the "failure" of the War on Poverty

This year marks the 50th anniversary of the beginning of the War on Poverty-- at least in terms of legislation. (Appropriations expanded a bit under LBJ and especially under Nixon.)

One of the first observations to make is that progress against (measured) poverty stopped once the War on Poverty began. (Insert "government is inept" jokes here.) With economic growth, poverty was decreasing rapidly post-WWII. A rising tide was, not surprisingly, lifting most boats. Since the War began, the poverty rate has merely fluctuated between 11-15% over the last 50 years.) Once the War began, progress stopped. But why?

If we use lame statistics, univariate analysis, and false-cause fallacy-- common errors, especially for those on the Left-- the War on Poverty stands (easily) condemned. Poverty falls; the War begins; poverty quits falling; the War is a failure; QED. (Not surprisingly, the Left does not do their usual analysis in this realm.)

If we move to more sophisticated analysis, the answers are more complex, interesting, and realistic.

Why did measured poverty quit falling when the War began? There are three primary reasons:

1.) Paying people to be in a state (and removing the payments if they move away from that state) will generally encourage them to remain in that state. The more you pay them-- and the longer you make the payments available-- the greater the problem. (See also: other welfare programs like "unemployment insurance".) So, on the basis of this point, the War on Poverty will increase poverty. This is the favored argument of those on the Right-- and is clearly true to some extent, but how much? (Daniel Mitchell relies too much on the poverty rate in his opening, but otherwise has a nice post on this, including a discussion of the high marginal tax rates imposed on the poor.)

2.) It could be that government started the War at the moment when most of the curable poverty had been handled by the market. Perhaps the market grabbed the low-hanging fruit-- and the government happened to get involved just when the high fruits were still on the tree. In a most unfortunate coincidence (at least for statists and those who can only manage lame analysis), it looks like government was getting in the way of progress, when it was mostly treading water. This is not a particularly flattering view of government policy (especially at the federal level), but it takes some of the blame/heat off of their seeming ineptness.

3.) The measurement of poverty is profoundly flawed. Against its poverty lines (adjusted for family size and inflation), the government compares measured, annual, cash income for each household. The poverty lines are a proxy/standard set by the government-- adjusted by inflation (a proxy for the economy's increase in prices). All in all, one can certainly quibble, but at least by government standards, these lines were calculated with a reasonable methodology in the 1960s and have been reasonably well-adjusted since then.

The bigger issues are with the government's measurement of income. First, the government is only picking up reported "income" by each "household". Unreported income is likely-- when one is engaged in illegal activity or getting paid under the table to avoid taxes or welfare benefit reduction. If people are shacking up, they might live like a household with a bigger income, but be measured as if they're in two separate households with smaller incomes.

Second, the government is only measuring *cash* income. So, cash benefits paid by the government are counted (e.g., Social Security), but non-cash / in-kind benefits are ignored. We could give every poor household $50,000 in food stamps and they would still be counted as poor by the government. In this, the poverty rate is much more of a measure of dependence on government than a measure of those living in poverty. Or putting it more bluntly: unless they're refusing assistance, no one lives in material poverty in the U.S. (as measured by the government's poverty standards).

Third, the government ignores wealth and focuses exclusively on annual income. One might have considerable wealth, but modest income and be measured as poor. This explains the impressive data about the poor in terms of home ownership and other consumer goods.

Similarly and finally, the government's measure of income is only a snapshot-- the statics of one year rather than  the dynamics of many years. Thus, it says *nothing* about the larger question of how "poor people" are doing five and ten years into the future. As an example, I had a number of colleagues in grad school who were "poor". Likewise, the country's highest poverty rates are found in college towns.

Is there poverty? Sure. Using static analysis, has government increased the living standards of many poor people? Sure. Using dynamic analysis, has government increased poverty by subsidizing it? Sure. The first two questions are obvious and not all that interesting. The third question is complex and difficult to measure. But any layman concerned about poverty should know the basic point-- that the statistics used to measure poverty are lousy.

How to adapt policies to better help the poor? Again, a complex and difficult question-- for another day. How to stop crony capitalism and government policies that harm the poor? Ahhh, I know a good book or two on that topic-- if you care...