Magazine article Public Welfare

Measuring Poverty

Magazine article Public Welfare

Measuring Poverty

Article excerpt

In the spring of 1978, Mollie Orshansky and others joined in a round table session to explore some of the issues intrinsic in determining an official poverty line. When this article was published, the author was with the Office of Research and Development, Social Security Administration, Washington, D.C.

What are the problems in measuring poverty? For one, the choice of a method for measuring poverty depends on who is counting and why. The real difficulty is that there is not one right way to do it--there is not one set of poverty statistics on which to base a measure. To coin a phrase, poverty lies in the eyes of the beholder.

Besides the dollar levels designated as the official poverty criteria for counting the poor, other measures such as the Bureau of Labor Statistics (BLS) lower budget are often used. The food stamp program claims to base eligibility on the BLS poverty lines but includes add-one for taxes, housing, or other expenses that meant, in 1475, a variation from the official poverty figure of $5,500 to $6,700 for food stamp households. But apart from the fact that different people like different definitions, there is not in place a continuous series of official counts.

The present measure was first published in 1965, to apply to income for 1963. We then lacked scientific norms for family needs for any items other than food, so we constructed poverty lines by estimating the dollar costs of food that would provide an adequate diet for the least amount of money (we used the Department of Agriculture 1955 economy food plan), and then by estimating further how much total money income a family had to have before it could afford to purchase this amount of food and still have enough left for other necessities. The factor by which food costs were transformed into minimum total family income requirements became known as the multiplier. For families of one or two, the factor multiplier was higher to reflect the relatively higher fixed costs such families face.

The same situation exists today; we have not come even close to a consensus on the amount of money needed for items other than food. Therefore we resort to the same surrogate procedure--one that originally implied, and probably still does, that families strapped for funds could cut expenditures for nonfood below the average to the same degree we estimated for food, although reduction of such amounts in housing and some other expenses, as we said at the outset, was probably unrealistic. In any event, for good orbed, the 1963 costs of the food plan, originally developed with an eye to 1955 family food choices, remain the core of the official poverty lines. These lines are adjusted only for year-to-year overall price change as measured by the Consumer Price Index of the Bureau of Labor Statistics.

Since 1969 the Census Bureau numbers--based on the Social Security Administration definition--have been designated the official statistics on the poor. They have been carried back to 1959, with adJustments for price change as reflected by the Consumer Price Index. There have been some technical revisions by the Census Bureau in processing and collecting income data through the years--data that we use to define who is poor. Admittedly these are improvements over making no adjustments in measuring at all. However, the altered methodology for processing in a single year has sometimes resulted in as large an annual decrease in the numbers we count as poor, just on the basis of procedures, as in other years we may attribute to an improved economy or effectiveness of antipoverty programs. The Census Bureau counts from the Current Population Survey, which is a regular annual series, showed almost 26 million poor in 1975. Had there been no changes in procedures since 1974, the number would have read 27 million. A new set of numbers for 1975 has derived from a survey with a larger sample and a better income data collection--at least we think it is better--and accordingly, instead of 26 million that the March Current Population Survey had recorded for 1975, they came up with 24 million. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.