Skip to main content

In 2006 and 2008, a nonprofit executive named Donna Edwards ran against incumbent Congressman Al Wynn in Maryland's Fourth District, which includes portions of Prince George's County and Montgomery County.  In 2006, she narrowly lost, but in 2008, she easily won, in what is considered something of a landmark of "netroots" influence on elections.

Back at Swingstateproject, we were talking about the effects of the VRA, and I mentioned a book, David Canon's Race, Redistricting, and Representation.  Canon claimed that the white populations of majority-minority or VRA districts could exert a (sometimes decisive) influence by siding with one of several minority candidates in a contested primary.  Someone mentioned that the Edwards-Wynn primaries would be interesting case studies.  I decided I would study them.

My approach was pretty simple--I simply downloaded the precinct-by-precinct elections results for the two primaries at the Maryland Board of Elections and (laboriously) added in the 2010 precinct-by-precinct racial percentages for each of the 169 or so precincts.  (I say "or so" because, for some reason, the 2008 results include some precincts absent from the 2006 results).  I then ran a series of multivariate regressions using Edwards' percentage of the vote in each precinct and one or more of the racial variables.  Note that this approach of course does not necessarily mean a candidate does better or worse with voters of a particular racial group--it's all about areas.

Let's start with 2006.  Here are the results of the regression analysis program (I used Winpepi, which is apparently for epidemiologists, because it's free):

 Regression equation:
    Edwards = 0.460 + 0.003(White) - 0.001(Black)

   Variable     Coefficient      SE     Two-tailed P
    White          0.003      < 0.001      0.000
    Black          -0.001      0.000       0.013

   Standard error of estimate = 0.077

   R-squared (coefficient of determination) = 0.636
      Adjusted coefficient of determination = 0.631

Now, if you've had no formal training in statistics...that makes two of us.  However, here's what I believe this means.  On average, Edwards did better the whiter the precinct's population, and independently from that, she did worse the blacker the precinct's population (several precincts also had significant Asian or Hispanic populations).    The two-tailed P test means that both coefficients are statistically significant.  However, the R-squared value is low at .636, which means there's a lot of error going on.  

Now, it's a bit harder to graph three variables than two, and I'd like to include a picture of what's going on. so let's run the same analysis using just Edwards' share in each precinct and the white population, since the two-tailed P is lower in that case:

Simple linear regression:
       Equation:  Edwards = a + (b x White)
                  Edwards = 0.379 + (0.004 x White)
        a = 0.379 (S.E.: 0.008)
        b = 0.004 (S.E.: 0.000; 95% C.I.: 0.004 to 0.005)
            P = 0.000 (for difference from zero)
       Coefficient of determination (r-squared) = 0.622
          Adjusted coefficient of determination = 0.620
       Standard error of estimate = 0.078

Note that the R-squared has only decreased a little bit.  My impression is that this means that we haven't lost too much accuracy by looking at just two variables.  Here is the scatterplot: Each dot is a precinct, Edwards' 2006 share of the vote is on the Y-axis, and the white % of the population is on the X-axis.

Note why a linear regression might not have the best fit--one line with a pretty sharp slope seems to fit the cluster of heavily-nonwhite precincts on the left, while after that, Edwards hovers around 60% throughout.  Even the equation Edwards=.379+(.004 x White) would predict Edwards got a majority of the vote in any precinct with more than, say, a 30% White population.

Another way to look at this is by county: Just looking at the numbers, Edwards seemed to pretty much romp in Montgomery County (whiter overall) while her performance in Prince George's County was much more varied.  Here's the above graph, just in Montgomery County:

The linear regression says:

Simple linear regression:
       Equation:  Edwards = a + (b x White)
                  Edwards = 0.478 + (0.002 x White)
        a = 0.478 (S.E.: 0.027)
        b = 0.002 (S.E.: 0.001; 95% C.I.: 0.001 to 0.003)
            P = 0.000 [ 3.1E-5 ] (for difference from zero)
       Coefficient of determination (r-squared) = 0.226
          Adjusted coefficient of determination = 0.214

And here it is in Prince George's County:

To try to get a better fit, I looked at PG with both white and black percentages thrown in, but no luck:

  Regression equation:
    Edwards = 0.316 + 0.008(White) + 0.000(Black)

   Variable     Coefficient      SE     Two-tailed P
    White          0.008      < 0.001      0.000
    Black          0.000       0.000       0.339

   Standard error of estimate = 0.064

   R-squared (coefficient of determination) = 0.467
      Adjusted coefficient of determination = 0.456

You can see that the graphs are pretty well-separated, and the R-squared--the quality of the fit--for both is quite poor.  So the relatively not-bad R-squared from before might have just been an artifact of some kind of county-level divide.  However, note that "White" remains statistically significant in both counties.

