Living in New York City is expensive. Everybody knows that. Goods and services cost 22 percent more than the national average. The average home price in the city is more than $500,000. The city's many millionaires may grace style and society pages, but the average household earns about $50,000 per year—and the Census Bureau estimates that about one in five New Yorkers lives in poverty.

But Christopher Wimer and his team of researchers think even that official poverty rate is too low. For the past two years, the Columbia Population Research Center has regularly called, emailed, and snail-mailed 2,300 New Yorkers to ask them about their finances. The researchers calculate that in 2012, almost 40 percent of New Yorkers experienced a persistent hardship, such as being unable to make the rent or afford a visit to the doctor's office.

By surveying the same people over and over again, the Columbia researchers hope to shed light on who is struggling to get by in one of America's costliest cities. Timothy Smeeding, former director of the Institute for Research on Poverty at the University of Wisconsin (Madison), calls measuring hardship and access to economic opportunity "the new frontier" for poverty research.

The old frontier was improving the way the federal poverty line is calculated—a challenge Smeeding remembers explaining to National Journal decades ago. The Census Bureau has calculated the poverty rate pretty much the same way since the early 1960s: Take the cost of a minimal diet, multiply it by three, and adjust for family size, composition, and age. The formula takes into account only government benefits issued in cash—leaving out benefits such as food stamps—and poverty thresholds don't adjust to reflect the local cost of living.

A few years ago, the Census Bureau started to issue an experimental poverty measure based on the recommendations of the National Academy of Sciences. It's called the Supplemental Poverty Measure, and it takes into account both cash and noncash benefits; defines minimal household expenditures as a function of food, clothing, shelter, and utilities; and factors in other required expenses, such as child support payments.

The SPM's estimate for the national poverty rate and the official rate aren't that different. But the SPM tells us much more about how government programs work. Without food stamps, for example, the poverty rate for children would have been 3 percentage points higher in 2012. Without Social Security, the poverty rate for the elderly would have been almost 40 percentage points higher.

At the state and local level, policymakers are developing SPM-like measures and using them to shape antipoverty policy. New York City started estimating such a measure in 2008, and the data have informed Mayor Bill de Blasio's antipoverty agenda. The most recent city report found that almost half of New Yorkers were living in poverty, or close to it.

The Columbia study tries to go even further than the city's measure—which relies on census survey data—by following the same people over a two-year period. Funded by the New York City-based Robin Hood Foundation, the project is called the Robin Hood Poverty Tracker.

"We know that falling one dollar above or below any given poverty line may not capture the full dimensions of disadvantage," says Wimer, codirector of the Center on Poverty and Social Poverty at Columbia University. "So we spend a lot of time asking about what we call material hardships."

The study began with a baseline phone survey that asked people in-depth questions about their personal finances, and mental and physical health. Questions such as: Over the past year, were you forced for financial reasons to stay in a shelter? Do you have any health problems that prevent you from working?

The researchers visited social-service agencies to recruit participants who don't have phones or home addresses, and who don't usually get counted. Every three months, they've followed up with additional questions. Survey findings are statistically adjusted to represent the city as a whole.

According to the researchers' own SPM-like poverty measure, New York City experienced a 23 percent poverty rate in 2012. The federal poverty rate for the city that year was 20 percent, and City Hall's rate was 21.4 percent. The full results of the Poverty Tracker's baseline survey have been published in a nifty online graphic.

The follow-up surveys, which have yet to be published, may shed light on how people cycle in and out of poverty over time. "It's not rocket science, but one of the strongest predictors of someone's need getting better and them getting all the help they need, is them actually seeking help," Wimer says. That could be assistance from a family member, a community group, or a social-service agency.

Some follow-up questions, designed in partnership with City Hall, zero in on recent policy changes. For example, one follow-up survey asked participants whether they've gone to work sick less often since New York City introduced a paid sick leave law.

The Robin Hood Foundation decided to support the survey with a multimillion-dollar grant because it wanted to know which poverty-alleviation services people really need—and, therefore, whether the foundation's money is being spent wisely. Surveys like these are expensive. But they can help policymakers make decisions, too.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.