Measuring Poverty the Way We Understand It

August 22, 2016

It’s a well-known fact that the official poverty measure we use doesn’t accurately tell us how many poor people live in the U.S. It was never supposed to be more than an initial baseline for War on Poverty programs.

But here we are, more than 50 years later, and it’s still the basis for setting income eligibility levels for most safety net programs, plus some others that fund services in low-income communities.

The Census Bureau has a better measure — prudently named the Supplemental Poverty Measure so as to not imply it will supplant the official measure. It’s still unfinished in some important ways, e.g., sufficient adjustments for differences housing costs.

But even a more refined measure would still produce poverty rates based on pre-tax income, plus the income equivalent of some near-cash benefits, e.g., SNAP (food stamps), and the refundable tax credits.

We’ve had other measures proposed. Some economists, for example, recommend a consumption-based measure — basically, what a household owns, spends and, in some cases, has through government spending.

A team at Columbia University, supported by the Robin Hood Foundation is using a different definition to track poverty in New York City. The aim, the Foundation says, is to “take a deeper, more realistic look at disadvantage.”

This, we see, involves several things. First, the researchers use Census data, its poverty thresholds, adjusted for local rental housing costs, and an adapted version of the SPM to measure poverty in the city.

Second, the team surveys a sample of households in all income brackets, asking questions to identify those that experience “material hardships, i.e., deprivations due to lack of money.

In other words, the project uses a concept of poverty more like the way we actually think about it — not how much money a family has, but whether it has enough to meet basic needs. The survey implicitly defines them as food, housing, utilities and medical and dental care.

A household that sometimes comes up short is said to experience “moderate material hardship.” The hardship becomes “severe” if a household experiences it chronically or acutely — for example, is so short of money for housing that it has to move to a shelter.

The hardship concept isn’t altogether novel here in the U.S. For example, a household that has very low food security, as the U.S. Department of Agriculture defines it, can’t always afford to buy enough food for everyone to have enough to eat.

The Census Bureau’s Survey of Income and Program Participation includes a section it classifies as “adult well-being.” The questions are actually about the household’s well-being, including food security, the conditions of the housing it lives in and whether it can afford housing-related expenses.

SIPP uses the same sample of households repeatedly — some for as long as four years. It’s similar, in this way, to the third aspect of the “deeper, more realistic look” that gives the Robin Hood Foundation project its name.

It too tracks the select material hardships over time by surveying the same households for several years running. But it’s focused on them, rather than participation in programs that compensate for lack of enough income.

The survey does, however, include some questions about receipt of public benefits as part of its effort to localize the SPM. And it asks what I infer are half a dozen questions about requests for help and responses. (I’ve looked in vain for a copy of the survey itself.)

The project aims ultimately to capture the dynamics of poverty — and certain related material hardships. Understanding them, the partners say, will better equip policymakers, nonprofits and private-sector donors to develop solutions.

We’ve had three reports thus far. Each slices and dices the survey results differently. Here’s a handful of things we learn about the poverty/hardship dynamics — none, I suspect, peculiar to New York, except at the data level.

Far more New Yorkers experienced at least one material hardship than fell below the relevant poverty line. Far more, in fact, experienced at least one severe hardship. Poverty was nevertheless a major factor, as you’d guess.

Very few New Yorkers lived in poverty both when first surveyed and a year later. Slightly more fell into or rose out of it, presumably by losing or gaining income, certain cash-equivalent benefits or both.

Material hardship seems more persistent. Of the eye-popping 48% identified as experiencing at least one, nearly half still experienced it — or maybe some other — a year later, if not longer. (The first survey, as reported, didn’t ask how long respondents had already experienced the hardships covered.)

Not all New Yorkers who suffered a material hardship sought help. But majorities did — more for some needs than others. Very few said they got all the help they needed — at least, within the short timeframe the survey then covered.

Help they may have received somewhat later didn’t do all that much good. Slightly less than a third no longer experienced the hardship — a troubling finding for the project’s target audience.

Help did even less to lift households out of poverty, though the survey asks about help that could have — finding a job, for example, or getting public benefits. Only 2% more households that said they got no help remained poor than those that said they got all the help they needed.

On the other hand, getting timely help apparently averted poverty for a higher percent of those that swiftly sought it and said they got all needed. This may include people who sought help with a health problem, rather than a hardship that a health problem may have caused.*

Some of the dynamics Poverty Tracker reports aren’t altogether new. The Census Bureau, for example, has used its survey data to follow near-poor people, including their lapses into full-blown poverty.

