Research
Party Polarization, Ideological Sorting and the Emergence - Supplemental Appendix
-
University:
Boston University -
Course:
CAS PO 501 | Formal Political Theory Academic year:
2024
-
Views:
194
Pages:
31
Author:
princesa035c6tg
7
Supplemental Appendix
7.1
Data Coding Appendix
We attempted to collect and code a consistent set of political attitudes and demographic
variables for all polls conducted by the Gallup Organization between 1953 and 2012 that have
individual-level data posted on the Roper Center iPoll database One challenge in this effort was
that Gallup frequently changed both the constructs they were trying to measure and the questions
used to measure these constructs over time. Because there were many more questions asked than
we could feasibly code, we limited ourselves to coding only responses to questions that were asked
frequently over time, and in a consistent enough manner that responses over time were comparable.
Table A.1 presents the political attitudes that are included in our dataset. We coded responses
to questions about presidential approval, partisan identification, and ideology. The standard presidential approval question is “Do you approve or disapprove of the way that
is handling his job as president?” Because there is variation across surveys in how Gallup coded
responses like “Don’t Know” or “Neither Approve or Disapprove” in the raw data, we code any
response other than “Approve” or “Disapprove” as “Other.” Gallup also occasionally asks domainspecific presidential approval after the standard presidential approval question. When asked, we
also used a similar scheme to code responses to questions about the president’s handling of the
economy and foreign affairs.
Table A.2 displays the number of observations and surveys that contains responses to the
standard presidential approval question by quarter. We observe approximately 20,000 responses
from about 15 surveys in a modal year, with the number of responses and surveys observed in a year
increasing somewhat over time. There are a few quarters in which we do not observe any surveys.
This happens because Gallup stopped asking presidential approval immediately prior to some
presidential elections. To assess our coverage of these Gallup polls, we examined whether there were
polls that had aggregate totals listed at http://www.ropercenter.uconn.edu/data_access/
data/presidential_approval.html in July 2013 but did not have usable individual-level data
in the Roper Center iPoll Databank. Table A.3 lists the 135 polls that fit this description. Given
A1 Table A.1: Description of Political Variable Codings
Variable
Variable Name
Coding
Presidential Approval:
Job as President
pres approve
Approve
Disapprove
Other
=
=
=
1
-1
0
Handling of Economy
pres approve economy
Approve
Disapprove
Other
=
=
=
1
-1
0
Handling of Foreign Affairs
pres approve foreign
Approve
Disapprove
Other
=
=
=
1
-1
0
party
Republican
Democrat
Other
=
=
=
1
-1
0
party2
Republican
Democrat
Other
=
=
=
1
-1
0
ideo
Very Conservative
Conservative
Liberal
Very Liberal
Other
=
=
=
=
=
2
1
-1
-2
0
Partisan Identification:
Consider Yourself
Lean More to
Ideology
Notes: -9 indicates missing value, -99 indicates variable not included in series.
A2 that we observe over 1,400 polls with presidential approval, this suggests that we are observing a
high percentage of the possible surveys.
We report a similar breakdown of the number of observations and surveys that contain responses to partisan identification by quarter in Table A.4. Unlike with presidential approval,
Gallup asks about partisan identification in just about every survey we coded. The exact wording
of the partisan identification question varies slightly across surveys. The two most common forms
of the question are: “in politics, as of today, do you consider yourself a Republican, Democrat, or
Independent” and “in politics today, do you consider yourself a Republican, Democrat, or Independent?” There are also a few times in the early 1950s when instead the question was worded:
“Normally, do you consider yourself a Democrat, Republican, or Independent.” While respondents
sometimes provide alternative answers (e.g., support a third party, don’t know, refused to answer),
these responses cannot always be differentiated from “Independent” in the raw data. Thus, we
again jointly code all responses other than Democratic or Republican into an omnibus “Other”
category. In some surveys, Gallup also asks a follow-up question to individuals who do not initially
identify as a Democratic or Republican about whether they lean towards either party. The exact
question wording is “As of today do you lean more to the Democratic Party or the Republican
Party?” This question was asked somewhat frequently in the 1950s, quite rarely in the 1960s or
1970s, and then frequently again beginning in the 1980s. Responses to these questions are coded
when available.
Gallup has asked about ideology for less time than either presidential approval or partisan
identification. While questions about ideology were occasionally asked in the 1980s and the early
1990s, Gallup only began regularly asking about ideology using a consistent question wording in
