Measuring poverty

By Kathy G.

I wanted to follow up on Ezra's post yesterday which discussed poverty measures. As Ezra mentioned, the official measure of poverty in this country is highly problematic. When LBJ began the War on Poverty in the 60s, he quickly realized that he'd a need a benchmark by which he could measure the results. What they came up with was an index that had been developed by Mollie Orshansky, a statistician who worked in the Social Security Administration.

What Orshansky did was to take the cost of the so-called "thrifty food basket" (the cheapest food plan that could feed a family of four nutritiously) and then multiply that by three (because about one-third of the average family budget went to food). Since then, the poverty threshold has remained constant in real dollar terms; the government has simply taken the figure Orshansky came up with, and adjusted it for inflation. The fact that food is a lot cheaper now than it was then, and that food takes up much less of the average family's budget, has not been accounted for.

And that's not the only problem with the measure. For one thing, it's based on pretax income, so the impact of payroll taxes (which are much higher now than in the 60s) and the Earned Income Tax Credit  (which did not even exist back then) are not taken into account. Benefits like food stamps aren't factored in, either. In addition, items like child care, transportation, and medical costs, all of which can be quite substantial, are not considered, either. Another major item that the current measure does not include are regional differences in the cost of living, specifically housing costs, which tend to be the largest single component of the family budget and which vary greatly according to geography.

In the mid-90s, a special commission was appointed to consider developing a new poverty measure. They came up with some very sensible recommendations. First of all, they wanted the poverty index to reflect the family budget not just for food but for clothing, shelter, and other needs. They also wanted to use consumer expenditure data to determine the what share of the budget should go to each category of goods, and they wanted to update these figures every couple of years, so they could reflect what other Americans were spending. In addition, they thought the poverty threshold should be based on after-tax income, and they wanted to take into account benefits like food stamps, as well as regional housing costs, and out-of-pocket expenses for things like child care and medical treatment.

Their recommendations were thoughtful, responsible, and analytically consistent. And they went absolutely nowhere. One of my former professors was on this commission, and I asked him why the recommendations were not accepted. As I recall it, he said the reason was opposition to it by some members of Congress who worried that their districts would get less money for anti-poverty programs.  Even just a few people in Congress is enough to stop something like the proposed new poverty measure dead in its tracks, and President Clinton decided it wasn't worth spending scarce political capital on it.

I can't say I blame him. And I won't be surprised if President-to-be (I hope) Obama does nothing about it, either. Even for me, a wonky social policy type, instituting a new poverty measure is pretty damn low on my list of political priorities -- way below getting out of Iraq, universal health care, global warming, and probably several dozen other policies besides.

One positive result of the commission's work, though, is that the Census Bureau now releases not just the official poverty rate but also the poverty rate according to alternative measures; you can find those measures here. The alternative measures tend to show higher rates of poverty. For example, in 2006 (the most recent year for which poverty statistics are available), the official poverty rate was 12.3%, but according to the alternative measures it is as high as 14.2%.

You can also look at how the alternative measures affect the poverty rates in various demographic groups. Interestingly, in 2006, the alternative measures show higher poverty rates among whites and Asians than does the official measure, but a lower poverty rate among African-Americans (for Latinos, it's mixed). When regional factors are taken into account, the alternative measures show higher poverty in the East and West and lower poverty in the South and Midwest (as compared to the official measure). And compared to the official measure, the alternative measures show lower poverty rates among children but higher poverty rates among adults and the elderly.

Here are some fun facts from those 2006 poverty statistics. For old time's sake, I'll go with the official measure here:

-- Poverty rate among children, 17.4%; among adults 18 through 64, 10.8%; and among those 65 and over,  9.4%;

-- Poverty rate among whites, 10.3%; among Asians, 10.3%; among Latinos, 20.6%; and among African-Americans, 24.3%;

--Poverty rate among married couple families: 5.7%; among male-headed households, 13.8%; among female-headed households, 30.5%.

As for that last little statistic -- wow. Just wow. I knew female-headed households had a substantially greater poverty rate, but I had no idea it was that much higher. (And in case you were wondering, the alternative measures come up with different figures of course, but the overall picture of poverty rate according to family status does not change much).

One final point about this topic -- the question of how to measure poverty is actually a very deep one, and I've barely scratched the surface in this post. Should it be some absolute measure, based on a minimum standard for adequate food, clothing, shelter and the like? Or should it be relative -- how much income a household has compared to that of a typical household in a given society?

The OECD and the European Union, for example, define poverty as being half the median household income.(And by that standard, btw, the U.S. does very poorly; for the most recent OECD report on poverty I could find, which looked at poverty rates from 1995 to 2000, the poverty rate for the U.S. was 17.4%, as compared to the OECD average of 10.4%) The U.S. measure originally started out as being partly absolute (because it's based on the cost of food) and partly relative (because it's also based on how much of the average family's budget was spent on food). But at this point the measure is so analytically incoherent I don't know how I'd classify it.

