In an excellent Op-Ed in the New York Times David Shipler (who had a Magazine article on the “working poor” in the Times a few weeks ago) writes: “About 35 million Americans live below the federal poverty line.” I think it is important that we take a moment to think about where these numbers come from.
In 1963-1964, Molly Orshansky of the Social Security Administration developed poverty thresholds. Orshansky based her poverty thresholds on the “thrifty food plan,” which was the cheapest of four food plans developed by the Department of Agriculture. The food plan was “designed for temporary or emergency use when funds are low,” according to the USDA. Based on the 1955 Household Food Consumption Survey from the USDA (the latest available survey at the time), Orshansky knew that families of three or more persons spent about one third of their after-tax income on food. She then multiplied the cost of the USDA economy food plan by three to arrive at the minimal yearly income a family would need. Using 1963 as a base year, she calculated that a family of four, two adults and two children would spend $1,033 for food per year. Using her formula based on the 1955 survey, she arrived at $3,100 a year ($1,033 x3) as the poverty threshold for a family of four in 1963.
From Doug Henwood’s book, After the New Economy, we learn that when the Johnson administration set the poverty line in 1965, they ignored more recent research by Orshansky and others which showed that
… the share of income the average household devoted to food had declined from one-third to a quarter, which suggested that the multiplier should be four, not three, and which would translate into a significantly higher poverty line.
Even more troublesome, these numbers are still the same numbers used today, adjusted only for inflation! This article points out some of the changes which would make the numbers even higher:
The guidelines have never been updated to account for changing household consumption patterns. Families no longer spend one-third of their income on food and two-thirds on other basic needs. “Food now accounts for more like one-sixth of the family budget. Housing, transportation and utilities are much larger components of family spending,” said Laura Connolly, an OSU economist.
- The guidelines fail to take into account the extra costs of two-earner families, such as clothing, transportation and, perhaps biggest of all, childcare. “The cost of child care was not figured in to the official guidelines because the typical family in the 1950s had one wage earner and a stay-at-home mother,” Connolly said.
- The guidelines do not recognize geographic differences, even though the cost of food, clothing and housing varies from state to state and within states. “In a high cost state such as California, the official poverty thresholds are probably too low. In a lower cost state, they are too high,” Connolly said.
Henwood suggests that it isn’t enough to simply adjust the numbers underlying the poverty line, he suggests replacing such a fixed number with a “relative measure”:
A relative measure has several advantages. First, price indexes are notoriously devilish to work with … How do you adjust for improvements in quality? If the price of an item goes up, but its quality improves, did it really get more expensive? if the price of apples goes up, and that of pears goes down, what does that do to the price of fruit? The problems multiply over the long term: there were no VCRs or Palm Pilots in 1965; what does their development mean to the changed cost of living thirty-five years later? A relative measure avoids all of these problems.
And second, relative measures comport much more closely with the way people perceive themselves. As no less revered an authority than Adam Smith (1976, bk V. Chap. 2, pt. 2, art. 4) defied it in 1776, poverty was characterized by the want of “necessaries,” which he in turn defined as “not only the commodities which are indispensably necessary for the support of life, but whatever the custom of the country renders it indecent for creditable people, even of the lowest order, to be without.” People perceive themselves and others as poor or nonpoor by comparing them to their neighbors, not some number invented by statisticians fourty years ago and adjusted by another set of statisticians every year.
Support for such a relative measurement can be made even stronger, when based on Amartya Sen’s studies which show that mortality rates track, not absolute income, but relative income. As I stated in my last post,
… it is harder to function as a poor person in a rich society than in a poor one, even if you have more material possessions. An argument borne out by the fact of lower life expectancies amongst poor and minority populations in industrialized nations when compared with materially poorer populations in developing nations.
Henwood follows the widespread practice (among academics) of drawing the poverty line at “half the national median income.” He argues that this is not only supported by Gallup polls on subjective measurements, but it is also “about where Orshansky’s original line fell.” The effect on poverty measurements is dramatic:
In 1998, the poverty line for a family of three was $13,003 and $16,660 for a family of four. That year, half the median income for a family of three was $24,466 and $28,030 for a family of four — 88% and 68% above the official poverty lines. Even the modest assumption that the poverty line should be half again as high as it is would suggest a poverty rate of 22% in 1998, rather than the official 12.7% figure…
In a recent post at Calpudit, Kevin Drum worries that (based on the lower numbers given by the official poverty line) “it’s hard not to simply decide that the problem is intractable and give up,” but Henwood tells us that it wouldn’t take much money at all to balance the inequalities:
The amount of money it would take to bring all officially poor households up to the pverty line is amazingly small: 0.5% of GDP, or jsut over 3% of the income of the richest fifth of households. It would take a bit more money to bring the poor up to a civilized standard, but not that much. Doubling the incomes of the poorest 20% of households — form an average of $10,136 (in 2001) to $20,272, which is still less than half the median — by taxing the richest 20% would require the affluent fifth to sacrifice less than 7% of their income, bringing it down from an average of $145,970 to a mere $135,834. That would reduce their share of the national pie to late-1980s levels, hardly a period when the upper orders were suffering. But clearly it’s much more important that the affluent be able to buy Hummers than to accomplish this bleeding-heart goal.
- More on poverty at Wikipedia.