Upon further review, there is no proof of a near-poor spike

A recent census report appeared to suggest that the nation has a huge segment of its population hovering near poverty. It was a striking piece of information, one I reported on in my column. Unfortunately, it's not accurate.

The U.S. Census Bureau report featured a new poverty assessment formula called the Supplemental Poverty Measure (SPM). Creating the furor was one poorly designed graphic indicating that the nation's population of "near-poor" is much larger than previously thought.

The information caused a small stampede into print, one that drew much media attention, ranging from the venerable New York Times to this columnist.

The offending graphic employs a side-by-side comparison of two bar graphs, a format that normally involves apples-to-apples analysis. It shows the U.S. population of 306 million broken into brackets of income-to-poverty-line ratios.

What leaps off the page is the difference in the "near poor" bracket, those whose incomes are one to two times above the poverty line. Using the new measure that number of near poor is 31.8 percent, compared to 18.8 percent using the long-standing official measure. That's 97 million citizens versus 58 million.

Developed in the 1960s, the official measure fails to account for the impact of the many anti-poverty programs launched since then. It is based on "the cost of a minimum diet multiplied by three (to allow for expenditures on other goods and services)"; and household income includes only "before-tax cash income."

The formula misses in-kind assistance, such as food stamps, a $50 billion program, and many housing subsidies. The new SPM measure captures these previously overlooked in-kind assistance programs.

But why the seeming huge increase in citizens considered near poverty? A closer examination shows that while the new SPM measure is undoubtedly better, it is also different, a fact camouflaged by the misleading apples-to-apples format of the Census report graphic.

SPM income, it turns out, is an after-tax figure, while the official poverty measure of income is a pre-tax number. No wonder there was a seemingly big change.

Obviously, the two are very different. As household income increases, the household qualifies for fewer anti-poverty programs and pays an increasing amount of taxes (FICA and income taxes). So, taxes represent the major difference between the two bar graphs.

For example, consider a household with pretax income of $50,000, or more than twice the poverty threshold. Assuming a combined FICA and income tax rate of 20 percent, this equates to after-tax income of $40,000, or less than twice the poverty threshold - dropping the household into the enlarged "near poor" population. Nothing changed in the household's economic circumstances; it dropped solely as the result of a change in poverty measurement methodology.

This statistical warping does not exist at very low levels of income, when households pay little or no taxes. In fact, both the official poverty measure and the new SPM measure arrive at approximately the same figure for the poverty line: $22,113 for a family of four under the traditional formula and $24,343 using SPM.

Just to be sure I wasn't missing something I spoke with Census statisticians. I came away convinced that the tax differential is the predominant factor responsible for the much greater number of people being captured within the SPM bracket of up to twice the poverty rate, or near poor.

In my rush to share the census data I bought into the rise in poverty angle, a sin shared by the New York Times, but one it has not conceded to. While the population of low-income citizens has certainly grown over the last three years, it has not skyrocketed in some previously unrecognized way based upon any sea change in the economic circumstances.

Basing decisions on the deceptive graphic would lead to bad policy. Policymakers need to take a closer look, as I did.

Red Jahncke heads the Townsend Group, a business consulting firm in Greenwich and is a regular contributor to The Day.