I was going to talk about 2008 too, but it's not very interesting.

Edwards won nearly every precinct in the district on her way to a landslide victory.

Now for some questions: First, does anyone know a good way to estimate income or education levels by precinct?  Is there one, or would the error be too high for something like that?  I was thinking of using ACS estimates by census tract, but they're not the same divisions.  Overall, this area is one of the wealthiest and best-educated in the county, and I'd be very interested to know if education or income levels provided better fits, especially within the counties--or if racial percentages were still statistically significant after accounting for education and income levels.

Second, of course, I'd be interested to hear the reaction from people who know this election well--I deliberately kept this post entirely quantitative.  

Finally, I'm not sure if anyone else has done this analysis--a quick Google of "Edwards Wynn precinct regression" didn't seem to show anything, nor did a more detailed googling just now.  This fellow has done some similar stuff with 2000 census data for other MontCo races:

It was as much an exercise for me to learn how to use the software and such as anything else, but I thought it might have some interest.  If nothing else, it gave me a chance to run a regression analysis, like an actual scientist instead of the mathematician I am.  Anyway, looking forward to your reactions.  

Your Email has been sent.
You must add at least one tag to this diary before publishing it.

Add keywords that describe this diary. Separate multiple keywords with commas.
Tagging tips - Search For Tags - Browse For Tags


More Tagging tips:

A tag is a way to search for this diary. If someone is searching for "Barack Obama," is this a diary they'd be trying to find?

Use a person's full name, without any title. Senator Obama may become President Obama, and Michelle Obama might run for office.

If your diary covers an election or elected official, use election tags, which are generally the state abbreviation followed by the office. CA-01 is the first district House seat. CA-Sen covers both senate races. NY-GOV covers the New York governor's race.

Tags do not compound: that is, "education reform" is a completely different tag from "education". A tag like "reform" alone is probably not meaningful.

Consider if one or more of these tags fits your diary: Civil Rights, Community, Congress, Culture, Economy, Education, Elections, Energy, Environment, Health Care, International, Labor, Law, Media, Meta, National Security, Science, Transportation, or White House. If your diary is specific to a state, consider adding the state (California, Texas, etc). Keep in mind, though, that there are many wonderful and important diaries that don't fit in any of these tags. Don't worry if yours doesn't.

You can add a private note to this diary when hotlisting it:
Are you sure you want to remove this diary from your hotlist?
Are you sure you want to remove your recommendation? You can only recommend a diary once, so you will not be able to re-recommend it afterwards.
Rescue this diary, and add a note:
Are you sure you want to remove this diary from Rescue?
Choose where to republish this diary. The diary will be added to the queue for that group. Publish it from the queue to make it appear.

You must be a member of a group to use this feature.

Add a quick update to your diary without changing the diary itself:
Are you sure you want to remove this diary?
(The diary will be removed from the site and returned to your drafts for further editing.)
(The diary will be removed.)
Are you sure you want to save these changes to the published diary?

Comment Preferences

  •  Tip Jar (13+ / 0-)

    25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

    by Xenocrypt on Mon May 09, 2011 at 03:30:15 PM PDT

  •  guess (5+ / 0-)

    precincts that showed up in 2008 but not in 2006 may be new.  The county I was in had to create new ones when some current ones became too big.

    "Intolerance is something which belongs to the religions we have rejected." - J.J. Rousseau -6.38, -4.15

    by James Allen on Mon May 09, 2011 at 03:34:23 PM PDT

    •  Interesting. (0+ / 0-)

      I think there was growth in the district, absorbing people leaving D.C.  It needs to shed about 7,400 people but that still does mean a fair amount of actual growth.  Hoyer's district is the biggest, though.  

      25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

      by Xenocrypt on Mon May 09, 2011 at 07:17:37 PM PDT

      [ Parent ]

  •  Income/Education (3+ / 0-)
    Recommended by:
    Xenocrypt, Lujane, Odysseus

    I agree that this is something that definitely needs to be looked at. MD-04 has one of the largest (maybe the largest?) black middle class populations in the US, so it would be very interesting to see if poor heavily black areas voted differently from middle/upper class heavily black areas. The best way to do this would probably be by town. Just from a quick look at Dave's app, it looks like there aren't any precincts that cross town lines anywhere. Average income levels by town can easily be found in the census data, so you could add up the total vote from all precincts in each town and run a regression on that. It's a less precise way to do it than going precinct-by-precinct, but I don't know of any way to get income data for each precinct.