A study by an economist at the University of California, Davis tracked transitions in and out of poverty, linked to factors that help explain them. The Urban Institute cites many other studies that have parsed the in-and-out (and back in) dynamics.

But we don’t, to my knowledge, have anything as expansive for material hardship. Some other countries are far ahead of us, as a complex study by the OECD (Organisation for Economic Cooperation and Development) indicates.

The authors say what seems a suitable conclusion here. “Poverty is a complex issue, and a variety of approaches are required for its measurement and analysis.”

* The survey asks separate questions about health problems. These are certainly a disadvantage — the overall scope of the project. But analyses thus far have not linked them to material hardships, though one did look at the correlation to poverty.

 

 

 


Who Should Decide What Poverty Is?

May 20, 2015

Let’s step back for a moment — oh, lets — from all the budget and other hot-button issues that will make life better or worse for people in poverty here in the U.S. Let’s consider how we decide who those people are.

As I suppose you know, we decide, for official purposes, by using a measure developed more than 50 years ago. This is the measure that becomes the basis for deciding who is poor enough to qualify for most of our major safety-net benefits.

Knowing it’s outdated — and was crude from the get-go — the Census Bureau has developed a “supplemental” measure, which some other analysts now use. Though more complex and sophisticated than the official measure, it still reflects needs experts have decided are essential, e.g., food, shelter and utilities, clothing, health insurance.

This, economist Stewart Lansley and coauthor Joanna Mack say, is a technocratic way of going at what’s essentially a philosophical question: what it means to be poor. What if we instead asked everyday people what they think necessary for an acceptable standard of living in our society?

The team ought to know because that’s what they’ve done for Great Britain, though thus far only as one of several alternatives to the official measure there. That measure, like ours, uses a straightforward income threshold. But unlike ours, it bases the threshold on median household income.

Below 60% of whatever the median happens to be at any given time means a household is officially poor. So the measure is relative, as it also is in other European Union countries, plus some additional countries in the OECD.

The threshold, however, still reflects a line experts and policymakers have drawn — in this case, to identify people whose resources are “so seriously below those commanded by the average individual or family that they are, in effect, excluded from ordinary living patterns, customs and activities.”

Lansley and Mack advocate a “consensual” poverty measure. It’s consensual in that it’s based on surveys that ask the public to identify items they think are necessities not merely for survival, but for living in their society.

So the surveys include not only food, “damp free” housing and the like, but some of those amenities the far-right Heritage Foundation cites to trash on our poverty measure — and our public benefits programs. The survey, in fact, goes beyond goods and services of any sort to include “social activities” no one should have to do without.

The researchers then take items and activities a majority of respondents have chosen as necessities of life. Adults who lack three or more fall into the poverty group, as do children who lack at least two. (Basing the counts on multiple lacks is intended to exclude adults who don’t have — or engage in — one thing or the other because they’ve chosen not to, even though they could afford it.)

A  basic premise here is that the deprivation we commonly view as poverty depends on cultural and social conditions. Whatever the type(s) or degree(s) of deprivation our poverty definition entails don’t properly apply everywhere and for always.

A second, related premise is that deprivation includes the experience of being marginalized due to the indirect consequences of not having enough income and/or sufficient public benefits. We see this in the fact that a majority of UK survey respondents view the ability to afford a school trip for one’s children as a necessity.

Beyond this, the method formally recognizes that “[p]overty is a value judgment,” as the inventor of our own official measure said.

So the question becomes who should make the value judgment — experts who define some set of minimal needs and the compute the costs or the public, whose views and everyday living activities set norms that, as Lansley and Mack have said, cause people who can’t afford to meet them “to be regarded as deprived and to feel deprived.”

The team argues that the public opinion methods is “the nearest we have to a democratic definition of poverty.” In the UK, at least, it’s a standard that has support from “all social groups,” they say, cutting across classes, age groups, gender and “very importantly, political affiliation.”

They view it hopefully as an approach that could “refocus the discussion” — heated debate actually — about the safety net and the government’s proper role in fighting poverty.

Whether such broad support for a poverty definition would make a difference in our public policies is, to my mind, doubtful. We know from polls, for example, that a large majority of American voters view SNAP (the food stamp program) as important for our country.

Has this protected the program from cuts, let alone produced the needed benefits boost and other changes I tend to harp on?

