Skip to content

The Majority In Mississippi PVI

January 28, 2010

On multiple occasions I have made reference to the Cook Partisan Voting Index (PVI) when speaking about Congressional districts. The Cook PVI measures the partisan leaning of the district compared to the nation as a whole.

I have taken that idea, and created a Majority In Mississippi PVI (what I will now refer to as MPI for Majority Partisan Index) which details the partisan voting measures of the 82 counties compared to the state as a whole.

To find the state’s partisan “average,” I averaged the results from the governor’s race in 2003 and 2007, and the presidential election in 2004 and 2008. Those four different races saw an average vote of 57-43 in favor of the Republicans. Therefore, 57 was the key number.

I then prepared a spreadsheet with the county results (for the Republican ticket) of those four elections. If they averaged, say, 60 percent for the GOP then they had a PVI of R+3. If they had a GOP average of 42, then they had a PVI of D+15.

Click here to view the spreadsheet with details on all counties

Please let me know if you find any errors in the post or the spreadsheet.

Some details: I rounded up if the final average came to .5 or higher, rounded down if it was less than .5. Also, if a third party candidate took a percentage of the vote I generally divided their vote evenly between the Republicans and Democrats. So if the final vote was 50-48 with two percent for a third party candidate, I simply made it 51-49. Not perfect but it works.

One Comment leave one →
  1. JackCUDA permalink
    January 29, 2010 6:27 pm

    Assuming this is only for statewide or POTUS general elections.

    I think to be truly valuable you need to somehow index or weight the counties by either population, VAP or, say, average turnout.

    Some of those counties are so small Holmes is D+33 but only 5400 total voters for Governor in ’07, for example. Humphreys D+23 but only 3900 total voters for the same election.

    Then you have the matter of neighboring counties. Hinds is D+18 but for the last decade of statewide and POTUS elections Hinds — as a result of Jackson’s substantial losses in population — has been neutralized as a electoral playing field because Madison R+6 and Rankin R+20 offset and exceed any net D vote advantage produced in Hinds.

    Some sort of volume weighting … maybe a corollary measure per county indicating % of the total vote. So, using Gov’07 as an example Holmes would be, ballpark, .0073, Humphreys .0052.

    Then consider the neighboring county matter for the same election. Hinds .082, Madison .033 and Rankin .048 explains how Madison+Rankin at R+26 offset Hinds at D+18.

    Finally, I think you should go back and include the ’99 Gov and ’00 POTUS elections to increase your sample.


What are you thinking?

Fill in your details below or click an icon to log in:

Gravatar Logo

Please log in to to post a comment to your blog.

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 1,300 other followers

Powered by