Guest editorial: Measuring poverty not as easy as it sounds

Published 9:08 am Friday, July 14, 2023

By Mike Walden

N.C State University

I can remember when President Lyndon Johnson proclaimed the war on poverty in 1964. Since then, a variety of governmental programs have been developed to address poverty with a goal of reducing the number of people who live in poor conditions. By some estimates the total amount spent since the war on poverty was declared exceeds $20 trillion.
One way to determine if this war is being won is to track the poverty rate or percentage of people who are living in poor conditions. But to do this, we must have a way of measuring poverty. Perhaps surprising to many, there are actually numerous ways of measuring the poverty rate, without total agreement about which is best.
Conceptually, determining if a household is poor should be straightforward. Two items are needed: the amount required for a household to cover basic needs and the amount it has to meet those basic needs. Both the amount of resources needed and the amount of resources available are measured in dollars.
Then, determining poverty is just a matter of comparison. If a household has enough resources to cover basic needs, then that household is not considered to be poor. Conversely, if the household’s available resources fall short of resource needs, then the household is categorized as poor. The total number of people – including adults and children – in households labeled poor as a percentage of the total population is the poverty rate.

You can probably anticipate why there are disagreements over the best poverty measure. Many questions have to be answered in calculating poverty. How are basic needs determined? How are these basic needs updated each year? What should be included in determining the basic resources of the household? Should only earnings from working be included? How should taxes on work income be handled? And what about the resources a household receives from government programs, such as Supplemental Nutrition Assistance Program (SNAP) benefits, or Medicaid and Medicare for health care? Should these benefits be counted as resources for the household just like earnings from work?

Let me begin with the question of determining basic needs. The U.S. Census Bureau, which publishes the official poverty rate, uses a rather odd calculation. It takes the average yearly expenditure on food for a household of a given size and multiplies it by three to obtain the annual dollar amount required for basic needs. When this method was developed in the 1960s, food expenses were one-third of total expenses for necessities, which included food, shelter, clothing and utilities. The number is updated each year to account for inflation.

Over the last six decades, this methodology has been criticized on two counts. First is the assumption that food costs have remained at one-third (33%) of the total costs of food, shelter, clothing and utilities. Indeed, today’s food share has dropped to 30%.

The second criticism is that only four items – food, shelter, clothing and utilities – are considered necessities. Notably absent is health care. Addressing both of these issues would increase the basic needs level for households and – everything else being equal – increase the poverty rate. But everything else is not equal because there’s also the income side of the poverty equation: what to include as income.

The official poverty measure calculated by the Census Bureau only counts cash earnings as income. The earnings can be from work, investments, Social Security, unemployment compensation and the federal cash assistance program called Temporary Assistance for Needy Families. No tax deductions or tax credits are included.

A criticism of this income calculation is it doesn’t include the financial value of noncash programs that provide resources. A good example is SNAP, the successor to food stamps. SNAP benefits are deposited in an electronic account the recipient can only use for purchases at authorized retail food stores.
In 2011 the Census Bureau began issuing a supplemental poverty rate, which includes as income the financial value of programs like SNAP and housing subsidies. Tax credits, like the Earned Income Tax Credit, are also considered as income, but taxes paid and any expenses related to work, child care and out-of-pocket medical costs are now subtracted.

Notice that missing from income is the financial value of the two big public medical assistance programs, Medicare and Medicaid. There have been some private efforts outside the Census Bureau to estimate the impact on the poverty rate including the financial value of these programs.

Has there been progress in reducing the poverty rate since the 1960s, and if so, how much?   Both Census Bureau measures – the official poverty rate and the supplemental poverty rate – suggest noticeable progress has been made, with the poverty rate falling from 22% in the early 1960s to 12% for the official rate and 8% for the supplemental rate today. The calculation made by some academics outside the Census Bureau indicates a bigger decline, with a current poverty rate of nearly 2% when the value of all government benefits is included.

Who knew calculating poverty would be so difficult and could produce such dramatically different results? However, just like baking a cake, the result will be determined by the ingredients. Which ingredients should go into calculating poverty? You decide.

Mike Walden is a William Neal Reynolds Distinguished Professor Emeritus at North Carolina State University.