If you had to guess which New Hampshire town has the highest poverty rate, which would you pick? Maybe the state’s major urban center, Manchester, or a struggling former milltown like Berlin or Franklin.
But would you guess that Durham, home to the University of New Hampshire, had the highest poverty rate of any community in the state?
According to the American Community Survey, the latest Census product to give us updated information for each town in New Hampshire, Durham’s poverty rate stands at 31.1 percent -- the highest in the state, and far above the statewide rate of 7.7 percent. But on its face, the Durham figure doesn’t make sense. The town’s residents are among the best educated in New Hampshire, it is home to one of the state’s largest employers (UNH) and its median income is above the statewide median. How can nearly one third of its residents live in poverty?
Part of the answer lies in terminology. The way the Census Bureau defines and measures poverty does not necessarily align with reality, especially when counting college students. And Durham has a lot of college students.
Let's see how that plays out in the numbers.
The Census Bureau (and all other federal agencies) measures a town’s poverty rate using something called a “poverty threshold” – a specific dollar amount that divides the poor from the not-poor. If your income falls below that threshold, you are classified as poor.
But most economists think the current definitions are badly in need of updating. The basic ways poverty thresholds are determined have undergone only relatively minor changes since they were established in 1964. Many of the recent proposals for changing the thresholds would result in an increase in the number of people classified as poor, and hence an increase in the poverty rate.
The strange poverty figures for Durham (and, to varying degree, other college towns) reflect one of the more obvious flaws in the definition. Children living in a household with an income level above the poverty threshold are assumed to have a share of that income -- even if they are a high school student working at McDonald's on weekends for poverty level wages.
But once those kids move out of the house -- and, say, go to college to live in off-campus housing -- they are no longer considered part of the family's household. Instead, they're counted as part of their college household, where they likely share an apartment with other students. And since most college students don't have full-time jobs or hold down low-paying part-time jobs, they wind up being classified as poor. This method of counting doesn't recognize that many students have their rents paid by their parents -- and maybe even some extra "spending money" to help support beer and pizza habits. Thus, it's a poor measure of actual poverty among college students.
(Students living in dormitories aren’t counted when it comes to poverty status. So towns that have colleges where most students live in dorms, such as Dartmouth College in Hanover, don’t show the same skewed poverty rate as a town like Durham, where many students live in off-campus housing.)
There is, however, a way to untangle this problem. Using data from the American Community Survey, we can calculate poverty by school enrollment, comparing non-college students to undergraduate and graduate students.
The table below compares this recalculated poverty rate with the official measure, as well as the difference between the two rates. Under the alternate calculation, Durham’s poverty rate drops from 31 percent to 6 percent. The rate for New Hampshire’s other college towns and surrounding communities also drops, in some cases significantly.
Since the poverty rate calculation in either definition above does not include college dormitory residents, this is really just a measure of students living off campus and their effect on the poverty rate in that town.
Is there a way to avoid such gross overstatements of poverty rates in college towns? The easy fix is to just change the definition of poverty, so that any person enrolled in college is excluded, whether that student lives in a dormitory or off campus. But that would change the poverty rate for a lot of cities and counties with high student populations, and an awful lot of federal money gets distributed based on those rates. So it becomes much more than just a matter of getting the statistics right – it becomes a matter of politics.