Skip to content
 

Senators and health care; also a discussion of pretty statistical graphics

Nate, Daniel, and I have an op-ed in the Times today, about senators’ positions and state-level opinion on health care. We write:

Lawmakers’ support for or opposition to reform generally has less to do with the views of their constituents and more to do with the issue of presidential popularity. . . .

For instance, Senator Blanche Lincoln, a Democrat who has been a less-than-strong supporter of the present health care bill, recently told The Times, “I am responsible to the people of Arkansas, and that is where I will take my direction.” But where does she look for her cue? Hers is a poor state whose voters support health care subsidies six percentage points more than the national average. On the other hand, Mr. Obama got just 40 percent of the vote there.

Likewise, in Louisiana, where the Annenberg surveys showed health care reform to be popular but where Mr. Obama is not, the Democrats are not assured of Mary Landrieu’s vote. . . .

Here’s our graph that makes this point:

senators.long-reduced.png

In putting together the op-ed, the art dept at the Times made some changes (with our guidance and approval). Here’s what they made:

senatorsnyt.jpg

Much nicer than our original, I have to say!

We also look at public opinion within states:

Using a statistical method called multilevel regression and post-stratification, we also mapped opinion on health care, breaking down voters by age, family income and state. We’re used to thinking about red states and blue states, but the geographic variation is dwarfed by the demographic patterns: younger, lower-income Americans strongly support increased government spending on health care, while elderly and well-off Americans are much less supportive of the idea.

And here are the maps that tell the story:

healthcare2004-StateAgeIncome.png

Again, the Times improved it (saving space slightly by combining the two highest income categories):

mapsnyt.jpg

The Times version is not just more attractive; it’s also easier to read, I think, in the sense of being more self-contained. (I still prefer our color scheme, though.)

Summary on the politics

Swing senators’ positions on health care are often presented in terms of worries about voter attitudes in the senators’ home states. Overall I don’t think this fits the data. Attitudes on health care vary more consistently by age and income than by state (compare to our graphs of ideology and partisanship), and constituents’ views on health care are not a strong predictor of senators’ stances.

Public opinion is certainly relevant to the health care debate, but not in the direct senator-follows-the-state way that it is sometimes imagined.

Summary on the graphics

I liked our graphs, but the Times versions are better. Our graphs took months of effort, but the Times versions were not immediate either. We had to go back and forth several times to get the clarity we all wanted. I’d like to think, though, that our effort was not wasted: by being able to make a bunch of graphs that were informative for us, we were able to home in on the story. At that point, the graphics professionals helped us to do better.

It’s tougher to make graphs for a newspaper than for a book, scholarly journal, or even a blog, I think. Even beyond the different audiences, a newspaper graph really has to be self-contained. In a book or article I can accompany the graph with a caption, and I make full use of captions to make each graph reasonably self-contained (to the benefit of people such as myself who jump from graph to graph when reading), and in a blog I can put whatever I want right below the graph. But in the newspaper, the graph really has to stand alone and with minimal captioning.

22 Comments

  1. Willem says:

    Andrew,

    what cost you months of effort? Not the creating of the graphs I hope, but the intellectual effort of creating the models needed for those graphics. Otherwise you should shore up your knowledge of creating graphs or start using other tools. Perhaps use the readers of your blog to try some graphs on / with.

    I like what Times did to your graphs, but they didn't do much more (imho) than work on appearance, not make the content more accessible.

  2. Willem says:

    Tip to self: Read back the history before you post.

    You do use your blog to try graphs on (including code) and don't like all the minicodes in 'par' in R. Getting up to speed there.

  3. Matt Frost says:

    It's nice to see that making graphs cute doesn't always destroy their content.

    But I'm having a hard time mapping the text to the graph. If "[Arkansas] is a poor state whose voters support health care subsidies six percentage points more than the national average," why does the "x" for Blanche Lincoln appear to fall at 4%? Am I misunderstanding the data? Were different surveys used?

  4. Andrew Gelman says:

    Matt: I'm not sure, maybe we misread the graph when writing the text?

  5. Matt Frost says:

    OK. IN that case, I understand the data, graphs, and overall thesis after all.

  6. Chris says:

    Why are percentages in the graph relative to the national average? Surely this creates the misleading impression that Lincoln is facing no electoral pressure back home?

    btw, how were these polls selected. this suggests AR is quite different:
    http://arkansasnews.com/2009/11/05/almost-half-of

  7. Komodo says:

    The Times version of the second graph has a typo: 30-34 rather than 30-44. So looks like the original is easier to read after all.

  8. Jinchi says:

    I'm astounded at the numbers with respect to age.

    The maps look almost identical between the 18-29, 30-44 and 45-64 age groups. Suddenly people hit 65, sign up for Medicare and want to slam the door shut behind them.

  9. Andrew Gelman says:

    Chris: You ask, why are percentages in the graph relative to the national average? We did it this way because our poll data were from 2000 and 2004, when the average support for increased federal health care spending was around 70%. When considering present-day opinions, it seemed more reasonable to extrapolate relative numbers than absolute numbers.

