"(R)esearchers used available information and estimates about people’s historical incomes, along with their access to various social services and their eligibility for tax credits, to come up with a measure of poverty since the sixties that resembles the supplemental poverty measure." (Emphasis mine)I was immediately conflicted when I cursorily reviewed the data provided in the article below, and I have to say I find it incredibly hard to believe the conclusions they're drawing. There are several reasons for this, including the "estimates" used to support them. I'm reminded as well of two things I've been taught about using statistics to arrive at conclusions:
- Statistics can be used to manipulate the data to support just about any conclusion one wants to arrive at;
- Garbage (information) in; garbage (results) out.
This makes it challenging for me to pay attention or give a crap about what they're saying because I either don't understand it or I cannot connect it quickly to something meaningful in my life and/or my work. As such, I haven't dug as deeply into this as I should, and I'll be the first to admit this.
But what is of particular interest to me was the data used to get to the conclusions, and my Grad courses on statistics repeatedly warned us of the critical importance of well framed research questions, sample sizes, and the data set itself, so I did a little additional digging.
I discovered that back in 2010, "an Interagency Technical Working Group (which included representatives from the Bureau of Labor Statistics (BLS), the Census Bureau, the Economics and Statistics Administration, the Council of Economic Advisers, the U.S. Department of Health and Human Services, and OMB) issued a series of suggestions to the Census Bureau and BLS on how to develop a Supplemental Poverty Measure. Their suggestions drew on the recommendations of a 1995 National Academy of Science report and the extensive research on poverty measurement conducted over the past 15 years."
So, let me get this straight. The supplemental poverty measure they're using is constructed from the data obtained 1980 to 1995. This means the data itself is almost 20 years old, and of course is prior to the "pre-Bush" economic collapse. Its relevance seems pretty insignificant - at least to me - at this point in our current economic climate. I don't know about you, but my bullshit radar is sending off alarm bells that are waking up half my block as I write this.
I still don't know what to believe about this, but from an anecdotal perspective, I have a very hard time swallowing the results. Not arguing with what the data is showing, just arguing that the criteria used to define the relevant data may be horribly missing the mark.
December 10, 2013
Has America Become Less Poor?
In 2006, John Cassidy wrote in the magazine about the difficulties of measuring poverty. For years, the U.S. government didn’t try to calculate how many people were poor, Cassidy wrote. Then, in the nineteen-sixties, a government statistician named Mollie Orshansky came up with a way to determine what a family would need to not be impoverished. She based her calculations on the cost of food, which, at the time, made up a big proportion of household budgets.
To this day, the government uses a version of
that approach to measure poverty, even as families’ budgets have
transformed: food expenses make up a much smaller percentage, and
low-income families benefit from new social services and tax benefits
that the poverty measure also doesn’t account for. Starting in 2010, the
government created a second way to calculate poverty that accounts for
these factors—the supplemental poverty measure.
But that figure has been published only for the past couple of years, which means we haven’t been able to see how poverty has changed over time under the supplemental measure. According to the official measure, meanwhile, the proportion of poor people in the U.S. appears about the same as in the nineteen-sixties. It was fourteen per cent in 1967 and fifteen per cent in 2012. That has led to a lot of puzzlement over whether the War on Poverty of the sixties—which set in motion many of the anti-poverty programs that we rely on today—had any effect at all.
Now, a new study by a group of Columbia University researchers suggests what some policymakers have suspected for years: if you account for a fuller range of costs like clothing and shelter, and for government aid like food stamps, poverty has declined over the years—and by a lot. That is, even as people have limited wages and contend with high prices in the supermarket and elsewhere, the help from services like food stamps, and from tax benefits like the Earned Income Tax Credit, appears to have kept more people out of poverty.
The researchers used available information and estimates about people’s historical incomes, along with their access to various social services and their eligibility for tax credits, to come up with a measure of poverty since the sixties that resembles the supplemental poverty measure. Their finding: the proportion of poor people in the U.S. fell from twenty-six per cent in 1967 to sixteen per cent last year. (In this chart, “OPM” indicates the official poverty measurement, while “Anchored SPM” represents the researchers’ version of a supplemental measure, which is adjusted for inflation.)
Some politicians on the left might resist a measure that shows that poverty is much lower than it used to be: couldn’t this minimize the problem and make it harder to gain support for new anti-poverty programs? But it also may show that some anti-poverty programs of the past several decades appear to have achieved what they were meant for—which, one expects, should come as good news to everyone. Orshansky, the government statistician who came up with the first poverty measure, herself was open to other measurements. “If someone has a better approach, fine,” she said in 1999, according to Cassidy’s article. “I was working with what I had and with what I knew.”
Chart by Christopher Wimer, Liana Fox, Irv Garfinkel, Neeraj Kaushal, and Jane Waldfogel of Columbia University.
Photograph by Michael Dwyer/AP.
But that figure has been published only for the past couple of years, which means we haven’t been able to see how poverty has changed over time under the supplemental measure. According to the official measure, meanwhile, the proportion of poor people in the U.S. appears about the same as in the nineteen-sixties. It was fourteen per cent in 1967 and fifteen per cent in 2012. That has led to a lot of puzzlement over whether the War on Poverty of the sixties—which set in motion many of the anti-poverty programs that we rely on today—had any effect at all.
Now, a new study by a group of Columbia University researchers suggests what some policymakers have suspected for years: if you account for a fuller range of costs like clothing and shelter, and for government aid like food stamps, poverty has declined over the years—and by a lot. That is, even as people have limited wages and contend with high prices in the supermarket and elsewhere, the help from services like food stamps, and from tax benefits like the Earned Income Tax Credit, appears to have kept more people out of poverty.
The researchers used available information and estimates about people’s historical incomes, along with their access to various social services and their eligibility for tax credits, to come up with a measure of poverty since the sixties that resembles the supplemental poverty measure. Their finding: the proportion of poor people in the U.S. fell from twenty-six per cent in 1967 to sixteen per cent last year. (In this chart, “OPM” indicates the official poverty measurement, while “Anchored SPM” represents the researchers’ version of a supplemental measure, which is adjusted for inflation.)
Some politicians on the left might resist a measure that shows that poverty is much lower than it used to be: couldn’t this minimize the problem and make it harder to gain support for new anti-poverty programs? But it also may show that some anti-poverty programs of the past several decades appear to have achieved what they were meant for—which, one expects, should come as good news to everyone. Orshansky, the government statistician who came up with the first poverty measure, herself was open to other measurements. “If someone has a better approach, fine,” she said in 1999, according to Cassidy’s article. “I was working with what I had and with what I knew.”
Chart by Christopher Wimer, Liana Fox, Irv Garfinkel, Neeraj Kaushal, and Jane Waldfogel of Columbia University.
Photograph by Michael Dwyer/AP.
