Gap in Life Expectancy Widens for the Nation

I don’t really have anything to say about this article in the New York Times, except for a comment on the graph:

0323-nat-HEALTH-web.jpg

It’s pretty good—excellent use of a common y-axis—but I think they should’ve put time on the x-axis and used three separate lines for “low status,” “intermediate status,” and “high status.” I prefer to put time on the x-axis where possible. Especially since the article’s focus is on the “gap,” which looks like a slope in the above graph. If time were on the x-axis and different lines were used for different-status groups, then the “gaps” really would be gaps between lines in the graph.

The article (by Robert Pear) had some interesting discussion about differences between rich and poor in smoking, risky behavior, access to health insurance, and adherence to treatment advice. The only thing that seemed silly to me was when someone was quoted as saying, “Middle-class and upper-income people have greater access to the huge amounts of health information on the Internet.” I mean, sure, the internet is great, but crediting it with years of life expectancy seems a bit too strong. The internet is a fine way to score some Vicodin, but it’s hard for me to imagine it can explain health disparities. (Maybe just a failure of my imagination, as usual).

P.S. Typo fixed; thanks, Garrett!

P.P.S. See Steve Kass’s comments below.

7 thoughts on “Gap in Life Expectancy Widens for the Nation

  1. I dunno… I can definitely see how more medical information at home could lead to (slightly) increased health outcomes. For example, being more knowledgeable about the signs of stroke or heart attack could lead to calling an ambulance earlier. Similarly, being aware (or being able to look up) the signs of cancer could lead a person to more rapidly seek diagnosis and treatment. For less life threatening diseases, being able to read up on (medically significant) home remedies could lead a person to heal themselves without going to a Drs. office and potentially catch something worse in the waiting room — though this reasoning is probably a bit more far fetched.

  2. I like the visual presentation. The main point that I immediately see is that, on average, the "most deprived" person of 1998-2000 had essentially the equivalent life expectancy of the "least deprived" person of 1980-1982.

    This is progress. Eliminating the remaining ~ 4 year gap should be secondary to protecting the genesis of such continued progress.

  3. Before speculating as to cause, I'd like to know what it actually means. My first thought was to wonder how they handle the fact that people's social status can obviously change during their life. Looking at the article, though, things seem even worse. They say that the researchers "developed an index to measure social and economic conditions in every county, using census data on education, income, poverty, housing and other factors." So
    the data doesn't pertain to individuals at all, but to counties. In addition to the general problems of such "ecological" (terrible term) studies, there's the problem that people can move during their lives…

  4. Radford,

    Yes, I agree. There's also a larger issue–not exactly a problem, but an issue–that "life expectancy" is a derived quantity based on some model. If you're talking about life expectancy for someone born in 2000, this obviously is not directly observable. They can't really go into this in the article, but I always wonder about how they put these estimates together.

    Alex,

    It is actually noted in the article that "life expectancy was higher for the most affluent in 1980 than for the most deprived group in 2000."

    Michael,

    Maybe you're right. I was just envisioning all the ways the internet can mess people up, by having them try all sorts of wacky things that don't work. Remember, the data in this article stopped at 2000. I believe the argument that richer people have access to more information and better tools for filtering out the crap, but I just don't see the internet as being the explanation in this case. It struck me more as a throwaway comment that the reporter typed in without thinking much about.

  5. Here are a few observations, based in part on the paper "Widening socioeconomic inequalities in US life expectancy, 1980–2000" by Singh and Siahpush. The paper is a free download at <a>http://ije.oxfordjournals.org/cgi/reprint/35/4/96….

    Singh and Siahpush conclude that "Between 1980 and 2000, those in higher socioeconomic groups experienced larger gains in life expectancy than those in more-deprived groups," and the Times headline says the "Gap in Life Expectancy Widens [with respect to affluence]. While I agree with the conclusion, I think there's a simple explanation, and I don't think the relationship between life expectancy and affluence is changing as the press would suggest. Between 1980 and 2000, those in higher socioeconomic groups
    simply experienced larger gains in socioeconomic status than those in more-deprived groups. Life expectancy just tagged along for the ride (faster in rich counties, slower in poor ones).

