Civitas Partisan Index

Assessing Voter Partisanship

Modeled after the Cook Partisan Voting Index developed for congressional districts, the Civitas Partisan Index compares the political leanings of voters in each state house and senate district with the partisan voting tendencies of the state as a whole.  The end result is a letter (D or R) followed by a number, indicating the extent to which each district leans one way or the other.

Please click on the maps below for complete District data and a expanded view of the maps.

CPI Scores & Maps for Each District
House CPI 2008
State House Detail
House CPI 2008
State Senate Detail

How Does It Work?

The CPI compares votes cast in each legislative district to votes cast in the state as a whole, allowing us to see the propensity, or political leanings, of voters in that district. The CPI does not by itself predict election outcomes. Rather, it is a valuable tool in determining the relative likelihood of voters to elect a Democratic versus Republican state legislator, all other things being equal.

As an example, a district whose voters allotted 5 more percentage points to the Democratic candidates compared to the state average receives an index score of D+5. A district whose voters allotted 5 more percentage points to the Republican candidates receives a score of R+5.

Beginning in 2009 the CPI was developed using adjusted 2008 data on the elections for Governor and other council of state offices.

How Well Does It Work?

In 2010, Republicans became the majority in both the House and Senate in part by winning in 17 districts where the CPI indicated voters would vote for a Republican, but at the time the district was held by a Democrat.  In addition, three more districts were won by Republicans that held a neutral value.  The seven districts won by the Republican candidates in 2010 that were not in line with the CPI score ranged from D+1 to D+4.

In 2008, of the 60 contested legislative seats, all but 10 seats went to the winner of the party indicated by the CPI (three seats had a neutral CPI). Among the 10 districts that were not in line with the CPI score, eight were in the range of R+3 to D+3, and most involved races with multi-term incumbents, well-known challengers, or significant spending differences between the candidates. When the value of incumbency is taken into account, the outcome of all but one state House race is correctly predicted by the CPI model.


The newest CPI is developed due to the new redistricted legislative districts – the voter data remains the same – only the district lines have been changed. The CPI will be updated with 2012 election data that should better reflect the composition of the new districts.

In early 2008, the Civitas Institute premiered the North Carolina Partisan Index using data from the 2004 General Election.

Compared to the 2004 CPI model, the 2008 CPI shows a greater concentration of Democratic-leaning voters in the urban population centers. While voters in most of the state were somewhat more likely to vote for Republican statewide candidates in 2008 compared to 2004, voters in much of Buncombe, Cumberland, Durham, Forsyth, Guilford, Mecklenburg, and Wake counties favored Democratic candidates more heavily. In fact, 19 of the 21 state House districts that voted significantly more Democratic in 2008 compared to 2004 – districts whose CPI moved more than 3 points Democratic – are in those seven most populous North Carolina counties.

In 2009, we updated the CPI to reflect voters’ choices in the 2008 General Election.  In 2010 the 2008 CPI map for State House Districts was updated to reflect the changes in the House Redistricting Plan modified by House Bill 1621. The changes were limited to House districts 16 and 18 in Pender and New Hanover counties.  The redistricting did little to impact the CPI data, only House District 16 showed a decrease from R+9 to R+8 while House District 18 remained at D+12.

More About The Model

The CPI is based on voter data from the 2008 general election results for governor and other council of state offices. Although president and U.S. Senate results were also available, we found that state-level races give a more accurate picture of how voters will vote in a state legislative race than do national races.

In council of state races, North Carolinians tend to vote for Democratic candidates – the average council of state vote (looking at only votes for Democratic or Republican candidates) was 53.4 percent Democratic and 46.6 percent Republican. However, many council of state seats have been unevenly contested in the past, making it difficult to predict results “down ticket” looking at raw numbers alone. By looking only at deviations from the state average, the Civitas model is able to mitigate the effects of incumbency and uneven contests.

The CPI model also adjusts for outliers due to the popularity of a specific candidate in a specific area (for example, the governor’s popularity in her home county). Other data assumptions are also applied.

CPI Archive »

Raleigh Web Design, WordPress & Web Development