No CrossRef data available.
Article contents
A LIMITED DEFENSE OF EFFICIENCY AGAINST CHARGES OF INCOHERENCY AND BIAS
Published online by Cambridge University Press: 14 August 2023
Abstract
Scholars have long debated the appropriate balance between efficiency and redistribution. But recently, a wave of critics has argued not only that efficiency is less important, but that efficiency analysis itself is fundamentally flawed. Some say that efficiency is incoherent because there is no neutral baseline from which to judge inefficiency. Others say that efficiency is biased toward those best able to pay (generally, the rich). This essay contends that efficiency is not meaningfully incoherent or biased. The most widely discussed forms of efficiency do not require any particular baseline, and even those that do require a baseline can still serve as useful approximations of more theoretically sound but computationally demanding measures. Moreover, arguments of bias do not account for the source of funds in public projects, produce unintuitive results, and draw an arbitrary cutoff between bias and non-bias that elides important distributional details. Ultimately, the tradeoff between efficiency and redistribution remains the most useful frame for policy debate.
- Type
- Research Article
- Information
- Copyright
- © 2023 Social Philosophy & Policy Foundation. Printed in the USA
Footnotes
*
Law School, University of Minnesota, jonchoi@umn.edu. Competing Interests: The author declares none. Thanks to Neil Buchanan, Erick Sam, David Schmidtz, the editors of the journal Social Policy and Philosophy, and the other contributors to this volume for helpful comments.
References
1 Hypothetical compensation may not be paid for a variety of reasons, including transaction costs and political difficulties.
2 The Kaldor-Hicks criterion has some well-known technical oddities, including the possibility of “reversals” illustrated by the Scitovsky paradox, that two different states could each be Kaldor-Hicks superior to each other. Tibor Scitovsky, Papers on Welfare and Growth (London: Routledge, 2003). The paradox results from differences between willingness-to-accept and willingness-to-pay. There is good reason, drawing on the behavioral economics literature, to believe that one might require greater compensation to give something up than one would have been willing to pay for that thing in the first place (“loss aversion”). Kahneman, Daniel, Knetsch, Jack L., and Thaler, Richard H., “Anomalies: The Endowment Effect, Loss Aversion, and Status Quo Bias,” Journal of Economic Perspectives 5, no. 1 (1991): 193–206 CrossRefGoogle Scholar. Duncan Kennedy, Duncan Kennedy, “Cost-Benefit Analysis of Entitlement Problems: A Critique,” Stanford Law Review 33, no. 3 (1981): 387–445 CrossRefGoogle Scholar.
Scitovsky (and later, Samuelson, along slightly different lines) proposed amending the definition of Kaldor-Hicks superiority to exclude the possibility of reversals. Allan M. Feldman, “Kaldor-Hicks Compensation,” in Peter Newman, ed., The New Palgrave Dictionary of Economics and the Law, Volume 2 (London: Macmillan, 1998), 417-21. Making this amendment solves the problem but increases the likelihood that, given two possible states of the world, neither will be Kaldor-Hicks superior to the other, decreasing the usefulness of Kaldor-Hicks analysis.
3 This could happen if two different policies would result in different sets of individuals. For example, given current state-of-the-world s and potential state of the world s’ that would result from a certain policy, if the policy would incentivize parents to have additional children, then s and s’ are no longer comparable, because s’ includes individuals (the additional children) who cannot be compensated (either actually or hypothetically) for a move to s. I hope to discuss this limitation to Pareto and Kaldor-Hicks in future work.
4 Both willingness to pay and willingness to accept can be positive or negative, and each reflects a discrete and competing approach to wealth maximization. Willingness to pay is a measure of an individual’s “compensating variation,” meaning the amount of money that would leave her welfare unaltered if the change does occur, as measured using prices in the world where the change occurs. Willingness to accept is a measure of an individual’s “equivalent variation,” meaning the amount of money that would render her as well off under the status quo if the change does not occur, as measured using prices under the status quo. For a variety of reasons, including differences in price levels between the status quo and post-change world, willingness to pay and willingness to accept are different and sometimes competing measures. Kornhauser, Lewis, “Wealth Maximization,” in The New Palgrave Dictionary of Economics and the Law, ed. Newman, Peter (London; New York: Macmillan, 1998), 679 Google Scholar.
