The Insider Monkey
Friday, September 17, 2010
Yesterday the Census Bureau announced a big increase in nationwide poverty rates in 2009. CBS reported the news as “One in Seven in Poverty: 44 Million Americans”. When it comes to statistics it is always wise to be a little bit suspicious. We’re talking about poverty rates here and that makes me even more suspicious. The number of Americans who are below the poverty line is not 44 million, it is more like 39 million. The Census Bureau is using a flawed methodology. Here is why:
1. The poverty line is $10,830 per year, before taxes. That single threshold remains the same for every American, whether he lives in Mobile, Alabama or the west shore of Maui in the Census Bureau’s methodology. To put it bluntly, that amount of money can go a lot further in Mobile than in Maui. Where would you rather live with that lump sum, in Mobile, where a gallon of gas goes for $2.50 or in Maui, where it goes for $3.75?
2. The Census Bureau does not take into the costs of housing, out-of-pocket medical care costs, clothing and utilities. I am not kidding. These types of costs are not uniform across the United States. In Harlem, for instance, residents saw their rents increased at a significantly higher rate over the past decade than did the residents of Mobile. But the Census Bureau is ignoring this.
(ARTICLE CONTINUES BELOW)
3. Also, the Census Bureau does not take into account housing subsidies, food stamps, tax credits, etc when they calculate the household income. Do I need to argue why we should include these when calculating the household income? They have been using a flawed methodology for 50 years, and they continue doing so simply because we can compare this year’s flawed results to last year’s flawed results. Brilliant!
This is only half of the story though. Despite all these methodological weaknesses I was curious enough to look for the list of states with the worst and best poverty rates. I skimmed through their 88 page report but I could not find a single table breaking down their flawed results by state. This must be top secret information, I said to myself. Fortunately there are public use files where anyone with a $10000 per year SAS program (yes, they only provided a SAS dataset) and 4 hours of free time can build a bottom-up model to calculate poverty rates by state, adjust the household incomes by government benefits and use poverty threshold levels that are adjusted for geographic cost of living (and took medical, housing, energy and clothing costs into account as well).
I am not going to get into the details of Census Bureau’s faulty “top secret” list of states with the highest poverty rates. (If you are curious enough you can see the rankings at Insider Monkey). Before presenting the state by state results of my superior methodology, let me tell you one thing: there are not 44 million people living in poverty. The number of Americans living below poverty line is around 39 million. The Census Bureau overestimates the poverty rate by around 1.4 percentage points in 2008. When we take into account government benefits, poverty rate declines by 3 percentage points. When we properly adjust the poverty threshold level, poverty rate increases by 1.6 percentage points. Considering that non-cash income and benefits are usually underreported, using the correct methodology reveals that headline poverty should at least be 1.4 percentage points lower. Alternatively, more than 4 million people should not have been classified below the poverty line.
If the Census Bureau had presented a list of 10 States with the highest poverty rate, States like Nevada, Washington, Nebraska, Georgia, and Massachusetts would have been on it. Washington and Massachusetts? Are you kidding me? Here is the more accurate TOP SECRET list of States with the highest poverty rate:
This article was posted: Friday, September 17, 2010 at 10:19 am