(ALL YEARS: ASKED OF Rs WHO DID NOT VOTE): 1948,1952,1956,1960,1964,1968: Who (1948: whom) would you have voted for if you had voted? 1980,1984,1988,1992,1996: How about the election for president? Did you prefer one of the candidates for president? (IF YES:) Who did you prefer?
Presidential preference among nonvoters · Asked in 16 waves, 1952–2020 · 30,963 responses · weighted estimates computed from individual responses
'No preference' (code 7) is shown as a non-answer: it was coded only in years that asked a 'did you prefer one of the candidates?' filter (1980-2000 and 2008; volunteered in 1968), so the Democratic, Republican, and other shares are what compare across years. Codes that include 'no Pre/Post IW' were dropped for respondents without that interview. Kept and flagged: 0 (other_missing), 7 (other_missing), 9 (other_missing). Codebook notes: This is a follow-up to VCF0704-VCF0706. GENERAL NOTE: In 1952,1956,1960,1964 no category appeared (coded) for ‘no preference.’ Although a category appeared for ‘no preference’ in 1968, R was not asked the filter question, ‘Did you prefer one of the candidates for president?’ and the absence of this filter was reflected in the small 1968 ‘no preference’ N (presumably volunteered responses). In 1980, 1984, 1988 and 1992 Rs who replied ‘no’ to the filter question were coded 7 here. 1956 NOTE: One case from dual ‘nonvoter, NA preference; voted but not for president’ has been coded 9 here. 1968 NOTE: One case originally coded ‘Nixon or Wallace’ has been coded 9 here. 2004 NOTE: Code 5, ‘Ralph Nader’ and code 7, ‘Other’ have been coded 3 here.
Dashed lines mark changes in how interviews were done, which can shift answers: 1. → mixed mode (2012).
Shares of people who gave an answer. Each column is one wave, evenly spaced rather than to scale in time; hover for exact shares.
| Group | Preference: Democrat | Preference: Republican | Preference: Other | n | MoE | |
|---|---|---|---|---|---|---|
| All respondents |
| Poll | Field dates | Mode | Respondents | Responses |
|---|---|---|---|---|
| ANES 2020 Time Series Study Source variable: V202077. | 2020-08-18 to 2021-01-04 | mixed | 8,280 | 675 |
| ANES 2016 Time Series Study Source variable: V162038x,V162062x. 77% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 2016-09-07 to 2017-01-09 | mixed | 4,270 | 3,671 |
| ANES 2012 Time Series Study Source variable: postvote_prefprwho. 88% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 2012-09-09 to 2013-01-13 | mixed | 5,914 | 5,510 |
| ANES 2008 Time Series Study 76% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 2008-09-02 to 2008-12-30 | face to face | 2,322 | 2,102 |
| ANES 2004 Time Series Study Source variable: V045028,V045029. 84% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 2004-09-07 to 2004-12-19 | face to face | 1,212 | 1,066 |
| ANES 2000 Time Series Study |
Dashed lines mark changes in how interviews were done, which can shift answers: 1. → mixed mode (2012).
Shares of people who gave an answer. A group’s column is left blank in waves where its effective sample is under 50; hover for exact shares.
| 43% |
| 47% |
| 10% |
| 3671 |
| ±2 |
| Democrats | 76% | 18% | 7% | 1259 | ±3 |
| Independents (incl. leaners) | 37% | 47% | 16% | 1351 | ±3 |
| Republicans | 8% | 88% | 4% | 1043 | ±4 |
ANES 2016 Time Series Study · US citizens 18+ (face-to-face samples: contiguous 48 states and DC; web samples: 50 states and DC). Shares of people who gave an answer. n is unweighted; ± is the 95% margin of error at 50% from the weighted effective sample. Groups with an effective sample under 50 are grayed; under 20, hidden.
| 2000-09-05 to 2000-12-22 |
| mixed |
| 1,807 |
| 1,555 |
| ANES 1996 Time Series Study Source variable: V961098,V961099. 77% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1996-09-03 to 1997-01-02 | mixed | 1,714 | 1,534 |
| ANES 1992 Time Series Study Source variable: V925634. 76% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1992-09-02 to 1993-01-16 | mixed | 2,485 | 2,255 |
| ANES 1988 Time Series Study Source variable: V880774,V880775. 70% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1988-09-06 to 1989-01-23 | face to face | 2,040 | 1,775 |
| ANES 1984 Time Series Study Source variable: V840805,V84080. 74% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1984-09-04 to 1985-01-24 | mixed | 2,257 | 1,989 |
| ANES 1980 Time Series Study Source variable: V801014,V801016. 71% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1980-09-02 to 1981-02-07 | face to face | 1,614 | 1,408 |
| ANES 1968 Time Series Study Source variable: V680316. 76% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1968-09-06 to 1969-01-28 | face to face | 1,557 | 1,388 |
| ANES 1964 Time Series Study 78% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1964-09-07 to 1965-02-01 | face to face | 1,571 | 1,450 |
| ANES 1960 Time Series Study Source variable: V600201. 82% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1960-09-12 to 1960-12-26 | face to face | 1,181 | 1,109 |
| ANES 1956 Time Series Study Source variable: V560203. 73% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1956-09-17 to 1957-01-19 | face to face | 1,762 | 1,762 |
| ANES 1952 Time Series Study Source variable: V520178. 73% of responses are code 0, which the codebook uses for respondents outside the question's universe as well as DK/RF/NA (kept as other_missing); compute substantive shares without it. | 1952-09-15 | face to face | 1,899 | 1,714 |