5 For example, Richard Posner has sometimes referred to wealth maximization and Kaldor-Hicks interchangeably. Posner, Richard A., “The Ethical and Political Basis of the Efficiency Norm in Common Law Adjudication,” Hofstra Law Review 8 (1980): 491 Google Scholar, 495.
6 Charles Blackorby and David Donaldson have highlighted several respects in which wealth maximization is different from, and in their view worse than, Kaldor-Hicks optimization. Increases in wealth are possible even when no Kaldor-Hicks improvement has occurred, so that an increase in wealth is a necessary but not sufficient condition for a Kaldor-Hicks improvement. Wealth maximization and Kaldor-Hicks will yield the same result if and only if price levels are the same in all scenarios under consideration, which is typically not the case. Blackorby, Charles and Donaldson, David, “A Review Article: The Case Against the Use of the Sum of Compensating Variations in Cost-Benefit Analysis,” Canadian Journal of Economics 23, no. 3 (1990): 471–94CrossRefGoogle Scholar.
7 Weisbach, David A., “Line Drawing, Doctrine, and Efficiency in the Tax Law,” Cornell Law Review 84 (1999): 1650 Google Scholar.
8 Okun, Arthur M., Equality and Efficiency: The Big Tradeoff (Washington, DC: Brookings Institution, 1975)Google Scholar.
9 Kaplow, Louis and Shavell, Steven, “Why the Legal System Is Less Efficient than the Income Tax in Redistributing Income,” Journal of Legal Studies 23, no. 2 (1994): 669 Google Scholar.
10 Christine Jolls argues from the perspective of behavioral economics that redistribution through legal rules may in fact cause less distortion than redistribution through taxation. Christine Jolls, “Behavioral Economic Analysis of Redistributive Legal Rules,” Vanderbilt Law Review 51, no. 6 (1998): 1656–67.
11 Chris Sanchirico argues that legal rules can facilitate efficient redistribution based on “ability,” for which labor income is a proxy. Conceptually, if one’s likelihood of incurring tortious liability were also an imperfect proxy for ability, it could be optimal to enact redistribution through legal rules in order to better redistribute on ability. Sanchirico, Chris William, “Taxes Versus Legal Rules as Instruments for Equity: A More Equitable View,” Journal of Legal Studies 29, no. 2 (2000): 802–3Google Scholar. Brian Galle argues that local tort rules can redistribute more effectively than local tax law, because local taxation incurs additional distortions that tort rules do not (primarily due to taxpayer mobility and choice of residence). Galle, Brian, “Is Local Consumer Protection Law a Better Retributive Mechanism than the Tax System,” NYU Annual Survey of American Law 65 (2010): 525 Google Scholar.
12 See, for example, Markovits, Richard S., “Why Kaplow and Shavell’s ‘Double-Distortion Argument’ Articles Are Wrong,” George Mason Law Review 13, no. 3 (2005): 521 Google Scholar; Fennell, Lee Anne and McAdams, Richard H., “The Distributive Deficit in Law and Economics,” Minnesota Law Review 100, no. 2 (2016): 1051 Google Scholar. Zachary Liscow also argues that because “individuals silo their policy views,” we cannot practically offset less redistribution through legal rules with more redistribution through taxes and transfers. Liscow, Zachary, “Redistribution for Realists,” University of Iowa Law Review 107, no. 2 (2022): 495 Google Scholar.
13 See, for example, Liscow, Zachary, “Reducing Inequality on the Cheap: When Legal Rule Design Should Incorporate Equity as Well as Efficiency,” Yale Jaw Journal 123, no. 7 (2015): 2478 Google Scholar; Revesz, Richard L., “Regulation and Distribution,” New York University Law Review 93, no. 6 (2018): 1489–1578Google Scholar; Markovits, “Why Kaplow and Shavell’s ‘Double-Distortion Argument’ Articles Are Wrong”; Fennell and McAdams, “The Distributive Deficit in Law and Economics.”
