Prog Blogs, Orgs & Alumni
Bank Slate
Betsy Devine
birch, finch, beech
Democracy for NH
Live Free or Die
Mike Caulfield
Miscellany Blue
Granite State Progress
Seacoast for Change
Still No Going Back
Susan the Bruce
Tomorrow's Progressives
Politicos & Punditry
The Burt Cohen Show
John Gregg
Krauss
Landrigan
Lawson
Pindell
Primary Monitor
Primary Wire
Scala
Schoenberg
Spiliotes
Welch
Campaigns, Et Alia.
Paul Hodes
Carol Shea-Porter
Ann McLane Kuster
John Lynch
Jennifer Daler
ActBlue Hampshire
NHDP
DCCC
DSCC
DNC
National
Balloon Juice
billmon
Congress Matters
DailyKos
Digby
Hold Fast
Eschaton
FiveThirtyEight
MyDD
The Next Hurrah
Open Left
Senate Guru
Swing State Project
Talking Points Memo
50 State Blog Network
Alabama
Arizona
California
Colorado
Connecticut
Delaware
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Missouri
Missouri
Montana
Nebraska
Nevada New Hampshire
New Jersey
New Mexico
New York
New York
North Carolina
North Dakota
Ohio
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Dakota
Tennessee
Texas
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
To better understand Granite State voting behavior, Judy Stadtman and I have compiled town-level New Hampshire voting data from each statewide election since 2000. One of the exercises in that analysis was to calculate and document the Partisan Voting Index for each of the state's 103 State House districts. That data has now become the basis for a model to forecast the results of the upcoming State House of Representatives election.
At a high level, the model is based on a strong correlation between the national partisan vote and results in New Hampshire House elections. I begin with the projected national two-party vote, adjust it for each House district based on the district PVI, and determine the probability of a partisan outcome for each seat. I then run 100,000 simulations of the outcome for each seat to project the total number of seats for each party.
In this post, I'll detail the methodology (warning: geek alert!) and test the model by retroactively applying it to the 2004, 2006 and 2008 House elections. Part 2 will follow with my fearless forecast.
Update: the links have been repaired... sorry about that
When Mr. Tucker and I first starting mining data for our New Hampshire election map series, my initial motive was simply to figure out whether specific regions and communities drove the Democratic vote in November 2008, and if geographic patterns of partisan vote distribution in Presidential election years had changed in a distinctive way since 2000. (Answer: not exactly, and yes.)
Since then we've discovered that there is a nearly endless supply of reliable, public-domain data on voting and demographic trends that anyone can fool around with to produce map-ready results. Lately, Mr. Tucker has been urging me to consider the possibility that our little hobby is getting out of hand - it's an engaging intellectual exercise, but to be honest there's a significant time suck factor. So why do it?
At the end of the day, the one sure thing that micro-parsing of election results can tell us is where we've been, plus a little bit about where we are now. As an astute observer noted at a strategy meeting I attended recently, in New Hampshire, every election is an anomaly. The political, cultural, economic, and social crosscurrents that effect election outcomes - not to mention the vagaries of human nature and random, unplanned events that inevitably get into the mix - are so complex and dynamic that election predictions based on raw historic patterns are just asking to be shot down. My theory is that the next best thing to owning a crystal ball that will magically reveal the political future is having a fact-based understanding of the structural conditions in the field.
Hence the maps - and all the old faves, plus a bunch of new ones, were uploaded to a special section of the Miscellany Blue blog today.
Update: Added "Party Distribution of House Seats by PVI" Chart.
The next (and last?) chapter in my look at the partisan makeup of New Hampshire voters is a study of the 103 districts for the state House of Representatives. This is a follow up to earlier looks at the voting wards and the state Senate districts.
I have consolidated the voter ward data from my initial New Hampshire Partisan Voting Index study to take a look at the partisan makeup of New Hampshire's 24 state Senate districts. (For an introduction to the Cook PVI and my methodology, refer to the original diary.)