1992: “How would you describe your political views - very conservative, conservative, moderate,
liberal, or very liberal?” Table A.5 shows the number of observations and surveys that contains
responses to this question by quarter. We cannot always differentiate in the raw data between
people who respond that they are moderate and those who give another answer (e.g., don’t know,
refuse to respond), so all responses that are not liberal or conservative are placed into an omnibus
“Other” category.
A3 Table A.2: No. of Obs. (Surveys) with Presidential Approval by Quarter
Quarter
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1
4,720
(3)
6,131
(4)
7,443
(5)
8,918
(5)
6,154
(4)
7,710
(5)
4,592
(3)
5,049
(3)
4,300
(2)
7,736
(3)
2
3,075
(2)
5,747
(4)
6,084
(4)
7,975
(4)
7,761
(5)
4,547
(3)
6,292
(4)
7,351
(4)
12,591
(5)
9,215
(4)
3
5,938
(4)
7,727
(5)
5,848
(4)
4,276
(2)
6,131
(4)
7,616
(5)
7,005
(3)
14,752
(7)
6,763
(3)
6,875
(3)
4
5,987
(4)
1963
4,468
(3)
1964
2,977
(2)
1965
3,043
(2)
1966
2,991
(2)
1967
4,514
(3)
1968
6,834
(4)
1969
6,497
(3)
1970
6,128
(3)
1971
8,014
(3)
1972
1
6,889
(3)
15,263
(6)
7,832
(4)
12,977
(5)
8,913
(4)
4,503
(3)
7,675
(5)
9,281
(6)
4,634
(3)
6,046
(4)
2
10,626
(5)
15,555
(6)
11,204
(5)
10,144
(4)
12,200
(5)
7,653
(5)
7,701
(5)
6,062
(4)
7,945
(5)
6,134
(4)
3
5,605
(3)
0
(0)
10,566
(4)
8,925
(4)
9,932
(4)
4,552
(3)
7,810
(5)
7,544
(5)
3,108
(2)
0
(0)
4
8,566
(4)
1973
2,498
(1)
1974
9,586
(4)
1975
9,760
(4)
1976
6,365
(4)
1977
3,027
(2)
1978
6,222
(4)
1979
4,662
(3)
1980
4,588
(3)
1981
2,966
(2)
1982
1
6,145
(4)
11,015
(7)
7,747
(5)
7,786
(5)
7,757
(5)
9,233
(6)
10,721
(7)
9,527
(6)
4,799
(3)
6,121
(4)
2
9,281
(6)
10,196
(8)
7,912
(5)
4,607
(3)
10,671
(7)
10,712
(7)
9,158
(6)
9,409
(6)
9,193
(6)
9,282
(6)
3
7,609
(5)
7,816
(5)
4,635
(3)
0
(0)
7,564
(5)
13,969
(8)
10,903
(7)
4,750
(3)
7,699
(5)
7,580
(5)
4
7,795
(5)
1983
7,823
(5)
1984
9,213
(6)
1985
1,559
(1)
1986
10,609
(7)
1987
4,658
(3)
1988
9,226
(6)
1989
3,100
(2)
1990
7,666
(5)
1991
6,123
(4)
1992
1
7,742
(5)
6,231
(4)
7,944
(6)
3,711
(4)
5,368
(6)
4,480
(4)
6,907
(6)
5,918
(5)
18,885
(20)
7,276
(7)
2
9,161
(6)
7,340
(6)
6,999
(6)
5,626
(5)
8,247
(5)
4,032
(2)
11,116
(11)
5,695
(5)
10,967
(11)
9,343
(8)
3
10,774
(7)
10,848
(7)
7,023
(6)
4,657
(5)
5,523
(6)
2,001
(2)
6,753
(6)
16,483
(16)
11,759
(11)
4,442
(4)
4
6,066
(4)
1993
6,052
(4)
1994
3,136
(3)
1995
5,638
(5)
1996
7,170
(8)
1997
1,025
(1)
1998
7,527
(7)
1999
15,294
(16)
2000
2,008
(2)
2001
4,046
(4)
2002
1
11,069
(12)
10,070
(10)
7,021
(7)
6,076
(6)
8,167
(8)
9,970
(11)
12,328
(13)
11,300
(10)
5,089
(5)
6,705
(7)
2
8,124
(8)
7,793
(8)
8,822
(9)
10,164
(10)
3,970
(4)
4,697
(5)
10,302
(9)
4,114
(4)
5,055
(5)
8,869
(9)
3
10,954
(11)
8,996
(9)
12,233
(12)
9,097
(9)
7,974
(8)
13,756
(17)
8,954
(8)
7,225
(7)
5,883
(6)
8,442
(9)
4
10,251
(10)
2003
10,117
(9)
2004
7,127
(7)
2005
4,850
(6)
2006
4,907
(5)
2007
12,600
(12)
2008
5,060
(5)
2009
5,121
(5)
2010
4,881
(5)
2011
7,779
(7)
2012
1
12,063
(12)
7,053
(7)
6,946
(7)
5,030
(5)
5,030
(5)
11,134
(8)
16,910
(20)
14,239
(17)
15,324
(18)
14,843
(14)
2
7,086
(7)
5,013
(5)
8,959
(9)
6,032
(6)
6,042
(6)
7,532
(7)
13,695
(17)
13,921
(15)
9,194
(11)
15,967
(15)
3
4,021
(4)
5,551
(5)
7,864
(8)
4,016
(4)
6,084
(6)
7,095
(7)
12,874
(14)
10,396
(11)
9,953
(11)
15,641
(15)
4
5,018
(5)
8,613
(7)
7,666
(8)
4,534
(4)
7,101
(7)
10,376
(8)
10,215
(10)
7,345
(6)
10,630
(12)
19,071
(19)
A4 Table A.3: Missing Presidential Approval Data Series
Year (# Missing)
Date(s) of Missing Series
1961 (1)
4/28-5/3
1964 (1)
12/11-12/16
1965 (1)
3/11-3/16
1968 (1)
3/10-3/15
1978 (1)
11/10-11/13
1984 (2)
5/3-5/5,6/6-6/8
1985 (4)
8/16-8/19,9/13-9/16,11/1-11/4,12/6-12/9
1986 (5)
5/16-5/19,6/6-6/9,8/8-8/11,9/12-9/15,12/5-12/8
1987 (4)
3/6-3/9,6/5-6/8,8/7-8/10,12/4-12/7
1988 (10)
1/22-1/25,3/4-3/7,4/8-4/11,6/10-6/13,6/24-6/27
7/15-7/18,8/19-8/22,9/25-10/1,10/21-10/24,12/27-12/29
1990 (6)
3/15-3/18,3/16-3/29,4/19-4/22,5/17-5/20,6/7-6/10,8/3-8/4
1991 (10)
7/11-7/14,8/19,9/5-9/8,9/13-9/15,10/3-10/6
10/31-11/3,11/7-11/10,11/14-11/17,12/5-12/8,12/12-12/15
1992 (1)
3/20/92-4/22/92
1994 (2)
9/20-9/21,10/18-10/19
1996 (3)
3/1-4/14, 4/23-4/25,8/16-8/18
1997 (2)
1/10-1/13,4/18-4/20
1998 (3)
8/7-8/8,8/21-8/22,9/10
1999 (4)
1/8-1/10,3/19-3/21,9/29-10/3,11/18-11/21
2000 (3)
5/18-5/21,8/29-9/5,9/29-10/5
2001 (9)
2/1-2/4,3/5-3/7,4/6-4/8,6/11-6/17,7/19-7/22
8/16-8/19,10/11-10/14,11/8-11/11,12/6-12/9
2002 (11)
2/4-2/6,3/4-3/7,4/8-4/11,6/3-6/6,6/17-6/19,7/9-7/11
7/22-7/24,8/5-8/8,9/5-9/8,10/14-10/17,10/21-10/22
2003 (8)
2/3-2/6,3/20-3/24,4/7-4/9,5/5-5/7,7/7-7/9,10/6-10/8,11/3-11/5,12/11-12/14
2004 (10)
1/12-1/15,2/9-2/12,3/8-3/11,5/2-5/4,7/8-7/11
8/9-8/11,9/13-9/15,10/11-10/14,11/7-11/10,12/5-12/8
2005 (11)
1/3-1/5,2/7-2/10,3/7-3/10,4/2-4/5,4/4-4/7,7/7-7/10
8/8-8/11,9/12-9/15,10/13-10/16,11/7-11/10,12/5-12/8
2006 (11)
1/9-1/12,2/6-2/9,3/13-3/16,4/10-4/13,5/8-5/11,7/6-7/9
8/7-8/10,9/7-9/10,10/9-10/12,11/9-11/12,12/11-12/14
2007 (3)
1/15-1/18,2/1-2/4,3/11-3/14
2009 (8)
1/21-1/23,2/9-2/12,2/19-2/21,2/21-2/23,2/24-2/26,3/13-3/15,6/5-6/7,6/16-6/19
Gallup polls listed at http://www.ropercenter.uconn.edu/data_access/data/presidential_approval.
html in July 2013 that do not have usable micro data in the Roper Center archive.