14 “The rules of property law … are fundamental in determining what people can and cannot do in the marketplace.” Neil H. Buchanan, “The Role of Economics in Tax Scholarship,” in Brennen, David A., Brown, Karen B., and Jones, Darryl, eds., Beyond Economic Efficiency in United States Tax Law (New York: Wolters Kluwer Law and Business, 2013), 11 Google Scholar.
15 Other scholars have made analogous arguments, perhaps most famously Liam Murphy and Thomas Nagel. Liam B. Murphy and Thomas Nagel, The Myth of Ownership: Taxes and Justice (New York: Oxford University Press, 2002).
16 Buchanan, Neil H. and Dorf, Michael C., “A Tale of Two Formalisms: How Law and Economics Mirrors Originalism and Textualism,” Cornell Law Review 106 (2021): 20 Google Scholar. In an essay that Buchanan published several years earlier, he offered as another example the property law rule of adverse possession, which awards property rights to an adverse possessor who maintains “open and notorious” possession of property against the property’s nominal owner. Buchanan argued that there is no natural or proper baseline property right that we can assess efficiency against here, because the concept of adverse possession is itself a property right. Neil H. Buchanan, “The Role of Economics in Tax Scholarship,” 12-13.
17 Buchanan and Dorf, “A Tale of Two Formalisms,” 602.
18 Kaldor-Hicks has its own problems, which are well-known and have been discussed for decades. See note 2 above, discussing the Scitovsky paradox and some potential solutions to the paradox.
19 Kaplow and Shavell note that “wealth maximization is not a well-defined concept; to compute wealth, one must know the prices of different goods and services, yet there is no natural set of prices to use… . [T]he absence of a natural set of prices is not a problem that can be resolved by a simple price index adjustment, similar to adjustments for pure inflation, because relative prices differ.” Louis Kaplow and Steven Shavell, Fairness Versus Welfare (Cambridge, MA and London: Harvard University Press, 2002), 36.
20 Lewis A. Kornhauser, “Wealth Maximization,” 679-83.
21 Kennedy, “Cost-Benefit Analysis of Entitlement Problems: A Critique,” 422–44.
22 Posner, Richard A., “Wealth Maximization and Tort Law: A Philosophical Inquiry,” in Owen, David G., ed., Philosophical Foundations of Tort Law (Oxford; New York: Oxford University Press, 1995) 99–100 Google Scholar.
23 See, for example, Adler, Matthew D. and Posner, Eric A., New Foundations of Cost-Benefit Analysis (Cambridge, MA and London: Harvard University Press 2006), 6 CrossRefGoogle Scholar (“Our argument is that CBA is best defended as a welfarist decision procedure.”); Kaplow and Shavell, Fairness Versus Welfare, 37 (“[M]aximization of wealth (defined, perhaps, with respect to current prices) may in fact reasonably approximate maximization of social welfare in many contexts. Thus, under welfare economics, although wealth is not in itself deemed to be valuable, analysis that assesses policies based on their aggregate impact on wealth will often prove useful.”). See also ibid., 36, making an analogous point about efficiency in general.
24 Of course, this argument will be unpersuasive to theorists that favor wealth maximization, Kaldor-Hicks, or Pareto over welfarism because they eschew interpersonal or cardinal comparisons of utility. The argument instead targets those who theoretically subscribe to welfarism but find it too demanding in practice.
25 Liscow, Zachary, “Is Efficiency Biased?” University of Chicago Law Review 85, no. 7 (2018): 1656.Google Scholar
26 Baker, C. Edwin, “The Ideology of the Economic Analysis of Law,” Philosophy and Public Affairs 5 (1975): 3Google Scholar, 16-19, 47-48; Posner and Adler, New Foundations of Cost-Benefit Analysis¸16-19; Posner, Eric A. and Adler, Matthew D., “Rethinking Cost-Benefit Analysis,” Yale Law Journal 109 (1999): 184 Google Scholar.
