GEORGIA POLLSTERS: Perdue +2.3 Ossoff 44.4, Perdue 46.7 REF'S AVG: Perdue +2.8 - Ossoff 45.3, Perdue 48.1 RETURN TO SENATE RACES 2016 Exit Poll - 34D/30I/36R without leaners (PDF) 2018 Exit Poll - 41D/11I/48R with leaners (PDF) Pollsters' Averages Chart Political Ref's Averages Chart Poll Date Sample + Party Distribution Result Pollster Avg Ref's Avg Landmark 11/1 500LV Ossoff 47, Perdue 49 x   Political Ref's Take on the above poll The party distribution is reasonable. No adjustment necessary. Included   x Emerson 10/29-10/31 749LV, 35/22/43 Ossoff 49, Perdue 48 x   Political Ref's Take on the above poll The party distribution is reasonable. No adjustment necessary. Included   x Quinnipiac Univ 9/23-9/27 1125LV, 35D/25I/33R Ossoff 49, Perdue 48 x   Political Ref's Take on the above poll The D+2 sample includes too many D's, will adjust to 34D/30I/36R. According to the poll data, Ossoff won D's 96-2, I's 53-43 and Perdue won R's 97-2. Applying these voter ratios to the more likely turnout, Ossoff receives 49.3 to Perdue 48.5. Ossoff 49.3, Perdue 48.5   x YouGov 9/22-9/25 1164LV, 37D/20I/40R Ossoff 42, Perdue 47 x   Political Ref's Take on the above poll The party distribution is reasonable. No adjustment necessary. Included   x Monmouth Univ 9/17-9/21 402LV, 34D/28I/37R Ossoff 42, Perdue 50 x   Political Ref's Take on the above poll Party distribution is reasonable. No adjustment necessary. Included   x Univ. of Georgia 9/11-9/20 1150LV, Party distribution not provided Ossoff 45, Perdue 47 x   Political Ref's Take on the above poll Neither party distribution or voter ratios are provided nor can be inferred. Adjustment is not possible. Excluded. Excluded.     Morning Consult 9/11-9/20 613LV, Party distribution not provided Ossoff 44, Perdue 43 x   Political Ref's Take on the above poll Neither party distribution or voter ratios are provided nor can be inferred. Adjustment is not possible. Excluded. Excluded.     HarrisX 8/20-8/30 34D/30I/36R Ossoff 40, Perdue 47 x   Political Ref's Take on the above poll The party distribution looks reasonable. No adjustment necessary. Included   x Fabrizio Ward/Hart Research Assoc 8/30-9/5 800LV, 31D/35I/34R Ossoff 48, Perdue 47 x   Political Ref's Take on the above poll The party distribution looks reasonable. No adjustment necessary. Included   x PPP 8/13-8/14 530RV, 35D/26I/39R Ossoff 44, Perdue 44 >1month   Political Ref's Take on the above poll The party distribution looks reasonable. Will apply a registered voter poll to the entire sample because this poll includes far too many respondents who will not actually vote. Assuming 31% of respondents will fail to vote, 55% of which are Dem leaning and 41% are GOP leaning. This results in an adjustment of Ossoff to 27 and Perdue to 31.3. Applying the 27/31.3 ratio to the 88% who chose either Ossoff or Perdue, the final adjusted result is Ossoff 40.8 to Perdue 47.2. Ossoff 40.8, Perdue 47.2   >1month SurveyUSA 8/11 623LV, Party Distribution Not Provided Ossoff 41, Perdue 44 >1month   Political Ref's Take on the above poll Party distribution is not provided but voter ratios are. According to the poll data, Ossoff won D's 88-3, Perdue won I's 41-38 and R's 91-5. Applying these voter ratios to the most likely voter turnout of 34D/30I/36R, Ossoff receives 43.1 to Perdue 46.1 Ossoff 43.1, Perdue 46.1   >1month YouGov 7/28 1101LV, 36D/26I/36R Ossoff 43, Perdue 45 >1month   Political Ref's Take on the above poll The even turnout somewhat oversamples Dems. Will adjust to R+2, 34D/30I/36R voter turnout assumption. According to the poll data, Ossoff won D's 84-5, Perdue won I's 43-39 and R's 87-4. Applying these ratios to the more likely voter turnout, Ossoff receives 41.7 and Perdue 45.9. Ossoff 41.7, Perdue 45.9   >1month Monmouth Univ 7/23-7/27 402LV, 32D/31I/37R Ossoff 43, Perdue 49 >1month   Political Ref's Take on the above poll The likely voter turnout model falls between 2016 and 2018. No adjustment necessary. Included   >1month Spry Strategies 7/11-7/16 701LV, 33.1D/24I/41.5R Ossoff 43.7, Perdue 46.0 >1month   Political Ref's Take on the above poll The R+8.4 sample oversamples Republicans. Adjusting to the more likely R+2 turnout similar to 2016, assuming 36D/23I/38R. Voter ratios within parties is not provided but can be inferred. Perdue had an 8.4 point turnout advantage but only leads by 2.3 points. For this reason we can assume Ossoff won D's and I's by a bigger margin than Perdue won I's and R's. Will assume just for purposes of calculation that Ossoff's gains came entirely from Independents although this is not true and done just for the purposes of simplifying the calculation. It will, nevertheless, produce a reasonably accurate adjustment. Assuming Ossoff and Perdue each won their party 90-5, and Ossoff won Independents 15-9, or 63%-37% (a fiction assumed for purposes of calculation), Ossoff receives 48.8 to Perdue 44.5. Ossoff 48.8, Perdue 44.5   >1month Garin, Hart, Yang Research Group 7/22 800LV, Party Distribution Not Provided Ossoff 45, Perdue 44 >1month   Political Ref's Take on the above poll Neither party distribution or voter ratios are provided nor can be inferred. Adjustment is not possible. Excluded. Excluded.     Gravis Marketing 7/2 513LV, 34D/29I/37R Ossoff 43, Perdue 48 >1 month   Political Ref's Take on the above poll The party distribution of the voter sample matched 2016 almost exactly. No adjustment necessary. No adjustment necessary   >1 month Fox News 6/20-6/23 1013RV, 41D/16I/44R Ossoff 42, Perdue 45 >1 month   Political Ref's Take on the above poll The R+3 turnout model closely matches 2016. No adjustment necessary. No adjustment necessary.   >1 month PPP 6/12-6/13 661RV, 38D/23I/40R Ossoff 45, Perdue 44 >1 month   Political Ref's Take on the above poll The R+2 sample matches the 2016 exit poll exactly. The party distribution is also very close. No adjustment necessary. No adjustment necessary   >1 month