Absolute Poverty

Paul Krugman suggests, using calculations by Tim Smeeding (see table 2), that the United States is second-worst among affluent countries on absolute poverty. I don’t think that’s quite right.

Smeeding calculates absolute poverty rates as of 2000 using two poverty lines — the official U.S. line and 125% of that line. The U.K. is higher than the U.S. using either line. Krugman suggests that the U.K. rate may be lower than ours by now due to the Blair government’s anti-poverty initiatives. That is possible — we won’t know until more recent data are available — but the U.K. rate as of 2000 was significantly higher than ours, so the progress would need to have been dramatic.

Sweden and Finland have lower absolute poverty rates than the U.S. using one of Smeeding’s lines, but higher rates using the other.

According to my calculations, using the same Luxembourg Income Study data, five additional countries that Smeeding does not include — France, Australia, Ireland, Italy, and Spain — have higher absolute poverty than the U.S.

Here are my calculations. They’re from this paper. I use absolute income levels at the tenth percentile of the income distribution (so higher is better) rather than poverty rates. I prefer P10 incomes because poverty rates ignore the depth of poverty, but the two approaches yield very similar results.

This is not to suggest that we should be satisfied with our absolute poverty ranking. Given our nation’s economic wealth, incomes for Americans at the low end of the distribution are far lower than they could be. And as Krugman rightly points out, and I discuss in detail here and here, an exclusive focus on income overlooks the relevance of work hours and of public services such as health care, schooling, and child care for the well-being of the poor.

Addendum: Contra Tyler Cowen’s suggestion, the data for the U.S. used here do include the Earned Income Tax Credit and Food Stamps (though not Medicaid).

10 thoughts on “Absolute Poverty

  1. Pingback: The Ambrosini Critique » Blog Archive » Relative poverty?

  2. Lane, which of the following do these calculations include:

    *TANF (about $3100 per family)
    *housing assistance (about $5,400)
    *Medicaid (about $6,000 for a family of four)
    the State Children’s Health Insurance Program (SCHIP) (about $1,000 per child)
    *energy assistance (about $400)
    *the school lunch and breakfast programs (as much as $600 per child)
    *the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) (about $400 per person)

    the income of cohabitors and nonfamily household members

    (2002 estimates)

  3. KDeRosa:

    TANF: yes
    Energy assistance: yes
    WIC: yes
    Medicaid: no
    SCHIP: no
    Housing assistance: no
    School breakfast/lunch: no

    A valuable feature of the Luxembourg Income Study database is that these inclusions and exclusions are harmonized across countries as much as possible — though that doesn’t mean it’s perfect in this respect.


  4. Lane

    I have problems with some of these comparisons from LIS but for different reasons from the normal critique.

    This is that the poverty figures and calculations of income in inome surveys don’t really match up well with what we know about people’s benefit entitlements in different welfare states.

    It’s worthwhile looking at OECD figures on social assistance entitlements in different countries, which are available at

    You then open each spreadsheet and can go and look at the level of welfare benefits for different family types in different years up to 2004. The figures are in national currencies, but you can adjust these to US dollars (or whatever) by using Purchasing power parities also on the OECD website.

    I haven’t time to do this, but if you take the example of a lone parent with two children, you will see that in the US their basic welfare entitlement (when private income is equal to Zero) was $9,960. (This refers to Michigan).

    This doesn’t include Medicaid or SCHIP, but it also doesn’t include Medical assistance in any other country. It does take account of housing benefits where relevant. (See the pdf country descriptions on the same web address.)

    For the UK a similar family would receive about £13, 800 (and in Australia the same family would receive just over A$25,000.) These figures are much, much higher than for the US in absolute terms! bout twice as high roughly.

    And the same applies to virtually all of the household types used in the OECD analysis. (There is a conceptual problem with the OECD assumptions about the benefits received by couples without children in Australia and New Zealand, which I can explain if wanted.)

    If you are able to look at the chart in the main publication http://www.oecd.org/document/33/0,3343,en_2649_34637_39619553_1_1_1_1,00.html (but this requires a subscription) then you will find that relative to a 50% of median income poverty line then Australian and UK benefits (including housing benefits) are three times higher than in the US.

