Economic policy is not a purely technical exercise. Every policy recommendation embeds normative choices — about whose welfare counts, how much weight to give to future generations relative to present ones, how to make trade-offs between efficiency and equity, and what degree of risk to impose on vulnerable populations in the pursuit of aggregate gains. Economists often present policy recommendations in the language of optimization and efficiency, obscuring these normative choices. This appendix makes them visible.
I.1 Welfare Economics: Foundations and Limits¶
The standard welfare criterion in macroeconomics is the utilitarian social welfare function: social welfare is the (possibly weighted) sum of individual utilities:
where is the welfare weight attached to individual . Most macroeconomic models use the representative-agent assumption, which implicitly sets and treats every household as identical. This suppresses distributional questions entirely: a policy that doubles the income of a billionaire while reducing a subsistence farmer’s income by 10% may increase “aggregate welfare” under this criterion.
The Kaldor–Hicks criterion weakens utilitarianism: a policy change is a potential Pareto improvement if the winners could in principle compensate the losers and still be better off. Many policy recommendations in macroeconomics rest implicitly on Kaldor–Hicks: trade liberalization, for instance, is argued to increase aggregate income even though it creates concentrated losers (displaced workers). The criterion is ethically problematic because the compensation is rarely actually paid. Policies that create aggregate gains but concentrated losses require a separate distributional judgment — and an ethical argument for why the gains outweigh the losses — beyond the Kaldor–Hicks calculation.
Rawlsian maximin offers an alternative: maximize the welfare of the worst-off member of society. Under this criterion, the social planner would never accept any policy that reduced the income or welfare of the poorest household, even if it greatly increased average income. Few economic policies satisfy this demanding criterion, but it provides a useful counterpoint that forces attention to the distribution of policy costs and benefits.
I.2 The Ethics of Discounting Future Generations¶
The discount rate used in long-run cost-benefit analysis — including climate change economics — is an ethical choice, not merely a technical one. The Ramsey rule for the optimal social discount rate is:
where is the pure rate of time preference (how much less we weight future welfare relative to present welfare purely because it is in the future), is the coefficient of relative risk aversion (capturing diminishing marginal utility of income as future generations are richer), and is the growth rate of per-capita consumption. With , , and , — meaning a dollar of harm to people living in 100 years is worth only dollars today. This dramatically reduces the present value of climate damages.
The choice of is deeply normative. A positive says that the welfare of people in the future matters less than the welfare of people today simply because they are more distant in time — a form of temporal discrimination analogous to geographic or racial discrimination. Many philosophers argue that is the only defensible position: future generations have the same claim on welfare consideration as present ones. This was the Stern Review’s position and explains its radically higher estimates of the social cost of carbon. Nordhaus’s use of reflects a different philosophical position: that we should discount future welfare at the market rate of return, incorporating market participants’ implicit preferences about time.
There is no purely economic resolution to this debate. It is a question about the moral status of future people, and it deserves explicit acknowledgment rather than burial in a technical parameter.
I.3 Distributional Consequences and Policy Incidence¶
Every macroeconomic policy has distributional consequences — it affects different groups differently. The analysis of who bears the costs and who receives the benefits of a policy is called incidence analysis. Macroeconomic textbooks often present policies in terms of their aggregate effects (the output multiplier, the effect on the price level) without examining incidence. This omission is analytically incomplete and ethically problematic.
Monetary policy distributes income and wealth across debtors and creditors, across workers and capital owners, and across holders of different financial assets. An interest rate cut reduces the income of savers but reduces the debt burden of borrowers. Quantitative easing raises asset prices, benefiting primarily wealthy households who hold most financial assets. A rate hike that reduces inflation benefits fixed-income holders and creditors but raises unemployment, imposing costs on the workers most likely to lose their jobs in a downturn — who are disproportionately low-income.
Fiscal policy incidence depends on the composition of spending and taxation. A consumption tax is regressive — poor households spend a larger share of income on consumption and bear a larger share of the burden. An income tax with progressive rates is progressive. Government spending on education, healthcare, and social insurance tends to be progressive (the poor benefit more per dollar of income); spending on higher education subsidies, mortgage interest deductions, and defense procurement may be regressive.
Trade policy creates winners and losers along industry, geographic, and skill lines. The Stolper–Samuelson theorem predicts that trade liberalization in a capital-abundant country (like the United States) benefits capital owners and harms low-skilled workers — because it increases the relative return to the abundant factor. The concentrated geographic losses (factory towns, rust belt communities) from deindustrialization are not captured in aggregate welfare calculations.