(Part put below the fold. - promoted by Dean Barker)
When I rolled out the New Hampshire voting ward PVI maps, one of the commenters referred to their predictive value. It's an interesting comment. In my mind, the index has clear value as a means of measuring and comparing voting patterns across districts. But can it also be used to help predict election results?
(OMG this is a tremendous amount of very useful work. Thank you! - promoted by Dean Barker)
The Cook Partisan Voting Index (PVI) compares how a congressional district votes relative to the nation as a whole. The index objectively measures each congressional district as a means to allow comparisons between districts that are relevant in both mid-term and presidential election years.
The index is derived by averaging a district's voting results from the previous two presidential elections and comparing them to national results. The result indicates the number of percentage points by which the party's vote exceeded the national average.
For example, a D+2 PVI means the district performed two points more Democratic than the nation did as a whole in the 2004 and 2008 presidential elections. The Democratic candidates would have received roughly 53.3% of the two-party vote compared to the national two-party average of 51.3%.
Using this same methodology, I have calculated the PVI for each New Hampshire voting ward and created a map of the results.
158 of the state's 299 voting wards with at least 25 votes in the 2008 presidential election lean Democratic. They are led by Hanover as the most Democratic ward in New Hampshire with a D+29 PVI. 118 wards lean Republican. The most Republican voting ward in the state (with at least 200 votes) is New Ipswich in Hillsborough county with a R+16 PVI.
Next, I calculated the 2004 PVI and compared it to the 2008 index. The results confirm the extent to which New Hampshire, and particularly CD-2, is trending blue. In just four years, 186 wards have become more Democratic by at least two points compared to only 23 wards that have become more Republican by the same margin.
This study utilizes data previously collected and cataloged by my wife, Judy Stadtman, for her excellent series of maps illustrating New Hampshire voting results since 2000. Thanks to Dean Barker for suggesting the study. All of the maps and data can be viewed and downloaded here.
Dante Scala has, most helpfully, pegged three cycles worth of PVI (Partisan Voting Index) for each congressional district of NH. Here it is in sum:
NH-01NH-02 2000: -2 0 2004: 0 +4 2008: 0 +3 PVI
(Now, before you get nervous about the CD2 one point decrease from '04 to '08, bear in mind that in '04 NH was the only state to flip from Bush to Kerry, while last cycle Obama racked up big margins throughout the US and flipped a number of states. So speaking in relative terms, both numbers are quite robust. Relatedly, elwood wondered aloud offline whether Nader had some dampening effect on the CD2 '00 PVI.)
And Dante:
Over in NH 02, however, the eventual Democratic nominee begins the general election race with some protection against a potential wave. And it's the Second District that gives New Hampshire its Democratic tilt for 2012.
This is very much in line with earlier voter demographic breakdowns he has been doing that show the rise of Democrats in Sullivan, Grafton, and Coos counties (on top of what is already a dominant Democratic region in Cheshire co.)
And I'll say what he doesn't: these Dems are more politically and culturally aligned with Vermont than with Boston-area Massachusetts. Although this is so broad a stereotype brush I am asking for a flame war, the western border of the state has become a solid base for progressive Dems. Bradley Dems were strong there in '00, Dean Dems in '04, and Edwards/Obama Dems in '08.
How could it be that Obama has opened up double digit lead in two different polls? It's easy if you follow what Dante said (and buy his current PVI projection for NH, which I do):
Let's start with a look at the national polls. It's clear that Obama has broken this race wide open, at least for now. Pollster.com estimates his lead nationally at 7 points.
Three months ago, we estimated New Hampshire's PVI (partisan voting index) at R-2 or R-3. In other words, given the results of the 2004 election and demographic trends since that time, one would expect a Republican presidential candidate to perform two or three percentage points below his national standing here in the Granite State.
In that context, a 10-point Obama lead here is perfectly plausible, in the context of a national race where the Democrat leads by seven. In that context, New Hampshire is behaving just as we might expect from a purple state with a Democratic tilt.