Have Antipoverty Programs Increased Poverty?

In “Poverty in the U.S. Was Plummeting—Until Lyndon Johnson Declared War On It,” Daniel J. Mitchell presented several different researchers’ conclusions about how poverty programs starting with the War on Poverty have shot themselves in the foot with their adverse incentives and slammed the brakes on poverty reduction rather than accelerating it.

There is nothing wrong with what is presented, which is very valuable information. But there is something that can make the case even stronger than utilizing overall poverty rates: breaking down the effects by the age of primary breadwinners. And while I have not seen such data brought up to the present, James Gwartney and Thomas McCaleb did that in “Have Antipoverty Programs Increased Poverty,” published in the Cato Journal in 1985.

Revisiting their work is important because it provides concise explanations of the types of disincentive effects involved and also why the productive disincentives that welfare programs create will have more adverse effects the younger the household, allowing disaggregating the data by age-groups to show the differences in effects more clearly.  

Gwartney and McCaleb talk about four disincentive mechanisms: the higher real benefit effect, the high implicit tax effect, the skill-depreciation effect, and the moral hazard effect.

The higher real benefit effect is that “Increases in the real value of benefit payments make dependency on the government even more attractive compared with the alternative of self-support.” That effect will be greater for younger workers, whose earnings potential are lower than older, more experienced workers.

The high implicit tax effect is that means-tested poverty programs reduce benefits as households earn more, imposing the equivalent of an additional income tax on increased earnings. When the reality of multiple programs is factored in, that implicit tax rate can be very high—far higher than the highest official tax rate on earned income, and for some circumstances, well over 100 percent. Consequently, “Such high implicit marginal tax rates pose a significant disincentive to work for those individuals whose potential earnings are relatively low.”

The skill-depreciation effect is that individuals who have not used their skills for extended periods, such as those who spend much time outside the labor force, find that those skills erode. Not only does this get worse the longer such incentives last, the effect is greater for younger workers: “As transfers make dependency more attractive relative to work experience, schooling and other forms of human capital investment, youthful recipients fail to develop skills that have in the past enabled the young to escape from poverty.”

The moral hazard effect is that welfare assistance can help finance some in choosing “a lifestyle that increases the likelihood of poverty.” That incentive is more damaging for one’s productive life the earlier it begins.

Not only are each of these effects more severe for younger households than others, they also have much smaller productive effects on older workers and virtually none for low-income families whose members are retired. Just ask yourselves how worried you are about those effects on your grandparents. That means the disincentive effects on younger low-income workers can be seen in contrast to older low-income households.

And those effects are large. As Gwartney and McCaleb found, after the substantial decreases in poverty for all age groups before the War on Poverty began, both official poverty rates and poverty rates adjusted for in-kind benefits (not officially counted as income) for the elderly (for whom the disincentive effects are minimal) continued to fall dramatically, from 15.9 percent in 1968 to 5.5 percent in 1982. For the 45-54 age bracket, adjusted poverty rates fell from 6.7 percent in 1968 to 5.8 percent in 1975, rising thereafter to 8 percent. For the 25-44 age bracket, adjusted poverty rates only fell from 8.6 percent to 8.5 percent at first but rose substantially after, to 12.3 percent in 1982. Finally, for the youngest group studied, householders under 25, adjusted poverty rates rose from 1968 on, going from 12.3 percent in 1968 to 24 percent in 1982.

Daniel J. Mitchell is an invaluable source of information for those of us who care about liberty. I owe him thanks for informing me on many occasions. It is just that looking at the differential effects between younger households and older ones can add a striking contrast to the evidence he presents so effectively. And as productive as he is, he might well extend such a comparison into the present. I know I’m hoping.