POLIDATA ® Corrigenda

General Note: PRESIDENTIAL RESULTS BY CONGRESSIONAL DISTRICTS

As discussed on the PRESIDENTIAL RESULTS BY CONGRESSIONAL DISTRICTS Project page, the results denominated as the PRCD information are compiled independently by Polidata in conjunction with various sponsors. As such, they are compiled right after the election from source material, most of which is still on paper or, at any rate, not data files. Regardless, most of this work is accomplished before the availability of precinct-level election datasets prepared by State Election Authorities.
This means that there may be some differences in the raw numbers, the percentages, or even the 'winner' of the district between the Polidata information and numbers that could be compiled from information finally available from the States.
As part of the ongoing Polidata data library development processing, precinct-level datasets are acquired and/or compiled when available. As time allows, comparisons are made between the State version and the Polidata version for the PRCD data.
It is also important to understand that these numbers are, generally, the result of compilation of precinct-level information, many of which are 'split', i.e., they contain portions of more than one district and/or are in jurisdictions that require the deallocation of centrally-counted absentee and/or early votes. The sheer size of the compilation effort may result in clerical errors. The deallocation process, for both split units and centrally-counted votes, requires judgment as to the factor to be used to allocate the votes to districts: in such districts there are no 'hard' numbers.
Moreover, even when the final datasets are available, inconsistency exists as to how votes in these situations are addressed by the dataset developer.
Differences, when noted, may be available here. Generally speaking, the datasets are not updated in a piecemeal fashion unless it is a clear and significant error.
Find more information on the development process at the development notes page: .

PRESIDENTIAL RESULTS BY CONGRESSIONAL DISTRICTS-Update for year-to-year shifts

Shifts in the year-to-year breakdown of the presidential vote for the 110th Congress were the result of changes of all districts in GA and a handful in TX with the following shifts.
- A1. Districts voting for Bush in 2000: 240 (was 240).
- A2. Districts voting for Gore in 2000: 195 (was 195).
- A3. Districts voting for Bush in 2004: 256 (was 255).
- A4. Districts voting for Kerry in 2004: 179 (was 180).
- B1. Districts voting for Bush in 2000 and Bush in 2004: 237 (was 238).
- B2. Districts voting for Bush in 2000 and Kerry in 2004: 3 (was 2).
- B3. Districts voting for Gore in 2000 and Bush in 2004: 19 (was 17).
- B4. Districts voting for Gore in 2000 and Kerry in 2004: 176 (was 178).
Districts shifting:
1. GA-2 (Bishop) from Bush:Bush to Gore:Bush;
2. GA-12 (Barrows) from Gore:Kerry to Gore:Bush;
3. TX-15 (Hinojosa) from Bush:Bush to Gore:Bush;
4. TX-25 (Doggett) from Gore:Kerry to Bush:Kerry;
5. TX-28 (Cuellar) from Gore:Bush to Bush:Bush
Aside from TX-25 (Doggett), and to a much lesser degree TX-15 (Hinojosa), the shifts were of only a few points

OHIO: 2004: Districts of 109th Congress

A discrepancy was discovered between the PRCD data and the official State dataset after it was compiled for analytical purposes.

The discrepancy affected the 'winner' of one district. (See CD-15, Deborah Pryce-R.)
The vote totals for several districts were also affected, some by more than 5,000 votes. (See CDs 6 and 17; and CDs 12 and 15.)
The discrepancy, alas, appears to be on the part of the state: the dataset does not allocate votes for split units and does not intrinsically flag split units.
The district with the 'change' in the 'winner' was CD-15, Deborah Pryce-R. The Polidata breakdown was 50.3% Bush to 49.6% Kerry, or Bush by 0.7%; the State breakdown is 49.5% Bush to 49.8% Kerry, or Kerry by 0.2%.
The change in CD-15 would mean 1 less Bush Republican and 1 more Kerry Republican.

The discrepancy does not result in a change to the dataset.

The dataset can be ordered at the dataset ordering page.

POLIDATA Home Page.

Contact information. EMAIL: info@polidata.org.

HOME: www.polidata.org. THIS: http://www.polidata.org/correct.default.htm updated: 2007-nov-19