|
Wednesday, September 05, 2007
Samuelson Abuses Census Data
In his Washington Post column this week, Bob Samuelson abuses Census Bureau's Income, Poverty, and Health insurance Coverage in the United States 2006 to launch a critique of immigration policy.
The gist of his "reasoning" is this: From 1990 to 2006, the number of poor people increased by 2.9 million people. In those same years, the number of poor Hispanic people increased by 3.2 million while the number of poor whites and blacks and fell by 0.6 million and 0.8 million respectively. If we subtract out the increase in poor Hispanic individuals from the increase in the total number of poor individuals, we are actually left with a net decrease in the number of people in poverty from 1990 to 2006. Ergo, satisfactory progress has been made in poverty remediation, and flawed immigration policies primarily responsible for strained social services, health care, and public education systems.
Why is it important to get this story straight?
One reason is truthfulness. It's usually held that we've made little, if any, progress against poverty. That's simply untrue.
...
We shouldn't think that our massive efforts to mitigate poverty have had no effect. Immigration hides our grudging progress.
A second reason is that immigration affects government policy. By default, our present policy is to import poor people. This imposes strains on local schools, public services and health care.
Samuelson, however, is simply peddling statistical misdirection and obfuscation.
First, Samuelson omits the inconvenient truth that Hispanic poverty rates are falling significantly faster than the nation's as a whole. He leaves out the rate of decline of the Hispanic poverty but blames a tide of impoverished illegal immigration for the widespread belief that little progress has been made against poverty generally.
Although the number of poor Hispanic individuals increased from 1990 to 2006, the Hispanic poverty rate declined 7.4 percentage points. This is massive compared to the 0.6 and 1.2 percentage point declines for the whites and the entire U.S. population respectively. From 1990 to 2006, the national poverty rate fell from 13.5 percent to 12.6 percent - a nine percent drop. Yet, in that same 17 year period, the Hispanic poverty rate declined from 28.1 percent to 20.1 percent - a 26.4 percent drop .
(click to enlarge)So let's play Samuelson's game and create a world with no illegal immigration (read: no Hispanics). Omitting Hispanics from the poverty rate calculation, the national poverty rate declines 1.3 percentage points in the past 17 years. When Hispanics are included, the national poverty rate falls 1.2 percentage points. In either scenario, the decline in the poverty rate is scant and worthy of the complaint that "little, if any," progress has been made in the fight against poverty.
(click to enlarge)Second, it is not so clear that Hispanics are straining social services and the public education system, yet Samuelson proposes to alter public policy based on his shoddy 800-word analysis. Median Hispanic household income increased 13.1 percent while median national household income increased 7.6 percent from 1990 to 2006. So, tight immigration policies may reduce the number of individuals using social services and public education resources, but they may also reduced a rapidly growing source of revenue for those same social services. Without the appropriate data on costs of and participation rates of social services, it is impossible to tell which is increasing faster - funding or use - but knowing that information is absolutely critical in analyzing public policy. Samuelson simplistically ignores these data to push anti-immigration policies.
(click to enlarge)Lastly, Samuelson does not specify why he chooses to compare 1990 and 2006. It's an odd, 17-year span. If those two years represent similar points in an economic recovery (poverty and economic cycles are related), he doesn't say. Had he compared 2005 and 2006, he would have written about how all ethnicities' poverty numbers are in decline. Similarly, a comparison of 1993 and 2001 would have revealed reductions in the number of poor in all ethnic groups. Had he compared 1990 to 2004, he would have opined on the success of combating black poverty while bemoaning the increase in white and Hispanic poverty. However, he probably would have failed to note the falling poverty rates for all ethnicities. One could go on. Samuelson's arbitrary use of comparison years does nothing to foster the impression that his analysis is agenda-free.
(click to enlarge)To be clear: I am not staking out a position on immigration policy. In fact, it's precisely because of my limited command of the relevant information that I am not voicing an opinion on immigration. Samuelson, on the other hand, demonstrates a similar lack of knowledge, and yet he uses a simplistic understanding of demographic data as a basis for immigration policy critique.
Posted by Craig Jennings
|