The notion of a poverty definition grounded in the public’s view of the necessities of life in our country is nevertheless intriguing. If nothing else, it gives us insights into rarely surfaced assumptions underlying our poverty measures.

That, in itself, is, I think, worthwhile as the debate over who’s truly poor, why and what’s appropriate for our government to do rages on.


More Earnings May Not Mean Less Hardship

August 20, 2014

Everyone with even a passing interest knows that the Census Bureau’s poverty thresholds are far too low — in part because they’re based on a long-outdated spending pattern.

The Urban Institute’s Molly Scott has a more fundamental objection. “All our national poverty statistics,” she says, “reflect economic poverty.” In other words, they measure total household income — both earnings and payments from programs like unemployment insurance and SSI.

The Census Bureau’s Supplemental Poverty Measure also includes the value of some near-cash benefits, e.g., SNAP (food stamps), housing subsidies, home energy assistance.

But Scott has something quite different in mind than a better version of our poverty measure. “The problem,” she says, is that “the arbitrary poverty line is a bad measure of material poverty, the amount of hardship people experience meeting their basic needs.”

People both above and below the poverty line often struggle to get through the month. The only difference between them is “the mix of resources they use and the costs associated with work,” Scott says.

She gives us two hypothetical single mothers in the District of Columbia. Both have two school-age children. They live next door to each other, so the rent on their apartments is the same. They both have minimum wage jobs. The difference is that one works part time, the other 60 hours a week.

The part-time mom’s family gets a larger SNAP benefit because the household’s income is lower. She’s somehow managed to get a housing voucher — again because her income is extremely low.

At the same time, her transportation costs are lower, presumably because she doesn’t work every day. And she doesn’t have to pay for child care because she works only while her kids are in school.

The end result is that her gross income is much lower, but her family is actually somewhat better off. Probably still facing struggles, but not actually in the hole, like the family headed by the other mom, whose earnings put them nearly $10,000 above the federal poverty line.

The moral of this story is that policymakers — and others — who champion work requirements and other strategies “to get people to work more” are often actually looking for more ways to minimize spending on programs that help poor people make ends meet.

We may spend less, but achieve little or nothing to alleviate hardship, as Scott’s time-and-a-half working mom’s situation shows.

Scott’s conclusion is more cautionary than prescriptive. “[W]e need to make sure our policies and programs do more than swap out subsidies for low-income wages that won’t change people’s quality of life.”

She refers to “real ladders of opportunity and supports along the way.” Which is all very well and good, but we need to do something about those low-wage jobs as well — and about supports for people who, for various reasons, can’t climb a ladder into a genuine living wage job.

For our single mothers in the District, that would be a job paying $32.95 an hour, assuming full-time, year round work. This would give them an annual income nearly three and a half times higher than the poverty line for their families — and about $1,950 more than the median for all households in D.C.

We’ve got bills in Congress that would raise the floor the “ladders of opportunity” rest on. There’s the long-stalled minimum wage increase, of course, but also a pair of bills that would, among other things, ensure that workers don’t get shorted if they’re sent home early or required to work for awhile and then again later because their employers go in for “just-in-time” scheduling.

We’ve got bills that would guarantee most workers some time off with pay so they could stay home when they were sick or for other compelling reasons, e.g., childbirth, an ill family member who needs care.

We’ve even now got a bill that would help ensure that some of the 26 million or so workers employed by federal contractors get paid what they earn.

And, of course, President Obama has used his pen — or as some Republicans say, disregarded the Constitution — to both raise their wage floor and better protect them against wage theft, as well as some other prohibited labor practices.

But the mighty pen can’t boost federal funding for child care — the second largest item in the living wage budget for our D.C. single-mother families. It can’t do anything about the cost of housing, which, as you might expect, is the largest.

And it’s highly doubtful Congress will either — any more than it will raise the minimum wage or pass all the other bills that would somewhat improve the financial circumstances of low-wage workers.

What’s more frustrating, in a way, is that there is no silver bullet — or round of silver bullets — ready for policymakers to fire, if they choose. Material poverty seems to me even more complex than plain vanilla economic poverty.

Which isn’t an argument for doing nothing. There’s a lot that can be done, much of which we already know. It is an argument, however, as Scott implies, for rejecting out of hand solutions that rely solely on getting more people into the workforce.

 

 


Who Is Poor? Depends Who’s Asking and Why.