A5 Table A.4: No. of Obs. (Surveys) with Party Identification by Quarter
Quarter
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1
6,276
(4)
6,131
(4)
7,443
(5)
8,918
(5)
6,154
(4)
7,710
(5)
4,592
(3)
5,049
(3)
6,502
(3)
7,736
(3)
2
4,623
(3)
5,747
(4)
6,084
(4)
7,975
(4)
7,761
(5)
4,547
(3)
6,292
(4)
7,351
(4)
12,591
(5)
9,215
(4)
3
6,225
(4)
6,262
(4)
5,848
(4)
10,714
(5)
6,131
(4)
7,616
(5)
7,005
(3)
14,752
(7)
6,763
(3)
6,875
(3)
4
5,987
(4)
1963
4,468
(3)
1964
6,051
(4)
1965
8,944
(5)
1966
4,532
(3)
1967
4,514
(3)
1968
7,422
(4)
1969
6,497
(3)
1970
6,128
(3)
1971
8,014
(3)
1972
1
6,889
(3)
15,263
(6)
9,452
(5)
12,977
(5)
6,205
(3)
4,503
(3)
7,675
(5)
9,281
(6)
4,634
(3)
6,046
(4)
2
10,626
(5)
15,555
(6)
11,204
(5)
10,144
(4)
12,200
(5)
7,653
(5)
7,701
(5)
10,730
(7)
9,570
(6)
12,892
(10)
3
5,605
(3)
10,842
(4)
10,566
(4)
8,925
(4)
9,932
(4)
7,563
(5)
10,329
(6)
7,544
(5)
4,613
(3)
6,029
(4)
4
11,162
(5)
1973
9,482
(4)
1974
9,586
(4)
1975
9,760
(4)
1976
6,365
(4)
1977
4,632
(3)
1978
9,318
(6)
1979
6,194
(4)
1980
6,156
(4)
1981
4,482
(3)
1982
1
6,145
(4)
11,015
(7)
9,307
(6)
12,404
(8)
10,770
(7)
9,233
(6)
10,721
(7)
9,527
(6)
6,339
(4)
6,121
(4)
2
9,281
(6)
11,739
(9)
9,777
(8)
10,286
(7)
13,726
(9)
10,712
(7)
10,669
(7)
10,939
(7)
9,193
(6)
10,838
(7)
3
7,609
(5)
7,816
(5)
7,755
(5)
6,198
(4)
7,564
(5)
13,969
(8)
10,903
(7)
6,288
(4)
7,699
(5)
7,580
(5)
4
9,383
(6)
1983
7,823
(5)
1984
9,213
(6)
1985
7,715
(5)
1986
10,609
(7)
1987
6,193
(4)
1988
9,226
(6)
1989
6,249
(4)
1990
7,666
(5)
1991
6,123
(4)
1992
1
7,742
(5)
6,231
(4)
8,964
(7)
3,711
(4)
5,974
(7)
5,687
(6)
6,907
(6)
7,591
(7)
21,124
(23)
9,210
(9)
2
10,701
(7)
7,340
(6)
10,055
(8)
5,626
(5)
9,818
(6)
4,032
(2)
14,214
(15)
7,972
(9)
11,735
(12)
11,303
(10)
3
10,774
(7)
10,848
(7)
7,023
(6)
4,657
(5)
5,523
(6)
3,031
(3)
9,178
(9)
17,293
(17)
10,894
(10)
6,622
(6)
4
6,066
(4)
1993
6,052
(4)
1994
3,136
(3)
1995
7,907
(7)
1996
7,170
(8)
1997
5,110
(5)
1998
8,027
(8)
1999
16,956
(17)
2000
2,786
(3)
2001
6,427
(8)
2002
1
14,693
(17)
12,385
(13)
7,021
(7)
6,076
(6)
8,794
(9)
10,620
(12)
12,990
(14)
12,331
(11)
7,606
(8)
6,705
(7)
2
11,268
(12)
10,132
(10)
11,631
(11)
11,493
(11)
5,000
(5)
4,697
(5)
12,493
(12)
8,396
(9)
5,696
(6)
8,869
(9)
3
12,917
(14)
8,996
(9)
12,873
(13)
19,433
(31)
7,974
(8)
14,205
(17)
8,954
(8)
19,193
(35)
6,464
(7)
8,442
(9)
4
10,776
(11)
2003
10,742
(10)
2004
8,514
(9)
2005
29,047
(39)
2006
6,510
(7)
2007
17,912
(19)
2008
5,721
(6)
2009
38,223
(48)
2010
4,881
(5)
2011
7,779
(7)
2012
1
14,779
(16)
7,053
(7)
8,190
(9)
5,030
(5)
5,030
(5)
11,134
(8)
33,682
(28)
19,218
(17)
20,749
(18)
16,377
(14)
2
8,767
(10)
5,013
(5)
8,959
(9)
6,841
(7)
6,042
(6)
7,532
(7)
23,239
(21)
16,393
(16)
11,871
(12)
19,332
(15)
3
4,021
(4)
6,220
(6)
9,722
(11)
4,016
(4)
6,084
(6)
16,256
(15)
17,809
(14)
13,909
(12)
11,447
(11)
16,103
(15)
4
6,686
(7)
9,053
(8)
10,194
(12)
4,534
(4)
7,101
(7)
17,470
(15)
11,202
(10)
12,426
(9)
16,142
(14)
19,535
(19)
A6 Table A.5: No. of Obs. (Surveys) with Ideology by Quarter
Quarter
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
4,859
(4)
2
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
4,320
(4)
3
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
3,180
(3)
4
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
0
(0)
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
1
3,011
(3)
6,098
(6)
4,030
(4)
5,028
(5)
4,106
(4)
7,973
(8)
9,790
(10)
12,331
(11)
6,945
(7)
6,705
(7)
2
6,280
(7)
3,008
(3)
4,830
(5)
6,039
(6)
3,970
(4)
4,031
(4)
11,313
(10)
7,172
(7)
5,696
(6)
8,869
(9)
3
5,885
(6)
3,034
(3)
6,063
(6)
16,364
(28)
2,837
(3)
7,839
(8)
8,946
(8)
19,193
(35)
5,069
(5)
8,442
(9)
4
6,289
(6)
8,078
(7)
3,160
(3)
27,050
(37)
5,911
(6)
10,650
(12)
5,060
(5)
35,104
(44)
4,881
(5)
7,779
(7)
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
1
12,063
(12)
7,053
(7)
5,943
(6)
5,030
(5)
5,030
(5)
11,134
(8)
33,682
(28)
14,239
(17)
15,489
(18)
16,377
(14)
2
7,750
(8)
5,013
(5)
8,959
(9)
6,032
(6)
6,042
(6)
7,532
(7)
23,239
(21)
14,424
(16)
8,992
(11)
17,395
(15)
3
4,021
(4)
5,551
(5)
8,489
(9)
4,016
(4)
6,084
(6)
16,256
(15)
15,826
(14)
12,937
(12)
10,188
(11)
15,147
(15)
4
6,022
(6)
8,613
(7)
10,194
(12)
4,534
(4)
7,101
(7)
17,470
(15)
10,215
(10)
12,426
(9)
13,356
(14)
19,071
(19)
Tables A.6 and A.7 present the demographic variables we collected about respondents. We collected information about respondents’ gender, race and ethnicity, age, marital status, employment
status, religion and education. We also collected information about household income, what industry the household’s chief wage earner works in, and whether someone in the household belongs to a
union. Finally, we collected information about the state of residence and the community in which
the respondent resides. Unfortunately, not all of these variables are contained in every survey we
coded. To provide a general sense of when we observe different variables, Table A.8 presents the
percentage of responses in which we observe a given variable by presidential term.
Finally, Table A.9 presents the variables we collected about the survey design. Most Gallup polls
are designed to be a nationally representative sample of the voting-age population in the United
States. To deal with the fact that some types of individuals within this population are more likely to
respond than others, Gallup has used weights since it abandoned quota sampling in the aftermath
of incorrectly predicting the 1948 presidential election. How these weights are represented in the
A7 Table A.6: Description of Respondents’ Characteristics and Locality Variable Codings
Variable
Variable Name
Coding
Gender:
Male
Female
male
female
Yes = 1, No = 0
Yes = 1, No = 0
Race and Ethnicity:
White
Black
Hispanic
white
black
hispanic
Yes = 1, No = 0
Yes = 1, No = 0
Yes = 1, No = 0
Age
age
18 to 99
Married
married
Yes = 1, No = 0
Household Income:
Minimum Value
lower bound income
Dollars
Maximum Value
upper bound income
Dollars
(Top Coded = -1)
No Response
missing income
Yes = 1, No = 0
Union Household
unionHH
Yes = 1, No = 0
State of Residence
state
Gallup State Code
Place of Residence:
Minimum City Size
lower bound citysize
Population
Maximum City Size
upper bound citysize
Population
Lives on Farm
farm
Yes = 1, No = 0
Near City of Pop. 100,000+
near100k
Yes = 1, No = 0
Suburbs in City Size
andsub
Yes = 1, No = 0
Area Code
area
201 to 999
Congressional District
cd
1 to 53
Notes: -9 indicates missing value, -99 indicates variable not included in series.