27 Liscow, “Is Efficiency Biased?” 1655.
28 Ibid., 1655.
29 Ibid., 1655.
30 Ibid., 1655.
31 As Daniel Shaviro has noted: “The distinction between taxes and spending … depends on pure form.” Daniel Shaviro, “Rethinking Tax Expenditures and Fiscal Language,” Tax Law Review 57 (2004): 191.
32 Mankiw, N. Gregory, Weinzierl, Matthew, and Yagan, Danny, “Optimal Taxation in Theory and Practice,” Journal of Economic Perspectives 23, no. 4 (2009): 147–74CrossRefGoogle Scholar.
33 Liscow, “Is Efficiency Biased?” 1650–51.
34 Erica York, “Summary of the Latest Federal Income Tax Data, 2020 Update,” https://taxfoundation.org/summary-of-the-latest-federal-income-tax-data-2020-update. These are the most recently available data on tax collections stratified by income level; 2018 data will likely show somewhat less progressivity, due to 2017 tax reform.
35 The situation is more complicated for local parks funded by local property taxes. Many of the poorest taxpayers rent, rather than own, so do not directly pay any property tax at all. However, the incidence of the property tax may still partly fall on poor taxpayers, who would pay higher rent as a result. On the other hand, to the extent that property tax is capitalized into the value of real estate, and particularly to the extent that a Georgist tax on unimproved land theoretically taxes pure economic rent and therefore falls solely on capital-owners, the incidence on the poor may be minimal. Finally, property taxes are typically flat, and property ownership as a percentage of income or wealth likely declines the richer one gets, so that property taxes may ultimately be flat or regressive in the upper-income range.
36 Liscow, “Is Efficiency Biased?” 1662–66. Liscow describes circumstances under which offsets would be more or less likely, and emphasizes that if offsets occur then rich-biased policies may be justified; nevertheless, in contrast to this essay, he appears to focus on specific offsets rather than the potential that expenditures would be implicitly offset through the general budget process.
37 Because the federal government may freely borrow, it is possible that new spending will not be paid for until far into the distant future. But because money is fungible, it is generally not possible to distinguish which expenditures are paid for now versus later, and given the difficulty of predicting future tax rates, current tax rates are likely our best approximation of the taxing side for distributional analysis.
38 Payroll taxes are also regressive, but they are theoretically earmarked to fund specific government programs like unemployment insurance, Social Security, and Medicare, whose overall effect (taking into account both taxes and spending) is progressive. As a result, it seems more plausible that marginal expenditures are funded by income taxes rather than payroll taxes. This may not always be the case given that unemployment insurance, Social Security, and Medicare are pay-as-you-go programs with historically underfunded accounts to pay for future expenses. However, we are now reaching the inflection point where expenses for these programs begin to exceed revenues, making it less likely that general social programs can be funded from payroll taxes in the future.
39 This calculation uses the average tax rate for the rich and the poor; it is possible that using marginal tax rates could give different results.
40 He cites as particular examples government spending on research, roads, law enforcement, voting, parks, and transportation. Liscow, “Is Efficiency Biased?” 1674–77.
41 Liscow, “Is Efficiency Biased?” 1672 n. 80.
42 Alternatively, a multiplier could be applied to civil or criminal penalties imposed on the rich. This could be analogous to Finland’s well-known system of income-adjusted fines.
43 The specific tweak to legal standard proposed above may be insufficient to render the rule neutral; or it could be excessive, rendering the rule poor-biased. However, there is theoretically some level of advantage that could be given to the poor in order to render the rule neutral overall, whether it is a 55 percent burden of proof, 60 percent, and so forth.
44 Liscow, “Is Efficiency Biased?” 1695.
45 See, for example, Kaplow and Shavell, Fairness Versus Welfare, 35.