photo of a neon sign with gas prices

Wally Skalij/Los Angeles Times/Getty Images

Los Angeles, October 7, 2022

How should the United States be dealing with inflation? For many decades, economists have given a straightforward answer: the Federal Reserve Board should increase interest rates over time, and it should tell people, loudly and clearly, that it is going to do exactly that. As interest rates rise, individuals and businesses will have to pay more to borrow money (including for the most expensive items, such as housing), which means that they will have less money to spend, which means that the demand for goods and services will go down, which means that prices will stop rising. If people know that interest rates will rise, they can feel confident that inflation will eventually come under control—which will help counteract the damaging psychology of inflation, by which people buy more right now and workers demand immediate wage increases, with the expectation that prices are going to rise.

For the past few months, the Fed and its chairman, Jerome Powell, have been following the usual economic prescription. But some people deplore this approach. Writing in The Wall Street Journal in July, for example, Senator Elizabeth Warren pointed to the risk that raising interest rates would create a catastrophic recession, with employers firing people, cutting hours, and slowing hiring, thus leaving families with a lot less money. She does not celebrate what she calls “the bloodless language of economists,” which emphasizes the importance of “dampening demand.” For Warren, a huge spike in unemployment would be an unacceptable consequence of higher interest rates. She favors what she sees as less damaging tools to combat inflation, including fighting corporate monopolies, which may take advantage of inflationary periods to raise prices; increasing supply by reducing backlogs at American ports; and investing in child care so that more parents can work.

Many people agree with Warren that economic analysis can be bloodless—cold, unfeeling, and unacceptably technocratic. They might be right. But before accepting the accusation, you might consider the fact that economics can be seen as applied utilitarianism, which the philosopher and economist John Stuart Mill defended in the following way:

In the golden rule of Jesus of Nazareth, we read the complete spirit of the ethics of utility. To do as one would be done by, and to love one’s neighbour as oneself, constitute the ideal perfection of utilitarian morality.

That’s not so bloodless.

In Mill’s account, utilitarianism rests on a firm moral foundation: Everyone’s well-being counts equally. To know what to do, we need to ask about the human consequences of different courses of action as rigorously as possible, putting our dogmas, our hopes, our fears, and our intuitions to one side. Following Mill, contemporary economists tend to describe themselves as utilitarians, or “welfarists.” Insofar as they make policy or advise policymakers, they generate a lot of numbers. Then they try to follow the golden rule, counting everyone equally and broadly assessing the potential impact of various policies on human welfare. Whatever we think is the right approach to inflation, that’s a helpful way to assess many different policy proposals.

Imagine, for example, that an economist in the Department of Transportation is asked to evaluate a regulation designed to require motor vehicles to be more fuel-efficient (and to move the market in the direction of electric vehicles). The economist will want to project the likely cost to the industry. Let us suppose that the best estimate is $2 billion per year, most of which will be passed on to consumers in the form of higher car prices. The economist will also seek to project the benefits of the regulation, which will include consumer savings from driving more fuel-efficient cars and reductions in air pollution (including greenhouse gas emissions). Let us suppose that the total consumer savings are $5 billion per year and the benefits from reduced air pollution $2 billion.1 Since $5 billion is higher than $2 billion, the economist will be strongly inclined to conclude that the fuel-economy regulation is a good idea.

Economists do this sort of calculation in evaluating policy proposals involving occupational safety, smoking reduction, food safety, free trade agreements, health care reform, and much more. Many lifesaving policies have been spurred by a cost-benefit analysis of this kind, and many ill-conceived plans have been stopped on the grounds that they would do far more harm than good. Some promising proposals have been reoriented after a demonstration that a modified approach could produce higher benefits, impose lower costs, or both. Careful economic analysis, focused on the best available evidence about the likely human consequences, can serve as a corrective to interest group lobbying, ideology, and so-called motivated reasoning, in which people end up thinking that something will be good (or bad) because that’s what they want to believe.


Skeptics have often raised two sets of objections to economic analysis of this kind. The first is epistemic: How can economists possibly know enough to estimate and monetize the effects of policies? The costs of fuel-economy regulations might turn out to be a lot lower than anticipated, and the speed of technological innovation, which regulation might itself spur, could be a lot faster. Free trade agreements might have a host of negative consequences that economists cannot readily foresee. If such agreements cause far more job losses than expected, they might have devastating effects on communities and families. For some problems, the economist’s before-the-fact calculation can turn out to be way off.

The second set of objections is ethical: What about the people who are hurt? Suppose that the construction of a new power plant delivers benefits in excess of costs but that it also increases air pollution in poor communities, producing spikes in sicknesses and deaths. In the US, dirty air has been found to have a disproportionate effect on people of color, with unsafe levels of pollution in (for example) Houston and Los Angeles. In Flint, Michigan, dirty water, producing health risks, came in part from an effort to cut costs.

It would be scant comfort to those who get ill, or die prematurely, that a project saved money or produced lots of jobs. Policies that do a lot of good can also do real damage. With that point in mind, skeptics of the standard economic analysis often focus on distributional issues: Who exactly is being helped, and who exactly is being harmed? For inflation, Warren has that question very much in mind.