A8 Table A.7: Description of Labor Market, Religion, and Education Variable Codings
Variable
Variable Name
Coding
Employed
employment
Full Time
Part Time
Not Employed
=
=
=
1
2
3
Industry of
Chief Wage Earner
industry
Farmer
Business
Clerical
Sales
Skilled
Unskilled
Service
Professional
Farm Laborer
Non-Farm Laborer
Non-Labor Force
Other
=
=
=
=
=
=
=
=
=
=
=
=
1
2
3
4
5
6
7
8
9
10
11
12
Religion
religion
Protestant
Catholic
Jewish
Other
=
=
=
=
1
2
3
4
Education
education
Not High School Graduate
High School Graduate
Technical College
Some College
College Graduate
=
=
=
=
=
1
2
3
4
5
Notes: -9 indicates missing value, -99 indicates variable not included in series.
A9 A10
2%
98%
0%
0%
85%
State of Residence
Area Code
Congressional District
City Size
63%
Union Household
Income
76%
Religion
0%
1%
Married
96%
99%
Education
Industry of Chief Wage Earner
96%
Age
Employment
0%
100%
Race
Hispanic
0%
100%
Gender
99%
19%
Partisan Identification:
Consider Yourself
Lean More to
Ideology
90%
0%
0%
Presidential Approval:
Job as President
Handling of Economy
Handling of Foreign Affairs
IKE
97%
0%
0%
97%
96%
99%
0%
38%
97%
4%
100%
99%
0%
100%
100%
0%
99%
1%
90%
0%
0%
JFK
& LBJ
100%
0%
0%
100%
97%
99%
1%
83%
98%
21%
100%
98%
0%
100%
100%
0%
100%
6%
76%
0%
0%
Nixon
& Ford
96%
0%
0%
100%
98%
98%
47%
50%
94%
51%
99%
99%
0%
100%
100%
0%
99%
5%
89%
1%
4%
Carter
90%
0%
0%
98%
82%
74%
73%
75%
82%
80%
100%
99%
44%
100%
100%
0%
100%
36%
91%
28%
26%
Ronald
Reagan
23%
0%
0%
99%
76%
23%
54%
23%
67%
58%
99%
99%
34%
100%
100%
7%
97%
52%
83%
0%
1%
G.H.W.
Bush
Table A.8: Variables Observed by Presidency
12%
0%
20%
99%
85%
13%
28%
8%
18%
32%
99%
99%
90%
100%
100%
69%
100%
95%
69%
15%
15%
Bill
Clinton
0%
78%
89%
100%
86%
0%
12%
8%
48%
49%
99%
99%
100%
100%
100%
95%
100%
97%
87%
23%
18%
G.W.
Bush
0%
0%
35%
100%
80%
0%
78%
10%
94%
98%
99%
98%
100%
100%
100%
91%
100%
100%
75%
8%
6%
Barack
Obama Table A.9: Description of Survey Variable Codings
Variable
Variable Name
Coding
Weighting:
Final Weight
final weight
Average Value
Sample Weight
weight
0 to 999
Times at Home
times
0 to 9
Duplicate Obs. in Raw Data
duplicates
0 to 26
Survey Contains Oversample
over
Yes = 1, No = 0
Oversample Unrepresentative:
Presidential Approval
Partisan Identification
Ideology
unrep approval
unrep partyid
unrep ideology
Yes = 1, No = 0
Yes = 1, No = 0
Yes = 1, No = 0
Survey Info:
Survey Code
series
Name of Series
Observation Number
obs num
Order in Raw Data
Start Date
start date
First Date in Field
End Date
end date
Last Date in Field
Survey Sponsor
survey
Gallup (In-Person)
Gallup (Telephone)
Newsweek
CNN/USA Today
Times Mirror
UBS
Other
=
1
=
=
=
=
=
=
=
1
2
3
4
5
6
7
Notes: -9 indicates missing value, -99 indicates variable not included in series.
raw data has varied over time. In earlier years, observations were duplicated in the raw data in
proportion to their weight (e.g., an observation with a weight of three would be placed in the
dataset three times). In later years, sample weights were provided with each observations. We
construct a common weighting variable, final weight, to use across all of the surveys; it has an
average value of one within each survey. Occasionally Gallup purposely oversampled a particular
group (e.g., African-Americans, State of the Union viewers). In such cases, we note whether our
weighting variable is able to reconstruct a representative sample. Finally, we code information
about the survey mode and the sponsor of the survey.
A11 7.2
Cross Validation
We use a leave-one-out cross validation procedure to select a bandwidth on the Epanechnikov
kernel function used to create a smoothed average of partisanship by date in Gallup data. The
cross-validation procedure is based on minimizing the mean squared difference between the actual
and predicted values of four different quantities in the 232 surveys conducted between 1975 and
1984, when the partisan gender gap was growing the fastest. We construct the average partisanship
level of males who graduated from college, females who graduated from college, males who did not
graduate from college, and females who did not graduate from college. For each of the 232 surveys,
we construct a predicted value for each of these four quantities at time ts using data from all of the
applicable surveys weighted with an Epanechnikov kernel function with a variety of bandwidths,
excluding the survey conducted at time ts . We then construct the mean squared difference between
the actual value and the predicted value of all four quantities at time ts . As Figure A.1 shows,
a bandwidth of 100 days minimizes the average mean squared difference between the actual and
predicted values of the four quantities over the 232 surveys.
.0012
Average Squared Forcast Error 1975 − 1984
.00125
.0013
.00135
Figure A.1: Leave-One-Out Cross Validation of Bandwidth
0
200
400
Bandwidth
A12
600
800 7.3
Checking for Structural Breaks
We look for any periods of rapid change in the partisan gender gap using Equation 2, which is
a standard parametric specification that tests for discontinuous changes in an outcome before and
after time t, with θ capturing the discontinuous change in gender gap among those survey after
time t (Imbens and Lemieux, 2008). The change in the gender gap from an additional year passing
prior to time t and after time t is captured by δ and δ + γ, respectively. Thus, δ + γ + θ capture
the total change in the gender gap between year t and year t + 1. To increase the plausibility of
the assumption that the effect of time on partisanship is locally linear, the sample is restricted to
only include surveys such that ts is within four years of t when estimating Equation 2.
P rtnshps,men − P rtnshps,women = α + δ(ts − t) + γ(ts − t)1(ts > t) + θ1(ts > t) + s
(2)
Figure A.2 illustrates the fitted values from estimating Equation 2 before and after 1964 and
1980. The figure shows the level and trajectory of the gender gap appears to be quite similar before
and after the 1964 and 1980 presidential campaigns.
Table A.10 presents estimates of Equation 2 with t equal to January 1 of every president election
year between 1960 and 1992 using Gallup in-person survey data. Column 5 shows that rate at
which the partisan gender gap was growing significantly increased in 1976. This suggests that much
of the growth in the partisan gender gap between 1976 and 1980 occurred prior to Ronald Reagan’s
1980 presidential campaign. However, some caution needs to be applied to this conclusion, as we
would expect one out of every twenty regressions to estimate a significant (p < .05) change even if
there was no change in the evolution of the partisan gender gap before and after the presidential
election year. We cannot reject the null of no difference in the trend before and after the presidential
election for the other eight presidential election years between 1960 and 1992.
