James Sullivan, associate professor of economics, recently completed a five-year study on poverty over the past five decades that utilized a consumption-based measure of poverty, unlike the official measure used by the U.S. government.
While the government's method finds the number of those below the poverty level to have increased greatly since the 1960's, Sullivan discovered there has been a noticeable decrease in poverty since then.
"When you measure poverty correctly over time you get a very different story [from the official measure]," Sullivan said. "There has been a sizable decline in poverty over the past five decades that you just don't see in the official poverty measure."
According to Sullivan's study "Winning the War: Poverty from the Great Society to the Great Recession," the measure used by the government to calculate the poverty level does not accurately reflect the number of citizens who live in poverty today.
The official measure of poverty takes the poverty line established in the 1960s, adjusts it for inflation and declares that anyone whose pre-tax income falls below that line lives in poverty, Sullivan said.
Instead, Sullivan employed a measure based on an individual's consumption rather than his or her income, he said.
"How do we capture the value of expenses in the Medicaid program [in the official measure]?" Sullivan said. "It's health insurance for the poor that ... frees up resources for those who would otherwise have purchased it. It's not captured in the official measure, but in consumption-based measure it would be."
The discrepancy in this measure often stems from the poverty measure only counting pre-tax income, Sullivan said. Unlike when the official measure was developed, the government now redistributes money through the tax program to low-income workers.
"It's not counted in the official poverty level because it's a tax program," he said. "But low-wage workers file taxes and if they're eligible for income tax credits they end up getting a check from the government. It's a substantial program for allocating refunds to the bottom distribution."
This incongruity between the two poverty measures exemplifies how the official poverty measure does not account for changes over time, Sullivan said.
Further, Sullivan said an income-based poverty measure cannot reflect variations in income that are unrelated to an individual's well-being. Income can fluctuate greatly for reasons that are unrelated to how well a person lives, he said.
"Think about a retired couple that has lots of assets, lives in a nice home and has a couple of nice cars," Sullivan said. "They're retired so they have no income, but they live off their assets so they live quite nicely. But because they have no money income they're classified as poor."
A consumption-based measure accounts for these gaps in the income-based measure, Sullivan said, and shows a clear decline in poverty.
Other factors skew the official measure of poverty besides post-tax income and wealth, Sullivan said. The price index used to adjust the poverty line for information is biased, he said.
"We're not accurately adjusting the threshold for time for inflation," he said. "We're over-adjusting and the threshold is artificially rising which will create more poor people."
Sullivan said the over-adjustment occurs because the inflation adjustment does not take into account two biases: a substitution bias and a new product bias.
The substitution bias occurs because when a product's price increases, the government adjusts the poverty line for that level of inflation, Sullivan said. However, sometimes an individual can buy a very similar product for significantly less money, thereby not affecting their overall well-being.
"If the price of Coke were to double, and that was all you consumed, do you need your income to double for you to stay just as well off?" Sullivan said. "If you substitute into Pepsi and it's much cheaper, so you only need an income to buy enough Pepsi to make you as well-off as you were."
A new product bias arises because it takes about 10 years for a new technological product to be included in the measure of prices of certain goods, Sullivan said. By the time the product is added, its price has likely decreased greatly since goods tend to be more expensive when they first appear on the market.
The faulty official poverty measure could have grave implications for poverty-related government policies, Sullivan said.
"Official numbers show poverty is higher today than it was in 1970," he said. "People have used that statistic to say we've lost the war on poverty. That has political implications, to say there's been a failure in [anti-poverty] programs and perhaps we should cut back on them."
Instead, Sullivan said the evidence that income tax credit lifts millions of individuals out of poverty demonstrates the United States is actually winning the war on poverty.
The incongruity must be fixed for the sake of those in the bottom distribution, he said.
"We use those measures to evaluate and design anti-poverty programs, and if we're using it to design policy it had better be a good measure," Sullivan said. "[An accurate measure] will help us make better policies, and in the grand scheme of things, hopefully better policies lead to decreases in poverty."