    The Times article notes that "after 20 years, [life expectancy in] the lowest socioeconomic group lagged further behind the most affluent." Since "lowest" and "most affluent" are measured by decile, not absolutely, the relationship between one's (absolute) deprivation and life expectancy could have remained the same (or even grown less important). If the top line in the
    Times graph were stretched to the left and right to reflect increased socioeconomic disparity,
    with perhaps year 2000 decile 3 above year 1980 decile 1, and year 2000 decile 8 above year 1980 decile 10, the conclusion would have been that the gap remains, not that the gap has grown. I think this is what's going on: the horizontal units have changed in the past 20 years. A socioeconomic decile in 1980 was smaller than a socioeconomic decile in 2000, and this is why the slope in years/decile has increased in the past 20 years. By many measures, the gap between America's richest and poorest has grown in the last 20 years, and data in the paper's tables show that it's the case for a number of the factors used to calculate the deprivation index.

    The idea that at least much of what Singh and Siahpush observed is just "antirestriction of range" is also supported by the 1980, 1990, and 2000 correlation coefficients between socioeconomic decile and life expectancy at
    birth, which they report as 0.94, 0.96, and 0.98 respectively. The lower correlation in 1980
    could be explained entirely by a more restricted range of socioeconomic status then, and there could be no change in the magnitude of the relationship between absolute deprivation and life expectancy.

    Singh and Siahpush considered this issue, but they don't say much of help in their paper: I'm not sure whether they want to suggest that the increased income gap does not explain the full increased life expectancy gap when they say
    "Our analysis of temporal state-level data (not shown) indicates that the relationship of income inequality [the ratio of household income at the 90th percentile to that at the 10th percentile]
    with life expectancy at birth in the US, even after adjusting for differences in absolute income levels, has become steeper over time, with standardized regression coefficients varying from -0.65 in 1969–71 to -0.71 in 1979–81 to -0.74 in 1989–91." In any case, I don't see what an analysis of state-level data of a different relationship over a different range of time contributes. I do think they had enough data to
    have done some analysis to address the question of whether the life expectancy gap is widening faster or slower than the deprivation gap. They note among other things that the "life expectancy gap" diminished between 1930 and 1960, in constrast to what happened in the last few decades. That's not surprising if the only factor is the rich-poor gap, since that it grew smaller in the mid-20th century.

    There's more of interest in the Times graph and the Singh-Siahpush paper. The graph has a noticable "kink" at the sixth decile, and not surprisingly, so do some of the factors that make up the deprivation index (some in one direction, some in the other, of course). The sixth decile is out of line with neighboring deciles in
    that it is more like the first decile in black population, homicide, and infant mortality, and more like the tenth decile in urbanicity. Two of these factors affect the calculated life expectancy at birth directly. (I didn't hunt down whether those two are used in the deprivation index, but it wouldn't be meaningful to include them on both sides of the relationship being studied.)

    In any case, a look at this sixth-decile kink might illuminate some interesting relationships. It also points out the problem with using counties as a basis for the analysis. The nation's second-largest county, and one of its least homogeneous socioeconomically is in this decile. Cook County, Illinois in fact, makes up about a quarter of the entire decile. Chicago's inhomogeneity, or particular ways in which it is not appropriately a sixth-decile place, probably explains the kink some way or another. Life expectancy is lower than expected at the sixth decile, and it is higher than expected at the eighth, of which one-third of the population is Los Angeles County. Could weather affect life expectancy as much as deprivation?

  6. Hmm. To me this looks like an effect for women, but neither the comments here nor the NYT article discuss this.

    If I just saw the male graph, I would see lines that are parallel (perhaps with some transformation). It's the female lines that are particularly interesting. There's not as big a change from 1980 to 2000 (lines closer together), and the lines show a considerable increase in slope.

    There's very little gain for women in deprived deciles over 20 years.

Comments are closed.