Table A.11 presents similar estimates with t equal to January 1 of every president election
year between 1992 and 2008 using Gallup Phone survey data. The first two columns suggest some
instability in the dynamics of the partisan gender gap in the 1990s, as the partisan gender gap
A13 −.05
Difference in Females’ and Males’ Partisanship Level
0
.05
.1
Figure A.2: Changes in Partisan Gender Gap near 1964 and 1980 Elections (Gallup)
01jan1962
01jan1964
Survey Date
01jan1966
Partisan Gender Gap in Survey (1959−1963)
Partisan Gender Gap in Survey (1964−1968)
Linear Regression (1959−1963)
Linear Regression (1964−1968)
01jan1968
−.1
Difference in Females’ and Males’ Partisanship Level
−.05
0
.05
01jan1960
01jan1976
01jan1978
01jan1980
Survey Date
01jan1982
Partisan Gender Gap in Survey (1975−1979)
Partisan Gender Gap in Survey (1980−1984)
Linear Regression (1975−1979)
Linear Regression (1980−1984)
01jan1984
Notes: Lines represent the best linear fit of the difference in men’s and women’s partisanship level in polls from 1960-1963 and 1964-1967 (top figure) and 1976-1979 and 1980-1983
(bottom figure).
A14 A15
0.016
(0.006)
-0.003
(0.010)
-0.002
(0.002)
-0.002
(0.005)
0.840
106
(1)
1/1/60
-0.003
(0.012)
-0.002
(0.013)
-0.004
(0.005)
0.001
(0.005)
0.896
97
(2)
1/1/64
-0.014
(0.004)
-0.001
(0.008)
-0.003
(0.002)
0.005
(0.003)
0.365
128
(3)
1/1/68
-0.007
(0.006)
0.000
(0.007)
0.002
(0.003)
0.000
(0.003)
0.998
158
(4)
1/1/72
0.002
(0.004)
-0.002
(0.006)
0.002
(0.002)
-0.007
(0.003)
0.026
182
(5)
1/1/76
-0.020
(0.004)
-0.002
(0.007)
-0.005
(0.002)
0.002
(0.003)
0.801
178
(6)
1/1/80
-0.035
(0.005)
0.009
(0.008)
-0.003
(0.002)
0.000
(0.005)
0.524
117
(7)
1/1/84
-0.039
(0.014)
-0.016
(0.018)
-0.003
(0.005)
0.009
(0.007)
0.255
73
(8)
1/1/88
-0.031
(0.011)
-0.005
(0.015)
0.006
(0.005)
-0.006
(0.007)
0.654
74
(9)
1/1/92
Notes: Each column represents a separate regression of Equation 2 based on the specified date of a structural break.
Regressions include all polls conducted within four years of the specified data. Robust standard errors reported in
parentheses.
Survey Date X
Survey After Structural Break (γ)
p-value on Ho : θ = γ = 0
N
Survey After Structural Break (θ)
Survey Date (δ)
Constant (α)
Date of Structural Break
Table A.10: Changes in Partisan Gender Gap Near Presidential Election Years (Gallup In Person) Table A.11: Changes in Partisan Gender Gap Near Presidential Election Years (Gallup Phone)
Date of Structural Break
Constant (α)
Survey Date (δ)
Survey After Structural Break (θ)
Survey Date X
Survey After Structural Break (γ)
p-value on Ho : θ = γ = 0
N
(1)
1/1/92
(2)
1/1/96
(3)
1/1/00
(4)
1/1/04
(5)
1/1/08
-0.053
(0.005)
0.010
(0.007)
-0.002
(0.003)
-0.007
(0.004)
0.049
247
-0.077
(0.005)
0.007
(0.007)
-0.008
(0.002)
0.010
(0.003)
0.004
325
-0.064
(0.004)
-0.003
(0.006)
0.001
(0.002)
-0.002
(0.003)
0.662
328
-0.070
(0.005)
0.009
(0.008)
-0.001
(0.002)
0.000
(0.003)
0.484
246
-0.064
(0.006)
-0.001
(0.007)
-0.001
(0.003)
0.004
(0.003)
0.416
332
Notes: Each column represents a separate regression of Equation 2 based on the
specified date of a structural break. Regressions include all polls conducted within
four years of the specified data. Robust standard errors reported in parentheses.
began contracting somewhat in the late 1990s.
A16 7.4
Polarization Across Time
Another potential explanation for the increased assessments of polarization displayed in Figure
1 is that ANES started asking respondents about the parties positions on more polarizing issues
during the 1980s. To rule out this explanation, Figure A.3 shows the evolution of polarization over
time after we control for differences in issues on which respondents were surveyed. We construct
this graph by regressing the difference in assessments of the Republican and Democratic issue
positions on a set of issue fixed effects and a set of year dummy variables separately for college
graduates and non-college graduates. The year fixed effects are identified in this regression by
variation over years in assessments of polarization on the same issue position questions. We added
the estimated year fixed effect to the averaged estimated issue-position fixed effects to construct a
measure of polarization in a given year for a given educational group. The trends in Figure A.3 are
similar, although slightly smoother, to what were observed in Figure 1 when we did not control
for differences in the issues on which respondents were surveyed over time.
Differences in Ideological
Assessments of Parties’ Issue Positions
0
.1
.2
.3
.4
.5
Figure A.3: Assessments of Polarization in the Parties’ Issue Positions Holding Issues Constant
(ANES 1970 - 1998)
1968
1972
1976
1980
1984 1988
Survey Year
1992
1996
College Graduates
Non−College Graduates
Black Lines Indicate 95% Confidence Intervals
Grey Lines Indicate Linear Trend
A17
2000
2004 7.5
Leaners
Because Gallup does not always follow up with Independents about their leanings, our baseline
specification treats partisan leaners as Independents. In some years of the ANES, grouping leaners
with Independents reduces the size of the partisan gender gap (Norrander, 1999, 571). Thus, it
is important to examine the robustness of the partisan gender gap in Gallup data to alternative
treatments of Independents.
Figure A.4 presents the evolution of the partisan gender gap separately for Republicans and
Independents. We observe that when the partisan gender gap first emerged during the 1980s, men
and women were equally likely to identify as Republicans in Gallup data. The partisan gender
gap emerged because men were more likely to identify as Independents, and less likely to identify
as Democrats, than women. Since the 1990s, men have been slightly more likely to identify as
Republicans than women. But the largest difference between men and women still remains that
men are more likely to identify and Independents, less likely to identify as Democrats, than women.
Table A.12 examines how estimates of the partisan gender gap change when Independent
leaners are classified as partisans on in-person Gallup surveys that followed up with Independents
about their leanings. The partisan gender gap is generally larger when leaners are classified as
partisans, and this pattern is more pronounced among college graduates. This suggests that our key
findings about when the partisan gender gap emerged and the presence of educational heterogeneity
in the magnitude of the partisan gender gap would hold if we were able to classify Independent
leaners as partisans in the full sample.