    Indeed after taking account of housing benefits, the UK and Australia consistently rank towards the top of OECD countries. The same is true of low income working families, as relative to median wages and also median incomes, the minimum wage in the UK and Australia are about twice as high as in the US (and in-work benefits are also much higher).

    The US comes consistently towards the bottom of OECD rankings in terms of relative generosity of basic entitlements for the non-aged. Where figures are available only Greece, Italy and Turkey have lower benefits – but Mexico and Korea are not included in this analysis, and are also likely to have low benefit entitlements.

    Now there are lots of reasons why entitlements may differ from recorded incomes for low income households. Benefits in the US for older people are probably more generous than in other Anglo countries (but I wouldn’t have thought so for the poor). There may be problems of take-up of entitlements, and the household at the 10th percentile in the US could be working, while in the other Anglo countries they may be on benefits. Alternatively they may be on social insurance benefits in the US and various European countries, whereas these don’t exist in Australia and are practically non-existent in the UK.

    However, I think we have a paradox – what people are entitled to in Australia and the UK is in relative and in absolute terms much more generous than what people are entitled to in the US.

  5. There’s another Smeeding paper from the LIS which gives us disposable incomes adjusted for PPP as a percentage of US median incomes.
    The result is (and Smeeding discusses why this isn’t perfect but is good enough) that the bottom 10% of US citizens get 37% of US median income and the bottom 10% of Finnish or Swedish get 37/38% of US median incomes.
    So the absolute living standards are the same. Of couse, the living standards of the 90% in hte US are higher than those of the 90% in either Finland and Sweden.
    One way (and given my evil classical liberalness) to look at this is that we might assume a trade off between high taxation and redistribution and growth in the general economy. “Too much” of the former and we get less of the latter. Now this might be justifiable if it makes the poor better off (a value judgement where anyone’s pick is as good as anyone else’s): but if the net result over time is no difference at all in hte living standards of the poor then the high tax, high redistribution route doesn’t seem to be the way to go.

  6. Worstall: “The result is (and Smeeding discusses why this isn’t perfect but is good enough) that the bottom 10% of US citizens get 37% of US median income and the bottom 10% of Finnish or Swedish get 37/38% of US median incomes. So the absolute living standards are the same.”

    This is a very dubious assertion. It is well known that PPP calculations are based on typical middle class households and cannot be used to accurately assess the purchasing power of poor households, which spend most of their income on food and necessities. This means that in order to compare the living standards of the poor across countries, you would have to develop PPP factors based on what poor people actually can afford, not based on what the middle class can afford.

    There are other reasons why absolute income comparisons may be misleading. The extent to which government subsidizes services that are vital to the poor has a big impact on their living standards. Four areas come immediately to mind:
    – Mass transit
    – Health care
    – Child care
    – Education

    Many among the US poor are not covered by Medicare and we all know that health care expenses can take a big toll on their income, which is less of a problem in most other developed countries. In a 2007 survey of health care experiences in seven countries (http://content.healthaffairs.org/cgi/content/abstract/26/6/w717), 36% of Americans responded that they had skipped a doctor’s visit, medication or diagnostic test at least once during the last year, and 19% reported having serious problems paying medical bills. Both figures are significantly higher in the US than in the other countries surveyed. Surely we have to agree that if you have to skip a doctor’s visit because you can’t afford it, you are poor.

    The same is true if for financial reasons you cannot attain your educational goals.

    Working low-income parents with children in the US often have to spend a high proportion of their income on child care.

    Finally, the cost of owning and maintaining a car can make a hole of several thousand dollars in a tight budget. In most places in the US, you just can’t survive without a car (or two for a family), let alone hold a job. In Europe, in general the car is less of a necessity. I have lived in Central Europe in cities of comparable size to where I live now. Not only did I never even think of owning a car, but I felt more mobile there without a car than I feel in the US despite owning a car, and I spent less on my transportation needs than I spend on my car here (not counting the plane tickets that I also need).

    For these reasons and from my own experience, I conclude there is strong evidence that you need a higher nominal income in the US to attain the same absolute living standard.

  7. Pingback: Poverty and immigration in the U.S. and abroad « Consider the Evidence

Leave a Reply