborers, is less scandalous even among the left. They are simply being paid in proportion to what they contribute.
By this argument, income inequality is justified because it is rewarding each worker in proportion to the productive contributions they make to the rest of the economy. Conveniently, markets enforce this moral principle through competition. If a man or machine adds $100 to the revenue of a widget factory, competing employers can profitably outbid each other until the input is paid a wage equal to the value of the extra goods it produces, i.e. its marginal product.
Despite its popularity and intuitive appeal, Behrle argues that this justification for income inequality is wrong. Admirably, he does not take the standard tack of denying that market wages measure productive contribution, nor does he remind us that some determinants of productive capacity are set unfairly by nature or circumstance. Instead, Behrle seeks to defeat this argument on its own terms.
Behrle’s Argument
Behrle begins by considering the argument for paying workers their marginal product in the first place:
A worker’s marginal product is the difference between output with the worker and output without them … Marginal revenue product has us compare the actual world, at a given time, with a possible world in which an individual worker is absent.
The justification for crediting workers with their marginal product is based on counterfactual impact. Were it not for the labor of Worker W, total output would be lower by $200k. Therefore, Worker W deserves $200k of total output.
But how does W produce that $200k exactly? Alone on a desert island W could only produce a few dollars worth of coconuts at best. In a modern economy there are dozens of groups without whom W could not produce a cent. To make their marginal product W needs coworkers, managers, truckers, electricians, plumbers, pharmacists, and police. Formally, there exists a Group G such that, were it not for the labor of that group, Worker W’s marginal product would be $0 instead of $200k.
Therefore, by the same counterfactual impact argument that justifies W’s claim on $200k of output above, Group G has a claim to W’s marginal product. Were it not for the labor of Group G, W’s marginal product would be lower by $200k. In Behrle’s words “G makes a difference to the difference W makes.”
Note also that Group G’s claim to W’s marginal product goes beyond mere causation. One might say that W’s marginal product also relies on the demand of consumers for the goods they create, or on the flap of a butterfly’s wings ten thousand years ago which set the world on its current course. But, far more saliently than these other causes, Group G causes W’s marginal product in the same way that W does, namely by contributing their labor.
If W deserves credit because their labor is necessary for this $200k difference, why should other workers receive no credit when their labor is likewise necessary for that very difference?
And, Behrle proactively responds, it is the case that these essential groups receive little to no credit for the difference they make. Take G to be truckers, for example. Together, they produce a service that’s essential for the whole economy and certainly for W’s $200k marginal product. But because there are already millions of truckers, adding or removing one doesn’t change any other worker’s marginal product or output much, so each one commands only a small wage. The sum of these wages is much less than the total value that would be lost if all trucking disappeared.
To rectify the difference between what G deserves (on the same basis as W) and what G gets (the sum of each member’s marginal product), W’s income must be split and shared with group G.
There are dozens of such groups in the economy. Thus, Behrle concludes, the distribution of desert is far more equal than the distribution of income created by paying workers their marginal product.
Response
There are several ways to respond to Behrle’s argument here. You might point out that even if marginal product payments don’t assign everyone exactly what they deserve, they do induce efficient behavior and so maximize the amount of value produced in the economy. Growing the size of the pie has been a far more effective lever for getting people what they deserve than fighting over how exactly to distribute the slices, so we’d do better to not kill the golden goose.
Alternatively, you might take a Nozick-style property rights view and just observe that at every step in the sequence of transactions that led to the income distribution that Behrle impugns, everyone simply made voluntary trades of their own labor and resources they legitimately own. No other justification for the income distribution is needed.
But neither of these responses engage directly with Behrle’s arguments. Must we accept his takedown of the counterfactual impact justification for marginal cost pricing and accept merely instrumental or axiomatic defenses of income inequality?
No. First, note that all of Behrle’s discussion of interdependence is merely an unusual way to say something that all economists already agree with. Namely, that there are lots of groups of inputs that are essential for economic output i.e. strong complementarity is widespread throughout the economy. The most common model of a production function, Cobb-Douglas production, combines inputs multiplicatively so that if any single input is zero, all of output is zero.
Marginal product already rewards inputs for this complementarity. In a simple Cobb-Douglas economy where output Y = K^α*L^(1-α), labor’s marginal product (∂Y/∂L = ((1-α)K^α*L^(-α)) rises with the capital stock, reflecting that labor makes a difference to the difference capital makes. If an input G increases W’s marginal product by $200k, then that’s added in to G’s marginal product as part of their counterfactual impact of joining the economy.
The problem Behrle has is not that any individual is being under-credited for the differences they make to output and to other's productivity. His problem is that groups of individuals are being under-credited for the differences they make only as a group.
He observes that the total value of truckers, say, is immense, but the marginal value of truckers, which sets their wages, is small, and sees this as an injustice. The injustice is not that any individual trucker makes a big impact to the economy that their wage doesn’t reflect. The injustice is that none of the trucker’s wages, nor their sum, reflect the massive economic impact of all truckers stopping work at once.
Behrle’s dispute with marginal product pricing, therefore, is just a reframing of the diamond-water paradox. Collectively, water is responsible for all life on earth and all economic output. But the price of a gallon of water may be thousands of times less than a diamond, all of which could disappear without catastrophic consequences. Why should diamonds deserve so much when their value depends on the continued existence of water?
The answer to this paradox, of course, is that scarcity is a source of value. Water would be far more valuable than diamonds if there was very little of it and every drop made the difference between life and death. But as it is, water is abundant, so any given gallon that might save your life is easily replaced with another that does the same.
Behrle’s view that desert should reflect the total value of a group of inputs rather than the marginal value of each one requires that somehow adding more workers or more gallons of water to the economy shouldn’t lower their value but should raise it because the group’s total impact goes up. It requires that every plate of food should be credited for the bite that saves your life in a famine. Or that every nurse should share the reward for the only one on staff at midnight on Christmas.
In fact, the value of something, its counterfactual impact, falls as the number of alternatives and substitutes grows. The workers of group G aren’t compensated for their collective role in W’s marginal product because they are each replaceable where W is not. Perhaps all together they are not replaceable, but they can all replace each other and that decreases the value of each one. If they band together to extract their total surplus, that’s called a monopoly or a cartel. I wonder whether Behrle would like all of the world’s oil companies to collect more credit for their role in powering the world economy.
The workers of G don’t deserve any of W’s marginal product because each of them, on the margin, is unessential for it. That W would suffer if G and all its substitutes disappeared together shouldn’t credit the members of G any more than the fact that food is worth everything in a famine should make the prices at the grocery store go up.
Truckers as a group may be necessary for the functioning of the modern economy, but rewarding any individual trucker for this interdependence is giving them credit for the value of truckers in a world where we have very few. In fact, in our world, we have millions of truckers, so the counterfactual impact of each one is low. Thus, the low marginal product of truckers is not a denial of their necessity as a group or an under-payment, it’s just a consequence of the fact that scarcity is a source of value.
There are other problems with Behrle’s claims and further proofs by contradiction one can point to, but this is the central response: The counterfactual impact of an entire group is not a source of desert for its members because the substitutability of the members within that group is a legitimate constraint on their value.