In The Tragic Science, George DeMartino, a professor of economics at the University of Denver, offers both epistemic and ethical objections to economic analysis, especially insofar as it influences public policy. He insists that economists cause damage that “can be severe, even devastating,” including the destruction of lives. That, he says, is “the tragedy of economics.” His central target is the “moral geometry” that, in his view, economists practice, reducing hard moral issues to math problems, tallying up all effects and turning them into monetized costs to balance neatly against benefits. In DeMartino’s account, this “represses the complexity” of suffering and loss and insulates the economic “profession morally from the harms people suffer because of the policies it advocates.” The result is that “economists sleep well while others suffer the consequences of their practice.”

DeMartino puts a bright spotlight on what he describes as the “irreparable ignorance” of economists, who often have to deal with what is unknown and unknowable. Consider the financial crisis of 2008: even the most experienced economists did not foresee what was to come (the total collapse of large companies, billions upon billions of dollars in losses, an enormous increase in unemployment). Or consider the inflationary period in which we now find ourselves: many of the nation’s best economists wrongly thought that any spike in prices would be modest and transitory—“Team Transitory,” as it has been called, isn’t looking so great.

DeMartino emphasizes that the real-world effects of policy interventions depend on a large number of individual responses, which cannot easily be anticipated in advance. (Puzzlingly, he does not engage with behavioral economics, which explores how people depart from perfect rationality in general and in their response to new policies; we might, for example, pay too little attention to a pandemic risk until it is far too late.)

To illustrate the epistemic problem, DeMartino points to the distinction, drawn by Frank Knight and John Maynard Keynes, between “risk” and “uncertainty.” For example, consider the suggestion from some economists that there is a greater than 50 percent chance of a recession in 2023: a recession is a risk, and we might be able to assign a number to the likelihood that it will occur. But in circumstances of uncertainty, we do not know enough to assign probabilities at all; consider the likelihood of a war in Europe in 2100. As Keynes once put it, on some topics “there is no scientific basis on which to form any calculable probability whatever. We simply do not know.” DeMartino thinks that economists often find themselves in that situation but pretend otherwise.

He uses the term “econogenic harm” to refer to damage that economists cause but do not see or properly count. To make that idea concrete, he explores the concept of Kaldor-Hicks efficiency, developed independently by Nicholas Kaldor and John Hicks in the 1930s and 1940s, which asks whether a policy confers gains on the winners that exceed the losses to the losers. The Kaldor-Hicks test is sometimes justified on the ground that there would be a surplus and that the winners could, in principle, compensate the losers—as would be the case, for example, if a fuel-economy regulation produced $8 billion in monetized annual benefits and $6 billion in monetized annual costs. DeMartino objects, “But here’s the trick. Kaldor-Hicks does not require that compensation be made.” He deplores the “brazen attitude of economists who advocate policy that harms some members of society on grounds that it benefits others more.”


DeMartino also argues that economists pay too little attention to the difference between harming people and merely neglecting to help them. If you are inflicting damage on someone, you need special justification; that is not the case if you decline to give something to that person. He also worries about the problem of “commensurability.” There are qualitative differences among things that people value. A heart attack can be described as a cost, but it is not the same as the loss of a specified sum of money. The loss of jobs is different from the destruction of a beach, and the loss of an endangered species is not the same as a reduction in corporate profits.

DeMartino thinks that welfarism, by reducing all harms to a single metric, “trivializes what it means to be harmed while ignoring the diverse forms of harm and ways people can be harmed.” In his account, some kinds of injury are not reparable at all, and some goods are priceless; monetary compensation does not suffice when someone has (for example) lost a child.

To address these problems, DeMartino endorses a proposed alternative to standard economic policy analysis: decision-making under deep uncertainty (DMDU). As he describes it, DMDU practitioners are aware that predictions can be perilous and that the past may not be prologue. Some policy problems are “wicked,” in the sense that they involve “nonlinear dynamics and unpredictable breaks (tipping points) in trends that are not identifiable in advance”—climate change and pandemics are plausible examples. To deal with these problems, DeMartino urges us to seek approaches that increase the likelihood of tolerable outcomes under most imaginable futures. We should, for example, take strong steps, right now, to increase resilience to risks associated with flooding, drought, wildfire, and extreme heat.

DeMartino does not claim to have a simple replacement for the moral geometry that he deplores, but he thinks that we can do a lot better than economists are now doing. He wants to focus on human “interests,” which are not the same as welfare. Watching a good romantic comedy might be fun; being a good parent, or having a fulfilling career, is altogether different. DeMartino is drawn to the “capabilities approach,” pioneered by Amartya Sen and Martha Nussbaum, which specifies a disparate set of human capabilities, helping people to achieve various “functionings.” In Sen’s account, these include having enough food, access to shelter, and preventing death. Nussbaum has also focused on health and housing, and includes “being able to laugh, to play, to enjoy recreational activities.”

