Empirics disprove income effect – and do not support reduction in labor incentives
Fleischer University of San Diego School of Law Professor of Law, and Hemel, University of Chicago Law School Assistant Professor of Law, 2020
[Miranda Perry Fleischer and Daniel Hemel “The Architecture of a Basic Income” The University of Chicago Law Review , Vol. 87, No. 3 (May 2020), pp. 625-710 JSTOR – pre-jcp-tm accessed 4.4.23]
The third and final component of the case for a UBI is the claim that redistributive cash transfers should be uncondi- tional—not limited to individuals of a particular age (for example, the young and the elderly), or to individuals with a demonstrable disability, or to individuals who participate in the workforce. This is probably the most controversial component of the case for a UBI and the point on which welfarism and resource egalitarianism are most equivocal in their support. (Libertarians may find the claim less problematic.) Again, we consider each theory in turn, high- lighting the ways in which lack of conditionality can be justified and evaluating potential counterarguments.
1. Welfarism.
Conditionality generates winners and losers. The winners are individuals who meet the conditions for aid (for example, by participating in the workforce or demonstrating disability). The losers are, of course, those who do not. The game is arguably zero- sum insofar as there is a fixed pie of resources available for redis- tribution: more recipients means less for each one. For welfarists, the trade-off between the scope of benefits (that is, reaching more people by removing conditions) and the size of benefits (that is, increasing the transfer per beneficiary) is one of the greatest chal- lenges in the design of a cash-transfer regime.
Conditionality would be more attractive if we knew who was capable of work and who was faultlessly in need. What we do know, by contrast, is that understandings of disability and de- servedness are ever evolving. Before the 1960s, for example, few in the United States understood alcoholism to be a disease.100 By the 1990s, alcoholism and drug abuse were considered to be disa- bilities that potentially qualified an individual for benefits under the Social Security Disability Insurance (DI) and Supplemental Security Income (SSI) programs—though Congress reversed course in 1996 and barred the Social Security Administration (SSA) from considering alcoholism and drug addiction as factors qualifying individuals for benefits.101 Prior to the 1980s, individu- als suffering from mental illnesses such as major depressive dis- order struggled to qualify for DI and SSI benefits;102 today, more than a third of DI beneficiaries and more than six in ten nonel- derly SSI beneficiaries fall into the “mental disorders” diagnostic category.103 Individuals unable to work on account of substance abuse disorders or depression might have been branded as “at fault” for their circumstances in an earlier era. Awareness of our own epistemic limitations should make us think twice before set- ting conditions that exclude individuals from aid today; it may be our own lack of knowledge—or lack of sympathy—that causes us to conclude that they are to blame for their circumstances.
Aside from the problem of defining disability in the abstract, any conditional cash-transfer program that denies benefits to able-bodied nonworkers encounters the additional challenge of applying the definition of disability in individual cases. Internal and external audits of SSA disability determinations find high er- ror rates, with studies suggesting that anywhere from 20 percent to 60 percent of disability benefit rejections are “false nega- tives.”104 Whatever the precise figure, welfarists must weigh the inevitability of error in a conditional transfer program against the putative upside of limiting aid to those who are “truly” in need.
The same welfarist logic that justifies redistribution in the first place would suggest that these error costs are asymmetric: providing nothing to disabled individuals who are erroneously classified as “work capable” is worse than providing somewhat less to disabled individuals who are correctly classified as such. This, again, is a function of the diminishing marginal utility of income: the first dollar matters more than the thousandth and the millionth. The welfarist who is worried about false negatives in disability determinations might well conclude that ensuring a subsistence-level income for all individuals is more important than trying to channel somewhat more in benefits to a subset of disabled individuals by excluding others from aid.
