How the Final 2004 and 2008 National Exit Polls were forced to match the recorded vote
Richard Charnin
Jan. 6, 2011
This is a quick summary of the changes that were made to the unadjusted, pristine exit polls in order to force them to match the recorded vote.
1988-2008 Unadjusted State and National exit polls vs. recorded votes and National True Vote Model
https://docs.google.com/spreadsheet/pub?key=0AjAk1JUWDMyRdFIzSTJtMTJZekNBWUdtbWp3bHlpWGc&output=html
2004
Table 1A is the unadjusted National Exit Poll (13660 respondents). Kerry had 51.7%.
Table 3A is the adjusted Final NEP Gender crosstab – forced to match the recorded vote (Kerry 47.8%).
Table 4A is the unadjusted NEP ‘Voted 2000′ crosstab (3182 respondents). Kerry had 51.7%.
Table 5A is the adjusted Final NEP ‘Voted 2000′ crosstab – forced to match the recorded vote (Kerry 48.3%).
Table 6A is the True Vote Model. Kerry had 53.3%.
Kerry had 51.1% in the unadjusted state exit poll aggregate (76192 respondents).
Based on the 3182 respondents who were asked how they voted in 2000:
1- The unadjusted 2004 NEP implies that Gore had 47.8%, Bush 48.4%, Other 3.8%.
2- The Final 2004 NEP implies that Gore had 44.6%, Bush 51.8%, Other 3.6%.
But Gore won the popular vote by 540,000 and had 50.8% in the unadjusted exit poll. Therefore, it is likely that the unadjusted 2004 exit poll understated Kerry’s True Vote share by nearly 2%.
2008
Table 3 is the unadjusted NEP (17836 respondents). Obama had 61.0%.
Table 7A is the adjusted Final NEP Gender crosstab – forced to match the recorded vote (Obama 52.7%).
Table 4 is the unadjusted NEP ‘Voted 2004′ crosstab (4178 respondents). Obama had 58.0%.
Table 12 is the adjusted Final NEP ‘Voted 2004′ crosstab – forced to match the recorded vote (Obama 52.9%).
Table 2 is the unadjusted state exit poll aggregate (82388 respondents). Obama had 58.1%.
Table 7 is the True Vote Model. Obama had 58.0%.
Based on the 4178 respondents who were asked how they voted in 2004:
1- The unadjusted 2008 NEP implies that Kerry had 50.2%, Bush 44.6%, Other 5.2%.
This is close to the unadjusted 2004 NEP (Kerry 51.7%), but the 3rd party (Other) recorded share was 1.0%, a 4.2% discrepancy from the implied share.
2- The Final 2008 NEP implies that Kerry had 42.3%, Bush 52.6%, Other 4.6%.
This is far from both the unadjusted and Final 2004 NEP (Kerry 48.3%). The discrepancy is due to the Final 2008 NEP forced match to the recorded vote.
National Exit Poll Timeline
This refutes the myth that early exit polls were biased to Kerry. He led from 4pm with 51% (8,349 respondents) to the final 13,660 (51.7%). The exit pollsters had to switch approximately 471 (6.7%) of Kerry’s 7,064 responders to Bush in order to force the Final NEP to match to the recorded vote. Given his 51.7% share of 125.7 million votes cast, Kerry won by nearly 6 million votes. But the True Vote Model indicates he had 53.6% and won by 10 million.
11/3/04 1:24pm, Final National Exit Poll, 13660 respondents
Adjusted Sample: Kerry 48% (6,557); Bush 51% (6,966)
Forced to match recorded vote by switching approximately 507 (7.2%) of Kerry’s 7,064 respondents to Bush.
http://www.richardcharnin.com/US2004G_3970_PRES04_NONE_H_Data.pdf
Unadjusted National Exit Poll, 13660 respondents
Sample Kerry Bush Other
13,660 7,064 6,414 182
Share 51.7% 47.0% 1.3%
http://webapps.ropercenter.uconn.edu/CFIDE/cf/action/catalog/abstract.cfm?label=&keyword=USMI2004-NATELEC&fromDate=&toDate=&organization=Any&type=&keywordOptions=1&start=1&id=&exclude=&excludeOptions=1&topic=Any&sortBy=DESC&archno=USMI2004-NATELEC&abstract=abstract&x=32&y=9
11/2/04 12:22am, 13047 respondents
Kerry 51%; Bush 48%
http://media.washingtonpost.com/wp-srv/politics/elections/2004/graphics/exitpolls_us_110204.gif
11/2/04 7:33pm, 11027 respondents
Kerry 51%; Bush 48%
http://www.richardcharnin.com/US2004G_3798_PRES04_NONE_H_Data.pdf
11/2/04 3:59pm, 8349 respondents
Kerry 51%; Bush 48%
http://www.richardcharnin.com/US2004G_3737_PRES04_NONE_H_Data-1.pdf
2004
Final Exit Polls
http://www.cnn.com/ELECTION/2004/pages/results/states/US/P/00/epolls.0.html
2008
Final Exit Polls
http://www.cnn.com/ELECTION/2008/results/polls/#val=USP00p1
Election Model Forecast; Post-election True Vote Model
2004 (2-party vote shares)
Model: Kerry 51.8%, 337 EV (snapshot)
State exit poll aggregate: 51.7%, 337 EV
Recorded Vote: 48.3%, 255 EV
True Vote Model: 53.6%, 364 EV
2008
Model: Obama 53.1%, 365.3 EV (simulation mean)
Recorded: 52.9%, 365 EV
State exit poll aggregate: 58.0%, 420 EV
True Vote Model: 58.0%, 420 EV
2012 (2-party state exit poll aggregate shares)
Model: Obama 51.6%, 332 EV (Snapshot)
Recorded: 51.6%, 332 EV
True Vote Model: 55.2%, 380 EV
