A Retrospective on the “War On Poverty”

On the 50th anniversary of President Johnson’s bizarre declaration of “unconditional war on poverty”, here’s Robert Rector discussing the results.

Please note that Rector is a fellow at the Heritage Foundation, so he is very definitely giving the conservative point of view.

The federal government currently runs more than 80 means-tested welfare programs that provide cash, food, housing, medical care and targeted social services to poor and low-income Americans. Government spent $916 billion on these programs in 2012 alone, and roughly 100 million Americans received aid from at least one of them, at an average cost of $9,000 per recipient. (That figure doesn't include Social Security or Medicare benefits.) Federal and state welfare spending, adjusted for inflation, is 16 times greater than it was in 1964. If converted to cash, current means-tested spending is five times the amount needed to eliminate all official poverty in the U.S. [emphasis added]

Let’s think about the scale of this:

• 80 different programs, all with offices and bureaucrats
• $900 B spent per year — that’s 1/4 of Federal outlays
• 100 million receiving aid — that’s a 1/3 of the population
• 16 times greater is a 1500% increase — you need to sustain a 5.7% growth rate for all 50 years to get that. In fairness, this should be adjusted for population growth too: so the 5.7% should be reduced by the 1.1% per year population growth rate.
• 5 times the official amount needed — the sad fact is that most of the anti-poverty spending goes to the paychecks of people who administer the programs: if we just gave 1/5 of the spending straight to the poor in cash we could raise all of them above the poverty line at once.

Rector points out one of the ways in which we can still officially have poverty in spite of these efforts:

  … The country has invested $20.7 trillion in 2011 dollars over the past 50 years [in anti-poverty programs]. What does America have to show for its investment? Apparently, almost nothing: The official poverty rate persists with little improvement.

That is in part because the government's poverty figures are misleading. Census defines a family as poor based on income level but doesn't count welfare benefits as a form of income. Thus, government means-tested spending can grow infinitely while the poverty rate remains stagnant. [emphasis added]

This is bad, but not as bad as it sounds. It is useful to measure poverty before welfare. But that doesn’t tell us anything about whether welfare is helping or not. What you really need to do is pair that with a second statistic. It turns out the government does produce that figure, but what’s goofy is that no one talks about it. Here they are charted together:

The blue line is the official poverty rate: the peaks roughly correspond to the period after a business cycle trough when most of the economy is recovering but the poor are still lagging. The red line is the poverty rate adjusted for taxes the poor pay, transfers they receive, and income and expenses not measured in the blue line. And this shows substantial improvement over the last 50 years: over a third of poverty eliminated.

Interestingly, Rector doesn’t mention a huge consideration: should we evaluate poverty by relative or absolute measures? Our government uses the relative measure, so this is all you hear about.

Relative poverty is where your income is judged relative to the national average. For example, if average income is $40K, perhaps we’d define anyone with income half that or less as poor. Fair enough. The problem is that this doesn’t account for economic growth. Suppose that average real income goes up to $60K. Then the threshold for poverty is now $30K. The problem is that someone who earns $30K now is considered poor, while someone who earned that much before was not considered poor. So this definition is really more about inequality than it is about poverty.

Absolute poverty is where we measure income against some standard: will you have enough to eat, enough shelter, enough leisure, and so on. Measured this way, all developed countries continue to have some poverty, but far less than in the past.

For policy purposes, the extent to which voters interpret the lack of improvement in the poverty rate as lack of improvement in absolute poverty is problematic. People are very sensitive to absolute poverty, and don’t mind supporting efforts to alleviate it. With relative poverty … not so much.