I.4 The Ethics of Macroeconomic Risk¶
Macroeconomic policies affect not just average outcomes but the distribution of risk. Policies that reduce average unemployment by making it more volatile — “hire and fire” labor market flexibility — transfer risk from firms to workers. Fiscal austerity in a recession reduces government deficits but increases the unemployment risk borne by workers, particularly in the public sector. Financial deregulation may increase average credit availability but also increases the frequency and severity of financial crises, which impose severe costs on vulnerable populations.
The justice of risk allocation is a legitimate ethical concern that pure expected utility analysis misses. Even if a policy raises expected welfare, it may be unjust if it imposes large risks on people who are poorly equipped to bear them — because they lack savings, insurance, or political power to protect themselves. A complete policy evaluation should include: (i) the change in expected welfare; (ii) the change in the distribution of welfare outcomes; and (iii) who bears the risk of adverse outcomes.
I.5 Macroeconomics, Power, and Political Economy¶
Economic policies are made in political contexts in which some groups have more power than others. Macroeconomic theory is not immune to the influence of these power relationships on what questions get asked, what models get built, and whose interests policy recommendations serve. Several critical observations are worth making explicitly.
The neutrality illusion. The presentation of economic policy recommendations as technically derived from neutral models that maximize aggregate welfare is often misleading. The models embed distributional choices (welfare weights), ethical choices (discount rates), and empirical assumptions (about multiplier sizes, price flexibility, and behavioral responses) that are not neutral. Different assumptions yield different policy prescriptions, and the choice among them often reflects whose interests the analyst is implicitly prioritizing.
The measurement problem. Because GDP measures market transactions and not welfare, policies that increase GDP while reducing non-market welfare (environmental quality, leisure time, domestic production) will be endorsed by GDP-maximizing approaches even if they reduce true welfare. The political incentive to present GDP growth as the measure of policy success can distort policy toward outcomes that are measured well rather than outcomes that matter most.
The creditor-debtor tension. The macroeconomic policy debates of the post-2008 period — about the pace of fiscal consolidation, the appropriate stance of monetary policy, and the treatment of sovereign debt — often reflected a conflict between the interests of creditors (who benefit from deflation, fiscal surplus, and high real interest rates) and debtors (who benefit from inflation, fiscal expansion, and low real interest rates). The presentation of “sound macroeconomic policy” often reflects one side of this conflict while claiming to represent a neutral technical judgment.
None of this implies that macroeconomics is merely ideology dressed up in mathematics. Positive macroeconomics can produce genuine knowledge about how economies work, even if normative macroeconomics requires ethical choices. The appropriate response to the power of vested interests in shaping economic analysis is transparency — making the ethical choices explicit, acknowledging where the evidence is contested, and subjecting policy recommendations to the same scrutiny as empirical claims.
I.6 The Role of the Economist as Expert and as Citizen¶
Economists occupy an unusual social position: they are experts whose technical judgments have major political implications. This creates a tension between the role of technical expert (providing neutral analysis) and the role of citizen (who has interests and values). Several principles can guide this tension:
Intellectual honesty requires acknowledging when the evidence is uncertain, when alternative models give different conclusions, and when a recommendation depends on value judgments that reasonable people might make differently.
Separation of analysis and advocacy suggests being clear about where positive analysis ends and normative advocacy begins. “According to our model, this policy increases GDP by X%” is a positive claim; “therefore the government should adopt this policy” is a normative claim that requires additional ethical premises.
Epistemic humility is warranted by the track record. Economics has a long history of overconfident predictions (the efficient markets hypothesis before 2008, the “Great Moderation” thesis, the claimed precision of structural estimates) that were subsequently revised. The appropriate posture is to present the best available analysis while being honest about its limitations.
Responsibility for consequences is harder to evade when economic recommendations have been adopted as policy. Economists who advised structural adjustment programs that caused severe social harm, or financial deregulations that contributed to crises, bear some responsibility for those outcomes even if they acted in good faith on the best available analysis at the time. This responsibility reinforces the case for epistemic humility and for broadening the range of considerations — including distributional impacts, political economy risks, and second-best constraints — that enter policy analysis.
Further reading on economics and ethics: Sen, A. (1987). On Ethics and Economics. Oxford: Blackwell. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Sandel, M. (2012). What Money Can’t Buy: The Moral Limits of Markets. New York: Farrar, Straus and Giroux.