    You could also try contacting the Maryland State Board of Elections and seeing if they have precinct income data on file somewhere. I doubt that they do, but it's worth a shot.

    Male, VA-08, Born CA-36, SwingStateProject expat

    by drobertson on Mon May 09, 2011 at 04:04:08 PM PDT

    •  Thanks (1+ / 0-)
      Recommended by:

      If I figure out how to do it, I'll add it as an update or another diary.  I assume you mean circa-2000 income level?  They don't ask about it in the census anymore, I don't think.

      25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

      by Xenocrypt on Mon May 09, 2011 at 04:07:00 PM PDT

      [ Parent ]

      •  Either 2000 data or ACS data (2+ / 0-)
        Recommended by:
        Lujane, Xenocrypt

        The actual census doesn't ask about income anymore, but the American Community Survey does. There are pros and cons to both. The ACS would give you the most timely data as it was taken mid-decade, but because it is a much smaller sample than the full census the margin of error on it is much higher, particularly when you are talking about small individual towns. The 2000 census wouldn't be that bad. It's almost as close to 2006 as the 2010 census is, and unlike the 2010 census it wouldn't show the effects of the Great Recession which probably changed income levels significantly.

        Male, VA-08, Born CA-36, SwingStateProject expat

        by drobertson on Mon May 09, 2011 at 04:15:22 PM PDT

        [ Parent ]

  •  There were reasons for this (5+ / 0-)

    Prince George's makes up about 65 percent of the district and Montgomery County makes up 35 in terms of votes.

    in 06, Donna raised and spent only about $200,000.  She did well in MoCo because Rep. Wynn had basically ignored that part of the district.  Furthermore, Montgomery voters were angered by his vote for the Iraq War.  Meanwhile, Wynn was campaigning for other Democrats in Prince George's.  His machine basically carried that part of the district.

    In 08, Donna raised a lot more money and had key institutional support from groups like SEIU and the Sierra Club.  The advertising from both the campaign and outside groups basically eliminated the advantage Wynn had from his machine.  Also, her message was much cleaner than Wynn's.  They were two very different elections.

    Disclaimer-my company was the direct mail firm for Donna Edwards in the 2008 campaign.

    by DavidatTruBlu on Mon May 09, 2011 at 04:08:49 PM PDT

  •  I'm going to be honest, I didn't really understand (2+ / 0-)
    Recommended by:
    Lujane, terjeanderson

    this. But good on you for this analysis, which clearly took some work!

    They tortured people to get false confessions to fraudulently justify our invading Iraq.

    by Ponder Stibbons on Mon May 09, 2011 at 04:18:08 PM PDT

  •  I live in md-04 (1+ / 0-)
    Recommended by:

    The republican usually gets about 15% of the vote. Donna is congresswoman for life if she wants to be.

  •  Outlier (8+ / 0-)

    This race was a little different from other races where the minority white vote bloc voted resulting in an Dem primary defeat for a black incumbent. It wasn't moderate-conservative voters putting a more centrist Democrat was progressives turning a more centrist Dem out.

    There were multiple reasons why MoCo voters didn't like Wynn. Though even the whitest precincts in the county have proven more than willing to support black candidates to all levels of offices...they are turned off by PG County and its dysfunctional machine politics, of which Al Wynn is a product. He voted well to the right of his constituents, black and white alike, voting for the Iraq War, the GOP bankruptcy bill of 05, and against much of the progressive agenda.  
    (Which makes me curious if Wynn fared any better in the Upcounty areas of the 4th, which might appreciate some of those more conservative votes. They correlate pretty well with the whitest MoCo districts in the 4th, so it can be figured out some. I think that's why the correlation line is not quite as straight as one might have guessed there. Contrast with white residents of northern PG in the 4th, who are overwhelmingly liberal and extremely pro-Edwards.)

    And name rec in his case only went so far... Wynn had been an incumbent for years but was one of the more anonymous members of the House, having a low profile and no significant accomplishments to his name. He was particularly unknown in most of the MoCo portion of the 4th, as much of that territory was given to him in 2002.
    And he had never really had to campaign much..he had a safe district as far as general elections were concerned, and in over a decade had not been seriously challenged.

    So this is a somewhat different case than Cynthia McKinney or Earl Hilliard.  