May 30, 2013

New York Times columnist Thomas Edsall recently took a thoughtful tour through a vexatious question: Who is poor? Or more precisely, what measure should we use to decide?

As he indicates, this is more than an abstract issue for wonks to debate. It has immediate consequences for public policy decisions — most immediately perhaps, for the moves toward “reforming” Social Security and Medicare.

Because if fewer seniors are poor than the Census Bureau’s official poverty measure indicates, there’s a case to be made for reducing their benefits and funneling the savings to children and youth.

Not necessarily a good one. And not what Congressional Republicans want to do with the savings. But these are separate stories.

As I’ve said countless times, everyone knows the official measure is outdated — and was over-simple to begin with.

The Census Bureau’s Supplemental Poverty Measure aims for greater accuracy.

One the one hand, it takes account of more income sources and, on the other hand, certain necessary expenses, e.g., taxes, out-of-pocket medical costs, child care and work-related transportation.

Then it adds and subtracts from what families at the 33% spending level pay for four basic needs — food, clothing, shelter and utilities.

The end result is a significant increase in senior poverty and a smaller drop in the child poverty rate.

Right-wing conservatives have argued for some time that the official poverty rate is grossly misleading — and the SPM rate worse — because people they classify as poor enjoy a standard of living that’s a far cry from what we think of as genuine deprivation.

Look, says the Heritage Foundation, at all the “amenities” officially poor households have — refrigerators, TVs, cell phones, etc.

This is a crude — and misleading — version of a consumption-based poverty measure. The details of what economists have done to come up with such a measure — or in some cases, a framework for developing one — are far beyond my expertise.

The basic notion, however, is that what people consume or spend is, over time, a better measure of their material well-being (or lack of same) than their income, whether it includes the value of in-kind benefits or not.

And besides, people often understate their income on surveys like those the Census Bureau uses, says economist Bruce Meyer, a lead proponent of the consumption-based method.

He and James Sullivan, also an economist, argue, among other things, that both the official poverty measure and the SPM overstate poverty among the elderly, mainly because seniors are more likely than others to be spending out of savings and not spending on some big-ticket items because they own them outright, e.g., homes and cars.

When Meyer and Sullivan use their method, the poverty rate for seniors becomes 3.2% in 2010, as compared to 9% according to the official measure and 15.9% according to the SPM.

Shawn Fremstad at the Center for Economic and Policy Research has proposed an even more complex framework. The object, if I understand it correctly, would be to capture both a lack of basic economic security and deprivation — or hardship — based on our evolving standard of living.

Thus, the “amenities” that the Heritage Foundation cites as proof that most officially-poor Americans aren’t poor become necessities when experts and we, the general public, view them as such.

This hearkens back to a very old idea. Adam Smith wrote, in 1776, that he understood “necessaries” as “not only the commodities necessary for the support of life, but whatever the custom of the country renders it indecent for creditable people, even the poorest, to be without.”

James Cassidy, among others, has used it to argue for a relative definition of poverty — somewhat like the SPM, but with a far higher poverty threshold.

Cassidy suggests half the median household income. This could be adjusted for different household sizes, he says, as the Census Bureau already does for its poverty thresholds.

The OECD, an organization consisting mainly of highly-developed countries, does something of this sort for its poverty measure. It sets its poverty line at 50% of median income and then calculates the gap between that and the average income of people who fall below it.

Countries within the European Union also use a percent of median or average income. The percents vary, and some countries use multiple measures.

Ireland, for example, begins with the EU-wide at-risk of poverty line — 60% of median income. Then it calculates the percent of people below it who can’t afford 11 specific items and activities, e.g., new clothes, keeping the house warm.

In all these cases, poverty is conceived of as deprivation, relative to the country’s standard of living.

EU policy, however, pairs low income and/or material hardship with common consequences — not only practical disadvantages like unemployment, social exclusion and restricted access to “fundamental rights.”

This concern with inequality and social exclusion is the fundamental rationale for a measure that will always define some people as poor.

Exactly what the Heritage Foundation so ferociously objects to, since it tends to justify continuing government spending on anti-poverty and upward-mobility programs.

The basic issue, in other words, isn’t how to most accurately measure poverty. It’s how we want to define poverty itself — and to what end.

As Professor Robert Haveman wrote about 20 years ago, “Defining poverty is not just a matter of measuring things in the right way; it also requires fundamental social judgments, many of them with moral implications.”

This is why I think the measurement debate won’t get settled — and why it’s something we all have a stake in.