A18 .8
Percent Affiliated
.4
.6
Figure A.4: Locally Weighted Average of Partisanship by Gender in Gallup Surveys
Male Rep. (In Person)
Female Rep. (In Person)
Male Rep. & Ind. (In Person)
Female Rep. & Ind. (In Person)
Male Rep. (Phone)
Female Rep. (Phone)
Male Rep. & Ind. (Phone)
Female Rep. & Ind. (Phone)
01jan1953
01jan1961
01jan1969
01jan1977
01jan1985
01jan1993
01jan2001
01jan2009
01jan1957
01jan1965
01jan1973
01jan1981
01jan1989
01jan1997
01jan2005
01jan2013
Survey Date
.2
A19 A20
0.001
(0.011)
0.015
(0.003)
-0.014
(0.011)
0.013
(0.010)
0.014
(0.003)
-0.001
(0.010)
(1)
53-56
103971
-0.040
(0.032)
0.028
(0.010)
-0.069
(0.034)
-0.015
(0.029)
0.019
(0.009)
-0.034
(0.030)
(2)
57-60
10556
0.051
(0.029)
0.008
(0.009)
0.044
(0.030)
0.045
(0.027)
0.008
(0.009)
0.037
(0.028)
(3)
61-64
10338
(8)
81-84
13760
as Partisans
-0.102
-0.120
(0.027) (0.019)
0.008
-0.035
(0.011) (0.009)
-0.110
-0.086
(0.029) (0.021)
(7)
77-80
7856
Classify Leaners
0.025
(0.020)
0.014
(0.008)
0.012
(0.022)
(6)
73-76
13863
Independents
-0.061
-0.091
(0.024) (0.017)
0.010
-0.032
(0.010) (0.008)
-0.070
-0.059
(0.026) (0.018)
(5)
69-72
0
Classify Leaners as
0.011
(0.018)
0.005
(0.007)
0.006
(0.019)
(4)
65-68
0
-0.104
(0.012)
-0.019
(0.006)
-0.085
(0.013)
-0.079
(0.010)
-0.012
(0.005)
-0.067
(0.012)
(9)
85-88
31354
-0.105
(0.011)
-0.022
(0.006)
-0.083
(0.012)
-0.077
(0.010)
-0.018
(0.005)
-0.059
(0.011)
(10)
89-92
32337
-0.098
(0.009)
-0.027
(0.005)
-0.071
(0.010)
-0.073
(0.008)
-0.023
(0.004)
-0.051
(0.009)
(11)
93-96
45246
Notes: Cells report estimate and standard error on the partisan gender gap by education level implied from a regression that
includes a female dummy, a college graduate dummy, the interaction between the female dummy and the college graduate dummy,
and survey fixed effects. The sample is restricted only to those surveys that ask Independents about whether they lean towards
a party. Robust standard errors are reported in parentheses.
Difference
Non-College Graduates
College Graduates
Difference
Non-College Graduates
College Graduates
Year of Survey
N
Table A.12: Leaners and Education Heterogeneity in the Partisan Gender Gap (Gallup In Person) 7.6
Comparing the Pooled Gallup Dataset to the ANES and GSS
Figure A.5 presents the size of the gender gap over time in the Gallup dataset compared to the
ANES and GSS. The dark line is the difference between men and women’s party identification over
time in Gallup in-person interviews, using the same Epanechnikov kernel function as in Subsections
5.4 and 5.5. The lighter line is the same, but for Gallup phone interviews. We overlay the gender
gap in every ANES and GSS survey in this period. The gap in the ANES is the black dots and the
gap in the GSS is the lighter dots, both with dashed lines for the 95% confidence intervals. When
viewed together, all three types of surveys (Gallup, ANES and GSS) appear to be following the
same trend. In 20 out of 22 ANES and 19 out of 21 GSS surveys, the point estimate of the gender
gap in the in-person Gallup data is within the benchmark survey’s 95% confidence interval. We
have fewer cases in which we can compare partisan gender gap in the phone Gallup, but the points
estimate is contained within the ANES 95% confidence interval in 8 out of 9 cases and the CSS
95% confidence interval in 13 of 14 cases. This is roughly the proportion one would expect to fall
outside a 95% confidence interval due to sampling error. This reassuringly suggests that, despite
the differences in question wording, the ANES, Gallup, and GSS data capture a similar construct.
The main difference is that the smaller sample sizes in the ANES and GSS produce much larger
confidence intervals which are unable to detect a significant gender gap in many years where it is
present.
An implication of Figure A.5 is that the development of the partisan gender gap is a smoother
process than one might conclude from the ANES or GSS. Although the gap is only occasionally
statistically significant, men tend to be slightly more Republican than women throughout the
1960s and 1970s. Since 1977, the gender gap remains distinguishable from zero at the 95% level,
two-tailed, with few dramatic shifts. Several surges and swoons in the gap’s size in the ANES
or GSS, which one might be tempted to imbue with political importance if one considered these
surveys alone, now appear to be mere sampling variation around the gradual trend. For instance,
it appears in the ANES that, after disappearing in 1958 and 1960, the old reverse gender gap
re-emerged for the last time in 1962. Years such as 1968 and 1974 stand out as points when the
A21 .05
0
−.05
−.1
−.15
Women’s Minus Men’s Partisanship Level
.1
Figure A.5: Comparing Partisan Gender Gap in Gallup, ANES, and GSS Surveys
01jan1953
01jan1961
01jan1969
01jan1977
01jan1985
01jan1993
01jan2001
01jan2009
01jan1957
01jan1965
01jan1973
01jan1981
01jan1989
01jan1997
01jan2005
01jan2013
Survey Date
Gallup In Person
Gallup Phone
ANES
ANES 95% CI
GSS
GSS 95% CI
modern gap first emerged at notable sizes (and marginal statistical significance). More recently,
1982, 1994, and 1996 stand out for particularly large gender gaps. It appears in the GSS that the
modern gender gap first emerged in 1976 (in contrast to the ANES), and then dissipated until
emerging again in the mid-1980s, then temporarily shrank in 1993-1994. But all this appears to
be largely sampling variation.
Figure A.6 shows the difference in the size of the partisan gender gap between college graduates
and non-college graduates in the ANES, Gallup, and GSS datasets. The solid black line shows this
difference in Gallup in-person interviews, again using the same Epanechnikov kernel function. The
college graduate versus non-college graduate difference in the ANES and GSS is shown with the
dark and light dots, respectively, with the dashed lines showing the 95% confidence intervals. The
comparison of datasets illustrates that the large sample size of the Gallup dataset is necessary
to detect the educational difference; the smaller sample sizes in the ANES and GSS produce
confidence intervals that are too large to detect this educational variation in the gender gap.
The Gallup point estimate is within the 95% confidence interval of every ANES and GSS survey
A22 .25
.2
.15
.1
.05
0
−.35 −.3 −.25 −.2 −.15 −.1 −.05
Difference in Women’s Minus Men’s Partisanship Level
College Graduates and Non−College Graduates
.3
.35
Figure A.6: Comparing Difference in Partisan Gender Gap by Educational Attainment in Gallup,
ANES, GSS Surveys
01jan1953
01jan1961
01jan1969
01jan1977
01jan1985
01jan1993
01jan1957
01jan1965
01jan1973
01jan1981
01jan1989
01jan1997
Survey Date
Gallup In Person
ANES
GSS
GSS 95% CI
ANES 95% CI
except the 1966 ANES. This pattern suggests again that all three surveys are measuring the same
underlying pattern. The difference is simply that, if you look only at the ANES or GSS, and use
conventional statistical significance levels, you do not have precision enough to know that the
partisan gender gap was consistently larger among the college educated since the early 1970s.