The trade-off between scope and size is not the only factor that a welfarist might consider when wrestling with the condi- tionality question. Proponents of making cash transfers condi- tional upon workforce participation or a demonstration of disabil- ity also worry about the effect of a UBI on labor supply. A UBI— so the argument goes—may lead to widespread exit from the workforce and thus to a reduction in total resources and overall welfare.105
There are two ways in which a UBI may affect labor supply: an “income effect” and a “substitution effect.” The term “income effect” refers to the effect on labor supply of making someone richer or poorer. If an individual becomes richer and therefore can meet more of her material needs or desires, then she may allocate more time to leisure; conversely, if she becomes poorer, then she may allocate more time to labor. The term “substitution effect” refers to the effect on labor supply of changing the terms of the labor-leisure trade-off. Labor becomes more attractive relative to leisure when after-tax wage rates rise (and, conversely, less at- tractive when after-tax wage rates fall).
A UBI financed through income or consumption taxes will in- crease the income of individuals at the low end of the income lad- der (who will receive more in benefits than they pay in additional taxes) and will reduce the income of individuals at the high end (who will pay more in additional taxes than they get back). We would expect the income effect on labor supply to be negative for low-income individuals (who now have more income) and positive for high-income individuals (who now have less). The substitution effect on labor supply of higher tax rates is negative across the board.
Past experiences with cash-transfer programs shed some light on the magnitude of these effects. One well-examined pro- gram involved members of the Eastern Band of Cherokee Indians in western North Carolina who began to receive annual disburse- ments of casino profits starting in 1996.106 The disbursements amounted to approximately $4,000 per year for each adult tribal member. A study comparing households with children that did and did not receive the casino-profits payments found no econom- ically or statistically significant effect on labor supply for mothers or fathers.107 Another program—the Alaska Permanent Fund Div- idend—has used the state’s substantial oil revenues to pay ap- proximately $2,000 per resident per year since 1982.108 A study comparing Alaska to other states with roughly similar de- mographics and industry compositions again found that introduc- tion of the dividend had no significant effect on overall employ- ment, labor-force participation, or hours worked.109 Note, though, that in both cases the cash-transfer program would have affected labor supply only through the income effect—there was no substi- tution effect, as tax rates (and after-tax wage rates) did not change. An inflow of casino profits or oil revenues is about as close as one can get to money dropping like manna from the heavens. The Eastern Cherokee and Alaska natural experiments thus sug- gest that the income effects on labor supply of a cash-transfer pro- gram in the range of $2,000 to $4,000 per person are small or non- existent, though they tell us nothing about the substitution effect. Studying the substitution effect of a UBI is somewhat more complicated. Much is likely to depend on the rate of the tax (or tax plus phaseout) used to fund a UBI—the higher the rate, the larger the labor-supply effect. Much will also hinge on whether the UBI supplements or replaces existing cash-transfer pro- grams. Some of those programs entail very high implicit tax rates. For example, a nonblind individual who earns more than $1,220 per month in 2019 will lose DI benefits, which can mean a loss of up to $771 per month for a single individual.110 The marginal tax rate on the 1,221st dollar is thus as high as 77,100 percent. Re- placing DI with a UBI and a less cliff-like marginal tax rate sched- ule may increase labor supply among former DI beneficiaries.
Finally, it is not obvious whether the labor-supply effects of a UBI—even if negative—should be considered a bug of a basic in- come or a feature. The aim of a “sufficient social income,” wrote the late Austrian-French philosopher André Gorz, “is not to force the recipients to accept any kind of work on any terms whatso- ever, but to free them from the constraints of the labor market.”111 A UBI, Gorz emphasizes, potentially will allow individuals to opt out of low-paying and otherwise unrewarding market labor and instead to devote their time to their families, their communities, or to other ennobling pursuits. While Gorz does not frame his ar- gument in welfarist terms, the welfarist could make a similar point: The objective of welfarism is to maximize well-being, not to maximize output. Some amount of labor-market leakage may be tolerable and even desirable if it allows individuals to live more fulfilling lives.