The High Probability of Being Poor

Late last month, the Associated Press ran a report about economic insecurity that managed to gain some traction in certain parts of the political internet, and since then, again and again in certain relevant debates. The statistical bomb dropped in the first sentence of the report really says it all:

Four out of 5 U.S. adults struggle with joblessness, near poverty or reliance on welfare for at least parts of their lives, a sign of deteriorating economic security and an elusive American dream.

To be clear, this figure pertains to the percentage of people facing these problems at least once in their life, not the percentage of people facing them right now. Also, it should be noted that this figure cannot, by itself, be a sign of deteriorating economic security. To show things are deteriorating, you'd have to know whether this figure used to be lower than 80 percent, and we do not know that.

Shortly after the AP report blew up, the Wall Street Journal's James Taranto responded with a few criticisms that are worth reading. Taranto argues that using near-poverty to derive the figure (meaning 150 percent of the poverty line) instead of the poverty line is arbitrary. Taranto also takes issue with the joblessness part of the figure because it counts even a day of joblessness, and takes issue with counting anyone who has ever received welfare benefits into the figure as well.

In light of Taranto's criticisms, an interesting question presents itself: What would this figure look like if it did not include those features which Taranto objects to? If we used the poverty line instead of the near-poverty line and did not include joblessness or welfare receipt into the calculation, how much lower would the figure be? Lucky for us, in a variety of papers and books that came out in the late 1990s and early 2000s, Mark Rank and Thomas Hirschl published data on precisely this question.

In their 2001 paper, Rank and Hirschl used PSID data to determine that 51 percent of people experience poverty (as defined by the Census) at some point in in their life between the ages of 25 and 75.

So in this graph, you notice that at age 25, around 6 percent of people have experienced poverty (presumably in that very year). And then from there, the number grows and grows. It cannot decline obviously because you cannot un-experience poverty. By the time folks reach age 75, 51 percent of them have experienced at least one year of poverty.

But as with all things economic, the aggregate picture obscures the fact that some demographics (namely women and people of color) have it much worse.

On the most extreme end, 98.3 percent of black women who have less than a high school education experience poverty at some point in their lives. For black women who have a high school education or greater, the figure is still a very high 77.5 percent, compared to just 30.7 percent for white males with a high school education or greater.

So even if you get rid of the various enhancements that the Journal's Taranto takes issue with, the picture is still pretty grim. A slight majority of people still spend at least one year of their adult life in poverty, and for some demographic groups, almost everyone experiences poverty at some point. The ultimate figure is not as high as 80 percent, but troubling nonetheless.


One of the reason of that is that million people are unemployed that is why many homelessness has increased by 50 percent in some cities.

buy keek like

The official poverty measure has been criticized for not accounting for several factors that can affect a family's economic well-being and for not having been updated, except for inflation, for four decades.

For example, while cash benefits from government assistance programs are included in a family's income when calculating the official poverty measure, benefits received in-kind such as food stamps, Medicare or Medicaid, employer provided health insurance, housing subsidies, and other social services are excluded. Taxes that families pay and tax credits they receive such as the Earned Income Tax Credit (EITC) do not enter into the official poverty determination.

Additionally, the threshold value a family must earn to escape poverty was developed in the 1960s by combining emergency food budget data from the US Department of Agriculture with an estimate of what fraction of income families spend on food. Although the thresholds are adjusted each year for inflation, some analysts believe that these numbers no longer accurately reflect the minimal resources a family requires.

Since 1979, the Census Bureau has published a variety of experimental poverty measures using expanded definitions of income and alternative methods to account for inflation.[5] These alternative measures tend to show lower levels of poverty than the official measure in any year, but the timing of increases and decreases in the poverty rate is very similar across measures. This similarity suggests that, despite the criticism it receives, the official poverty measure provides a reliable indicator of changes in the poverty rate from year to year.

In addition to the Census Bureau's alternatives, a 1995 panel appointed by the National Academy of Sciences (NAS) suggested a series of poverty measures based on alternative definitions of both income and needs (the amount of resources a family needs to escape poverty).[6] The income measures consider taxes paid, tax credits received, some noncash transfers, and deduct medical expenses and work expenses such as childcare and transportation costs. The needs thresholds are derived from a typical family's spending on food, clothing, shelter and utilities and some measures incorporate geographic differences in cost-of-living.

These alternative definitions tend to show higher levels of overall poverty than the official measures in any year, although the difference is usually less than one percentage point. The incidence of poverty across groups also differs under these measures. Single-mother households fare somewhat better and married couples tend to have somewhat higher rates of poverty than when measured using the official methodology. This discrepancy likely results from the differing level of noncash benefits received by these two types of families.

You need to be logged in to comment.
(If there's one thing we know about comment trolls, it's that they're lazy)