Generally, left-of-center types hold the view that it's better to define poverty in relative rather than absolute terms, but we have some interesting allies in this debate. For example, in Wealth of Nations, Adam Smith makes the argument that "necessities" (or, in the term of the day, "necessaries"), should be understood not only as the bare minimum of goods required to sustain life, but also as those things society deems it unacceptable to be without:

By necessaries I understand not only the commodities which are indispensably necessary for the support of life, but what ever the customs of the country renders it indecent for creditable people, even the lowest order, to be without. A linen shirt, for example, is, strictly speaking, not a necessary of life. The Greeks and Romans lived, I suppose, very comfortably, though they had no linen. But in the present times, through the greater part of Europe, a creditable day-laborer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty which, it is presumed, nobody can well fall into, without extreme bad conduct. Custom, in the same manner, has rendered leather shoes a necessary of life in England.

The sociologist Peter Townsend (no, not this guy!) also stresses the relative dimension of poverty. He has argued that poverty should be defined as lacking the economic resources to "play the roles, participate in the relationships, and follow the customary behavior which is expected of them by virtue of their membership in society." By this standard, if you lack, say, a telephone, you would be considered poor by the standards of American society. This is similar to the ideas of the Nobel Prize winning economist Amartya Sen, who has conceptualized poverty as being deprived of "capabilities," where capabilities are defined as things such as life, health, political participation, and economic transactions.

I'm partial to poverty definitions like Townshend's and Sen's, because they get at the fact that some of the worst features of poverty are not just the physical deprivations, but the social, psychological, cultural, and intellectual deprivations as well. Now obviously, it would be more difficult to create a poverty measure based on their definitions, but it wouldn't be impossible -- the U.N.'s human development indices, for example, get at some of this stuff.

Advanced industrial can well afford to provide for the basic physical needs of all their citizens, but we should be doing far more than that. The ideal should be that, as much as is possible, each and every person has access to what she needs to make the most of her talents and to participate fully in society. That's why measures that focus on physical privation are not enough, and why the U.S. measure of poverty is so inadequate. But sad to say, it's unlikely to change anytime soon.

Comments

[DSchneider]

Thanks for these thoughtful definitions of poverty and the Census Bureau link to alternative formulas. I work with 11th graders (high school juniors) teaching US history, and we explore "poverty in the 20th century and today" as part of that. I teach them about the current formula and its shortcomings; this will help. I will use this piece in the next academic year. You might be interested to read these definitions, too, from an entry on Edge of the American West:
http://tinyurl.com/3nwxld

[geemoney]

This is one area where I think that policy wonk distinctions have hugely different impacts.

In Germany, where I currently live, it is astounding to me, born and bred in the USA, the amount of support which is made available to the unemployed, or to those who fall below the poverty line. Should one become unemployed here, the government will provide enough to allow one to pay the rent, buy food, pay for a cell phone, and buy food for a pet. While there are certainly people that exploit this (via intentional, constant unemployment), everyone with whom I have discussed this here (the Germans, anyway) wouldn't change things to make it harder to "take advantage" of the system.

I know that the effects are really only tangentially related to the point of the post, but it amazes me to no end how different the outlook is here on what the necessities are, and how willing people are to support this for those that need it. The freedom here from fear of catastrophe is a wonderful, wonderful thing to behold.

[ninja_zombie]

One possibility, and the one I support, is the following.

Determine the basket of goods we consider necessary for modern life.

Then give everyone a guaranteed *job* (40 hours/week) with the aforementioned basket of goods given as payment. The jobs should of course include unskilled government work (e.g., courthouse janitor), so it is nearly revenue neutral (the revenue difference is cost of janitor - cost of basket of goods). If we need extra jobs, we can make some up: free rickshaw rides in central park, for instance (admittedly not revenue neutral).

This improves information. The public now knows precisely how deprived the poor are: they get whatever is in the basket of goods. It is perfectly adjusted to the cost of living, since the basket of goods is fixed (even if cost of the goods is not). And it's matter for the public political process whether the poor get enough, or whether they need more.

They also know the poor are not cheating, since they work 40 hours/week, barely less than a normal person.

[Hattie]

Americans used to have a reputation for generosity. No more. We are stingy, judgmental and grudging where poor people are concerned.

[Tim Worstall]

The worst effect of the standard US poverty measure is that it grossly skews decision making. As you've noted, it doesn't count many of the things that are done to alleviate poverty (rent vouchers, food stamps, in fact, all non cash transfers, Medicaid) and most importantly it leaves out the EITC which, for better or worse, is the major anti-poverty program in the country.
Which makes it all too easy for people to point to the way in which hundreds of billions of dollars are spent on these poverty alleviation measures (which indeed they are) but that the poverty rate doesn't shift (because we don't, when calculating the poverty rate, count the things we are doing to alleviate poverty).

Coming from the other side we get rhetoric like John Edwards'....that 12% of the US is in poverty so we should raise the EITC, have more housing vouchers etc. But this won't change the poverty rate because we don't include these things in our calculation of the poverty rate (although they clearly do in fact alleviate poverty).

It's aa horribly flawed measure and one that needs replacing.

[Linden]

The fact is, many wage-earning families get housing vouchers, food stamps, Medicaid, and EITC, and still are below the official poverty line. And I'm defining poverty by the standard we're using currently. I've worked with hundreds of low-income clients, so I know whereof I speak.

[Liz]

Linden is right. I worked on a project that tried to quantify the number of people who are still below a basic family budget for their area even after receiving public work supports. We found that approximately 1 in 5 people around the country lacked the resources to make ends meet (this is using EPI's basic family budgets). You can read more about our findings at the link above or at www.bridgingthegaps.org. Feel free to contact me if you have any questions.