  10. John Transue says:

    Cross posted from themonkeycage.org because there's more discussion here:

    It was fun to read this in the paper this morning. It’s better to see it online because the figures are much easier to process in color than they were in the print edition.

    I’m not sure that the argument is correct, though. In 2004 health care was not as salient as it was in 2008. It could be that Obama’s 2008 margin of victory is a better proxy for 2009 health care attitudes that 2004 health care attitudes. So the Senators may have better information about their states than the 2004 data. It could be using Obama margin as a proxy or maybe Obama margin is correlated with some unobserved (by us) information that they have.

    Once the 2008 data come out, it will be easy to see whether this is true. Does anybody know when they’re going to release the 2008 Annenberg? I can’t wait!

    ANES 2008 probably can’t support state-level inference like this. I wonder if anybody had health care questions on the 2008 CCES that could answer this.

  11. Will says:

    I understand what you are saying about the limits on the data and need to use 2004 polling results, but I think this is highly misleading. The healthcare debate has changed completely since then and is much more specific now. While it is likely that a large percentage of Americans support the general idea of healthcare reform, a much smaller number approves of the manner in which this administration is going about it. I think this is a very different issue than what you are showing the graphs and may vary differently by state (i.e. a state in which a larger than average proportion support general healthcare reform could be less supportive than average of this particular version of reform).

    You single out Senator Lincoln and imply that she is not representing her constituents because in 2004 more Arkansans than average supported healthcare reform. However, recent polls show that a plurality of Arkansans oppose a bill with a public option. Therefore, more recent, more specific polls seem to imply that Senator Lincoln would be properly representing her constituents by voting against the bill (or not voting for cloture).

    It's possible I am misreading this or misinterpreting it, but I'd be very curious to hear your response.

  12. Abe says:

    I strongly prefer your colors. Leaving "neutral" white or blank is very easy to understand. The NYT uses a sort of purple as the midpoint, which makes it really difficult to see the transition from pro to against.

    I'd even argue that the purple looks more like red then blue/green to me, leading the chart to make health care reform look less popular than the data shows.

  13. Eduardo says:

    What the hell!? One the color scale, positive numbers are at the bottom, and negative numbers at the top!

    I like the R version more and more…

  14. Martyn says:

    For the maps, it looks like you used a diverging HCL colour palette from the colorspace package in R (and if you didn't, now you know). The theory behind this is explained in a very nice paper by Achim Zeileis, Kurt Hornik and Paul Murrell (Escaping RGBland: Selecting colors for statistical graphics, Computational Statistics and Data Analysis, 53:3259-3270, 2009). Having read the paper, I have to say the Times colour palette looks terrible and is a lot less informative.

  15. BC says:

    No surprise: the boomers have always been selfish bastards. For all the idealistic reasons they fought against the war in Vietnam, there was also the fact that they didn't want to get killed.

  16. gb says:

    I think their graphics are not better than yours.

    The first one threw away a lot of the interesting information, and the second one has a far worse color scheme and a mislabelling.

    And if they're going to collapse categories why not collapse the ages down to

  17. gb says:

    damnit. my "less than" symbol got munged by the html interpreter.

    my third sentence should have said:

    And if they're going to collapse categories why not collapse the ages down to "less than 65" and "65+" ?

  18. Elaine says:

    BC, you are wrong. Boomers were born from 1946-1964, so they currently range in age from 45-63. The selfish over 65 age group are NOT Boomers.

  19. tom says:

    the much biggerp oint seems to me that this seems an absurd proxy for obamacare for three reasons all realated to the fundamental problem that the quesiton is do you support increased government spending on heathcare . -this is simply not the same logically as suppuort for Obamacare-after all the obama administration claims the bill would be financhialy netural . You could could hold any cobmination of postions on that question and on "do you support obamacare

    some reasons why the numbers will differ include but are by not means confined to

    a)

    - given it's a very complicated issue with lots of factual disagreemtns on what the legislation will do even in legal terms , peoples opini0on of Obama is likely to have a big effect on whether they support it or not -arkansas is a state where not only did obama lose but he ran massivley behind party lines

    b) support more "money" is not the same as public option competing with private plans , regulations on insuance industry etc etc it may be people from arkansa are more sceptical of aspects of plan-perhpas even the restricions on healthcare spending it !

    c) perhpas noteworhty the bill cross subsidiesed aboritons and is opposed by pro lif- arkansas is a fairly strongly pro

  20. Aleks Jakulin says:

    The plot with x's and o's could be improved if you just drew an arrow indicating a temporal change.

    instead of

    o x

    use:

    *———>*

  21. Manoel Galdino says:

    Prof. Gelman,
    could you tell us what package did you used to produece the maps? sp, ggplot2, maps etc?

    Thanks

  22. Andrew Gelman says:

    We used maps(). You'll be able to do it all yourself when we release our mrp package (very soon!).