    Stuck Between Stations : Thoughts from a bottomless pool of useless information.

    by Answer Guy on Mon May 09, 2011 at 04:49:13 PM PDT

    •  He took on Connie Morella's best precincts, no? (2+ / 0-)
      Recommended by:
      Xenocrypt, terjeanderson

      Ok, so I read the polls.

      by andgarden on Mon May 09, 2011 at 07:03:17 PM PDT

      [ Parent ]

      •  Without checking, sounds about right (nt) (0+ / 0-)

        25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

        by Xenocrypt on Mon May 09, 2011 at 07:14:42 PM PDT

        [ Parent ]

      •  Most of them (1+ / 0-)
        Recommended by:

        A few precincts of northernmost Montgomery are in the 6th; these are in the Damascus area, a mostly rural (though growing) area that is one of the few parts of the county to vote Republican with any regularity.

        MD-4 did take some parts of MD-8, such as Olney and about half of Germantown, that voted for Connie in the 80s and 90s and replaced them with precincts in the Silver Spring/Takoma Park area and a bit of Prince George's, areas that are one-sidedly Democratic and where a lot of voters didn't know Connie, leading to her defeat by Chris Van Hollen.  Since then, however, many of what had been Morella's best precincts have grown more Democratic (they mostly already leaned that way, and certainly wouldn't vote for a more conservative Republican; Connie was a special case.)

        •  Yeah (0+ / 0-)

          I haven't ran the income/education numbers or whatever, but, knowing what little I know, it's certainly plausible to me that there might a non-trivial number of Morella-Edwards voters.  Maybe people who "vote the candidate, not the party".

          25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

          by Xenocrypt on Tue May 10, 2011 at 08:52:38 AM PDT

          [ Parent ]

    •  The book was written earlier, actually (2+ / 0-)
      Recommended by:
      Odysseus, terjeanderson

      The examples are more like Mel Watt and Bobby Scott vs. Bennie Thompson and Maxine Waters.   So yes, it is different, but the question I had was whether the basic mechanism--the white minority of a district having a significant effect in a contested primary between black candidates--applied.  MD-04 has an unusually liberal white population, so the nature of their effect--to the extent they had one--would be different than in many other districts.  

      And I did suspect the point about machine vs. anti-machine voting, as I said.  I think we're all familiar with elections where (usually poor) minority areas get pretty machine-dominated.  This is where some information about education or income might improve things.  Of course, machine vs. anti-machine doesn't always load into the ideological scale very nicely, but it surely did in this case.

      Your local information is also appreciated--I didn't know how the district had changed over time.  It makes sense, though, since Van Hollen notoriously grabbed a slice of Prince George's County for himself.  

      25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

      by Xenocrypt on Mon May 09, 2011 at 07:09:31 PM PDT

      [ Parent ]

  •  Interesting. What happens if you add a county (2+ / 0-)
    Recommended by:
    Xenocrypt, terjeanderson

    dummy variable?

    •  Was wondering that myself (0+ / 0-)

      Too lazy to do it but now I did. Interestingly, County has a high p-squared and White remains the most statistically significant variable:

      Regression equation:
          Edwards = 0.471 - 0.034(County) + 0.003(White)
             - 0.001(Black)

         Variable     Coefficient      SE     Two-tailed P
          County         -0.034      0.024       0.152
          White          0.003      < 0.001      0.000
          Black          -0.001      0.000       0.105

         Standard error of estimate = 0.077

         R-squared (coefficient of determination) = 0.640
            Adjusted coefficient of determination = 0.634

      R-squared doesn't improve that much from White+Black, either.  I used 1 for PG and 0 for MontCo.

      25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

      by Xenocrypt on Tue May 10, 2011 at 07:43:55 AM PDT

      [ Parent ]

      •  Oh, okay. One more question-- what is the standard (1+ / 0-)
        Recommended by:

        deviation of Edwards? Because one reason this works so well (just race accounting for 63% of variance is pretty good) is that there might simply not be much variance. Or are there lots of precincts that went more than 60-40 one way?

        •  64 went to Wynn by more than 60-40 (1+ / 0-)
          Recommended by:

          (although the last was like .399 Edwards and a bunch were close to 40).  26 more went for Edwards by more than 60-40 (although again, some close).  So that means 90 went 60-40 or more one or the other way out of the 169 precincts.  Don't know if that counts as "lots".

          Standard deviation...even though I taught that in a summer class, I'm not sure I calculated it right, but pretty sure Edwards has a standard deviation of 0.126876317.  Again, I don't know if that's a lot by the standards of this sort of thing.

          25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

          by Xenocrypt on Tue May 10, 2011 at 08:51:16 AM PDT

          [ Parent ]

          •  For a fun completely apples-to-oranges comparison (0+ / 0-)

            The standard deviation of Obama's performance in 2008 across the 50 states was 0.0954478927.

            25, Dem, Dude seeing a dude, CT-04(originally), PA-02/NY-12(now)

            by Xenocrypt on Tue May 10, 2011 at 10:02:50 AM PDT

            [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site