7.7
Additional Tables and Figures
A23 .6
Average Partisanship Level in Survey
Higher Values = More Republican
.4
.5
Figure A.7: Partisanship Level by Gender in Each Gallup Survey
Males (In Person)
Males (Phone)
Females (In Person)
Females (Phone)
01jan1953
01jan1961
01jan1969
01jan1977
01jan1985
01jan1993
01jan2001
01jan2009
01jan1957
01jan1965
01jan1973
01jan1981
01jan1989
01jan1997
01jan2005
01jan2013
Survey Date
.3
A24 Table A.13: Partisan Gender Gap by Age, Race, and Education Across Time (Gallup)
Age
Black
18-39
40-59
60+
Yes
No
No HS
Degree
0.011
(0.004)
[45278]
0.013
(0.004)
[40576]
-0.010
(0.004)
[55255]
-0.026
(0.004)
[52669]
-0.023
(0.004)
[45790]
-0.020
(0.003)
[59796]
-0.020
(0.003)
[68326]
-0.042
(0.004)
[54984]
-0.048
(0.008)
[13546]
-0.053
(0.007)
[15158]
-0.041
(0.007)
[17872]
0.021
(0.004)
[40071]
0.034
(0.005)
[39691]
0.009
(0.004)
[56428]
-0.001
(0.004)
[53610]
-0.004
(0.005)
[40285]
0.003
(0.004)
[43056]
-0.011
(0.004)
[44758]
-0.021
(0.005)
[35129]
-0.026
(0.010)
[9150]
-0.035
(0.009)
[10855]
-0.037
(0.008)
[14106]
0.023
(0.007)
[19742]
0.040
(0.006)
[22477]
0.021
(0.006)
[34411]
0.000
(0.006)
[33344]
0.008
(0.006)
[26484]
0.027
(0.006)
[30153]
-0.007
(0.005)
[36528]
-0.018
(0.005)
[32234]
-0.012
(0.010)
[8698]
-0.018
(0.010)
[10027]
-0.029
(0.008)
[13110]
-0.029
(0.009)
[8616]
-0.001
(0.009)
[10107]
0.006
(0.007)
[15432]
-0.018
(0.007)
[11720]
-0.010
(0.007)
[9083]
-0.015
(0.006)
[14164]
-0.011
(0.005)
[16292]
-0.021
(0.006)
[13446]
-0.022
(0.013)
[3067]
-0.053
(0.012)
[3528]
-0.025
(0.010)
[5248]
0.013
(0.003)
[99080]
0.023
(0.003)
[97818]
0.002
(0.003)
[132252]
-0.012
(0.003)
[129947]
-0.007
(0.003)
[104685]
0.001
(0.003)
[121749]
-0.013
(0.002)
[134910]
-0.031
(0.003)
[109516]
-0.036
(0.006)
[28433]
-0.036
(0.005)
[32706]
-0.037
(0.004)
[40141]
0.000
(0.004)
[58104]
0.019
(0.004)
[55702]
0.002
(0.004)
[72177]
-0.019
(0.004)
[61132]
-0.007
(0.004)
[41765]
0.010
(0.004)
[42737]
-0.006
(0.004)
[42594]
-0.008
(0.005)
[29945]
-0.029
(0.011)
[6690]
-0.024
(0.011)
[6820]
-0.006
(0.010)
[7663]
Education
HS
Some
Degree
College
College
Degree
0.019
(0.005)
[30763]
0.021
(0.005)
[32985]
0.011
(0.005)
[47434]
0.004
(0.004)
[49749]
0.001
(0.004)
[41808]
0.010
(0.004)
[51078]
0.001
(0.004)
[58638]
-0.017
(0.004)
[47599]
-0.005
(0.009)
[12186]
-0.025
(0.008)
[14144]
-0.025
(0.007)
[18054]
0.012
(0.009)
[8466]
0.012
(0.010)
[9032]
-0.004
(0.009)
[13846]
-0.042
(0.008)
[14828]
-0.023
(0.008)
[14296]
-0.035
(0.007)
[19412]
-0.048
(0.006)
[23503]
-0.071
(0.006)
[22956]
-0.079
(0.012)
[6542]
-0.075
(0.011)
[7989]
-0.073
(0.009)
[10454]
Survey Year:
1953-1956
1957-1960
1961-1964
1965-1968
1969-1972
1973-1976
1977-1980
1981-1984
1985-1988
1989-1992
1993-1996
0.026
(0.009)
[9538]
0.050
(0.009)
[9515]
0.014
(0.008)
[13848]
0.030
(0.008)
[15604]
0.008
(0.007)
[15493]
0.002
(0.006)
[22179]
-0.004
(0.006)
[26124]
-0.030
(0.006)
[22132]
-0.045
(0.012)
[5936]
-0.033
(0.011)
[7071]
-0.045
(0.010)
[9075]
Notes: Each cell presents the coefficient, standard error, and the number of observations with
a given characteristic over the specified time period. The coefficient and standard error are
on the interaction between the given characteristic and a female indicator from a regression
of partisanship on gender, a set of characteristics that partition the sample, the interaction
between gender and a set of characteristics that partition the sample, and a survey fixed
effects. Robust standard errors reported in parentheses and the sample size is reported in
brackets.
A25 Table A.14: Partisan Gender Gap by Religious, Marital, and Household Union Status (Gallup)
Prst.
Religion
Cath.
Jwsh.
Oth.
Married
Yes
No
Yes
Union HH
No
Survey Year:
1953-1956
1957-1960
1961-1964
1965-1968
1969-1972
1973-1976
1977-1980
1981-1984
1985-1988
1989-1992
1993-1996
0.020
(0.004)
[51451]
0.027
(0.004)
[62059]
0.015
(0.003)
[99835]
-0.002
(0.003)
[92449]
-0.002
(0.004)
[70515]
0.009
(0.003)
[80801]
-0.005
(0.003)
[85627]
-0.021
(0.004)
[68128]
-0.035
(0.007)
[18422]
-0.042
(0.007)
[16779]
N/A
N/A
[0]
-0.004
(0.006)
[17096]
0.008
(0.006)
[22138]
-0.030
(0.005)
[33545]
-0.032
(0.005)
[33630]
-0.023
(0.005)
[29240]
-0.017
(0.004)
[36339]
-0.029
(0.004)
[40375]
-0.052
(0.005)
[32966]
-0.038
(0.010)
[8645]
-0.051
(0.010)
[8728]
N/A
N/A
[0]
-0.056
(0.015)
[2402]
-0.080
(0.014)
[2942]
-0.082
(0.012)
[4236]
-0.068
(0.012)
[4052]
-0.063
(0.013)
[3059]
-0.069
(0.013)
[3331]
-0.077
(0.013)
[3360]
-0.058
(0.015)
[2847]
-0.078
(0.035)
[707]
-0.055
(0.036)
[651]
N/A
N/A
[0]
0.000
(0.018)
[1997]
0.002
(0.014)
[3918]
0.015
(0.012)
[6258]
-0.006
(0.011)
[6239]
-0.014
(0.009)
[7819]
-0.023
(0.007)
[13790]
-0.033
(0.007)
[12668]
-0.032
(0.007)
[14181]
-0.031
(0.015)
[3394]
-0.034
(0.013)
[4067]
N/A
N/A
[0]
N/A
N/A
[0]
0.078
(0.028)
[1281]
N/A
N/A
[0]
0.013
(0.011)
[8561]
-0.012
(0.005)
[34713]
-0.016
(0.016)
[3425]
-0.014
(0.004)
[53117]
-0.011
(0.003)
[82099]
-0.013
(0.006)
[21378]
-0.024
(0.006)
[21694]
-0.020
(0.006)
[27664]
N/A
N/A
[0]
0.022
(0.054)
[319]
N/A
N/A
[0]
0.040
(0.021)
[2237]
-0.001
(0.008)
[11136]
0.013
(0.025)
[1218]
-0.028
(0.005)
[24587]
-0.059
(0.004)
[40863]
-0.069
(0.009)
[10119]
-0.065
(0.008)
[12031]
-0.054
(0.007)
[17725]
0.017
(0.006)
[20192]
0.026
(0.007)
[16760]
-0.015
(0.009)
[11530]
0.002
(0.007)
[15511]
0.000
(0.006)
[19034]
0.007
(0.005)
[32592]
0.001
(0.006)
[17714]
-0.001
(0.006)
[24598]
-0.020
(0.012)
[5973]
0.002
(0.012)
[5684]
N/A
N/A
[0]
0.007
(0.004)
[52452]
0.013
(0.004)
[46040]
-0.008
(0.006)
[33778]
-0.020
(0.004)
[47646]
-0.023
(0.004)
[57159]
-0.010
(0.003)
[100983]
-0.032
(0.004)
[57991]
-0.043
(0.003)
[91044]
-0.042
(0.006)
[25094]
-0.056
(0.006)
[26673]
N/A
N/A
[0]
Notes: Each cell presents the coefficient, standard error, and the number of observations
with a given characteristic over the specified time period. The coefficient and standard
error are on the interaction between the given characteristic and a female indicator
from a regression of partisanship on gender, a set of characteristics that partition the
sample, the interaction between gender and a set of characteristics that partition the
sample, and a survey fixed effects. Robust standard errors reported in parentheses and
the sample size is reported in brackets.