DeMartino argues that Sen’s and Nussbaum’s capabilities approach can help us to address policy questions while acknowledging that economic growth is not enough, focusing on the distribution of gains and losses, asking whether those who face harm could take steps to avoid it, and distinguishing between reparable and irreparable harms.

To come to terms with what DeMartino calls his “indictment” of economics, it makes sense to start with the epistemic problem—a problem to which economists themselves have given a great deal of attention. He does not deny that in many cases economists know enough to make reasonably specific projections about the consequences of proposed policies. In other cases, they may at least be able to identify a limited range of possible outcomes. It is true that for some important problems, projected effects are far more speculative: If the US and China don’t take aggressive steps in the coming years to reduce greenhouse gas emissions, what is the likely average global temperature in 2100? What would be the long-term benefits to the US, China, and the world if the US and China do take those steps?

Economists are acutely aware of the epistemic problem, and they have a host of strategies to deal with it. First, they favor “no regrets” actions, which are worth undertaking even if the relevant risks do not come to fruition. They might favor energy-efficiency requirements because they want to help consumers save money, whether or not such requirements will do much to combat climate change. In the domain of regulation, many economists are drawn to “break-even analysis,” which asks: What would the benefits of this proposal have to be in order to justify the costs? We might not be able to specify all of what we will gain from a new technology designed to make cars safer, but we might know that, even if the new technology helps much less than we hope, it will still be worthwhile. Some economists embrace the “precautionary principle,” which attempts to create margins of safety against risks we do not understand very well.

Within economics, there is an extensive and growing literature on how to proceed in the face of uncertainty, not least in the domain of climate change.2 Economists have explored the “maximin” strategy, which calls for the elimination of the worst-case scenario. With respect to global warming, we might not know nearly enough to be able to assign probabilities to the most horrific outcomes, so we might choose to take aggressive steps to eliminate those outcomes, or at least to reduce the risk that they will occur, both by scaling back emissions and by increasing resilience against flooding, extreme heat, drought, and wildfire. In light of the literature on these questions, DeMartino is, I think, far too hard on the economics profession. He would have done well to explore the various approaches, carefully developed by economists themselves, to deal with gaps in knowledge.

In lamenting the “tragedy of economics,” DeMartino is following in the footsteps of Mill, who criticized Jeremy Bentham for overlooking

the sense of honour, and personal dignity—that feeling of personal exaltation and degradation which acts independently of other people’s opinion, or even in defiance of it; the love of beauty, the passion of the artist; the love of order, of congruity, of consistency in all things, and conformity to their end; the love of power, not in the limited form of power over other human beings, but abstract power, the power of making our volitions effectual; the love of action, the thirst for movement and activity, a principle scarcely of less influence in human life than its opposite, the love of ease…. Man, that most complex being, is a very simple one in his eyes.

Fair enough. But does all this really count as an objection to standard economic analysis? Suppose that you love your dog. You might even describe your dog as priceless. Still, you make trade-offs; you will probably not be willing to spend all your money on treatment if your dog gets cancer. Every day, people make choices among goods that they value in different ways and for different reasons. In assessing costs and benefits, economists are not making controversial or obtuse philosophical claims about the nature of the good. They are engaged in a pragmatic exercise meant to recognize trade-offs and to see how we can get as much as possible of what we care most about.

To be sure, we need to ask about the distribution of costs and benefits, and not just their magnitude. If you are poor, cash in your hands is far more important to you than it would be if you were rich. There is a strong argument that policymakers should focus on helping those at the bottom of the economic ladder; the recent emphasis on “prioritarianism,” coming from both philosophers and economists, is a clear recognition of that argument.3 In drawing attention to distributional questions, DeMartino is hardly offering an external critique of economics; he is joining a long tradition within the field. Here as elsewhere, economists would be right to respond to DeMartino’s indictment by pleading not guilty.

Still, DeMartino is entirely correct to insist, with Senator Warren, on the need to attend to those who are struggling the most—a point that has broad implications for public policy. Even if the Fed should be raising interest rates (and I think it should be, although we might disagree about how much and how fast), Warren is convincing when she draws attention to the human suffering that might result—and to the need to develop policies to reduce or mitigate that suffering.

It is much less clear whether we should accept DeMartino’s claim about the importance of distinguishing between harming people and failing to help them. To make that distinction, we need some kind of baseline. If the Environmental Protection Agency refuses to issue an air pollution regulation, is it harming people or refusing to help them? If the Occupational Safety and Health Administration does not take steps to combat Covid at the height of the pandemic, is it just failing workers or is it harming them? Government inaction has consequences, and so does government action. You can harm people by failing to help them, and they might have a right to your help. For purposes of deciding whether to act and what to do, it is usually best to catalog all the consequences and choose the approach that has the best consequences on balance.

It is urgent that we develop new and better strategies to cope with uncertainty. Whether we are thinking about inflation, climate change, or occupational health, it is also urgent to ask who is helped and who is hurt, and to protect the interests of the most vulnerable. But in indicting welfare economics as such, DeMartino loses sight of its moral foundation: “To do as one would be done by, and to love one’s neighbour as oneself.” If our goal is to avoid tragedy, that is a good place to start.