In 2019, Ajit Banerjee and Annie Duflo won the Nobel Prize in Economics. They won the prize for their work on poverty across the world, using rigorous experimental methods to understand the nature of poverty in some of the world’s poorest regions.
In their 2011 book Poor Economics, Banerjee and Duflo share the results of an experiment to learn about the relationship between poverty and hunger. In this study, they went to people with the lowest of low incomes and asked them about calorie intake. They tracked what people were eating and how many calories they were consuming in their state of extreme poverty.
Their theory was this: if people with few resources have low caloric intake, they will increase their caloric intake as they receive money. So they tested it, giving people money and seeing how their caloric intake changed as they received this money.
A strange thing happened. Among people who received money, their caloric intake decreased when they received it. So people who had the least money in the world, given the chance to increase their caloric intake, actually decreased their intake of calories.
Well why was this? It turned out that people who were living with the lowest amount of resources in the world decided to change their diets when they received cash payments. Instead of just eating rice and lentils, they started to eat a meal or two a week of meat or fish. These meals have much lower calories per dollar than meals of just rice and lentils, but these low-income people preferred to have a little variety in their meals, even if it meant sacrificing calories.
This is one reason, among many, that the current Official Poverty Measure seems to be out of whack with our conceptualization of poverty in America. The Official Poverty Measure was developed by Mollie Orshansky, an economist working in the Social Security Administration during the Johnson Administration. At the time, the average American family spent about one-third of their income on food. For this reason, Orshansky believed an adequate income would be the cost of a thrifty food plan times three. The federal government agreed with her and it was adopted as the poverty threshold.
We still use this threshold as our measure for poverty today, adjusted for changes in cost of living each year by the Department of Health and Human Services. There are some problems with this measure, though. First, families spend much less on food today than they did sixty years ago because food costs much less. On the other side, families spend much more on health care and housing, which relatively cost more than they did sixty years ago. The official poverty measure also does not have adjustments for cost of living, so a household of four in rural Oklahoma has the same federal poverty measure threshold for poverty of about $32,000 as a family of four living in a flat in San Francisco.
More fundamental than these problems of changes over time, though, is this assumption that people below the federal poverty line will starve. Yes, food insecurity exists in the United States, but starvation is nearly nonexistent in the country. The reality of the poverty line is that we are not trying to define how much is needed to literally survive. There are plenty of people who live for years under the federal poverty line. What we are trying to define is how much is needed to live a dignified life.
The allure of a measure based on food intake is that it gives us a veneer of scientific reasoning. We look at Maslow’s hierarchy of needs (which I could write a whole other blog post on) and our eyes go straight to the foundation: we need food to survive. So let us calculate the cost of food and use that as a basis for this idea of what income people need to not die.
What Banerjee and Duflo show us is how much of a sham this approach is. We’re not getting any closer to a “survival” estimate by saying how much it costs to buy a thrifty meal plan. Even during Orshansky’s time, the federal poverty measure did not capture what we meant by “poverty.”
An alternate poverty measure that gets closer to the concept we really have when we talk about poverty is the Supplemental Poverty Measure. The Supplemental Poverty Measure is an improvement on the Official Poverty Measure that makes adjustments for geography and estimates the impact of public benefits and taxes on poverty rates. Most importantly, it pegs the threshold to average spending levels, implying that poverty is defined by the inability to spend within a range of average spending in the country.
This is what we mean when we’re talking about poverty: does your ability to provide goods for yourself and your family fall far outside of the norms for your community? In our Ohio Poverty Measure, we make a similar estimate based on American Community Survey data, getting geographic precision that becomes even smaller than counties in populated parts of the state.
This line of reasoning, though, makes me even more interested in purely relative measures of poverty. A common measure of poverty, and the measure that is used by the Organization for Economic Co-operation and Development (OECD), is 50% of the median income. One thing I find attractive about the measure of 50% of the median income as the poverty measure is that it is incredibly easy to estimate anywhere. If you know what the median household income is for an area, all you need to do is divide it by two, then compare someone’s income to that. Then you know if they are in poverty or not.
Poverty will always be a contested concept. There are certain researchers at think tanks like the American Enterprise Institute who argue that poverty has almost completely disappeared in the United States, arguing that people consume so much more than they did during the Johnson era and that should be our measure for poverty. There are others who add up a list of items they believe make up household essentials and say if you can’t pay for these at market rates, then you are in poverty. These measures put the poverty rate at as high as half the U.S. population.
Both of these definitions strain the definition of “poverty,” giving us answers to the question of who is in poverty that do not fit with our intuitions about what defines poverty. Ultimately, poverty is a socially-defined term. Having it connected to society through a relative measure is the most rigorous way to define poverty in line with the reality of how people see it in their communities.