A26 Table A.15: Partisan Gender Gap by Income and Labor Market Status (Gallup)
Avg. HH Income in
Chief Wage Earner’s Industry
High
Med.
Low
No Job
Median HH
Income
Above
Below
Employment
Full
Part
None
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
-0.035
(0.037)
[699]
-0.038
(0.005)
[34104]
-0.061
(0.004)
[50141]
-0.056
(0.008)
[14451]
-0.068
(0.007)
[16598]
-0.059
(0.006)
[21036]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
0.165
(0.081)
[142]
-0.010
(0.011)
[7010]
0.007
(0.009)
[11852]
-0.024
(0.018)
[3304]
-0.009
(0.016)
[3655]
-0.025
(0.013)
[4980]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
N/A
N/A
[0]
0.030
(0.036)
[686]
0.006
(0.005)
[31548]
0.010
(0.004)
[51020]
0.015
(0.009)
[13085]
-0.001
(0.008)
[13951]
-0.003
(0.007)
[18728]
Survey Year:
1953-1956
1957-1960
1961-1964
1965-1968
1969-1972
1973-1976
1977-1980
1981-1984
1985-1988
1989-1992
1993-1996
0.013
(0.006)
[20514]
0.011
(0.006)
[21289]
-0.010
(0.006)
[31699]
-0.017
(0.005)
[32796]
-0.019
(0.005)
[27922]
-0.011
(0.005)
[32299]
-0.028
(0.004)
[41570]
-0.048
(0.005)
[32793]
-0.052
(0.010)
[9498]
-0.056
(0.010)
[9107]
-0.060
(0.009)
[11946]
0.010
(0.004)
[45102]
0.030
(0.004)
[46680]
0.006
(0.004)
[63193]
-0.009
(0.004)
[59509]
-0.004
(0.004)
[45834]
0.002
(0.004)
[54021]
-0.002
(0.004)
[52851]
-0.018
(0.005)
[37843]
-0.023
(0.009)
[10297]
-0.036
(0.009)
[10234]
-0.013
(0.008)
[13603]
-0.004
(0.005)
[26707]
0.011
(0.006)
[22505]
0.006
(0.006)
[27508]
-0.008
(0.007)
[21795]
-0.018
(0.007)
[15808]
-0.010
(0.007)
[16636]
-0.004
(0.006)
[18918]
-0.024
(0.007)
[14082]
-0.028
(0.015)
[3904]
-0.039
(0.015)
[3472]
-0.017
(0.012)
[4975]
0.031
(0.017)
[2778]
0.018
(0.008)
[12996]
0.000
(0.007)
[23173]
-0.013
(0.006)
[24072]
-0.008
(0.007)
[20337]
0.014
(0.006)
[26828]
-0.012
(0.005)
[32577]
-0.029
(0.006)
[25124]
-0.015
(0.012)
[6671]
-0.034
(0.012)
[6699]
-0.051
(0.010)
[9185]
N/A
N/A
[0]
0.003
(0.017)
[3241]
-0.001
(0.004)
[81176]
-0.010
(0.003)
[80645]
-0.005
(0.004)
[66405]
-0.002
(0.003)
[72148]
-0.012
(0.003)
[82055]
-0.028
(0.004)
[69260]
-0.032
(0.007)
[17652]
-0.034
(0.007)
[19995]
-0.042
(0.006)
[24865]
N/A
N/A
[0]
0.039
(0.022)
[2162]
0.013
(0.004)
[62330]
-0.003
(0.004)
[53792]
-0.004
(0.004)
[45478]
0.013
(0.004)
[57659]
-0.003
(0.003)
[66215]
-0.013
(0.004)
[50602]
-0.022
(0.008)
[13422]
-0.033
(0.007)
[15493]
-0.018
(0.006)
[19568]
Notes: Each cell presents the coefficient, standard error, and the number of observations with
a given characteristic over the specified time period. The coefficient and standard error are
on the interaction between the given characteristic and a female indicator from a regression
of partisanship on gender, a set of characteristics that partition the sample, the interaction
between gender and a set of characteristics that partition the sample, and a survey fixed
effects. Robust standard errors reported in parentheses and the sample size is reported in
brackets.
A27 Table A.16: Partisan Gender Gap by Region and Size of City of Residence (Gallup)
Region of Residence
New
Engl.
Mid.
Atl.
Cntrl.
-0.003
(0.010)
[7314]
0.031
(0.011)
[6657]
-0.001
(0.011)
[7551]
-0.011
(0.010)
[7480]
-0.018
(0.011)
[6253]
-0.001
(0.009)
[7934]
-0.031
(0.008)
[8782]
-0.034
(0.010)
[6650]
-0.020
(0.020)
[1630]
-0.041
(0.018)
[2163]
-0.040
(0.015)
[2875]
0.002
(0.006)
[25376]
0.005
(0.006)
[25284]
-0.014
(0.006)
[32129]
-0.015
(0.006)
[32553]
-0.003
(0.006)
[26951]
-0.009
(0.006)
[28250]
-0.013
(0.005)
[33393]
-0.034
(0.006)
[26171]
-0.047
(0.012)
[6314]
-0.035
(0.012)
[6956]
-0.036
(0.010)
[8816]
0.015
(0.005)
[32396]
0.021
(0.005)
[32554]
0.006
(0.005)
[42386]
0.005
(0.005)
[41167]
-0.013
(0.005)
[32606]
0.014
(0.004)
[38745]
-0.002
(0.004)
[42819]
-0.028
(0.005)
[33136]
-0.014
(0.010)
[8320]
-0.046
(0.009)
[9638]
-0.025
(0.008)
[11734]
South
Rocky
Mtn.
West
0.018
(0.005)
[26922]
0.030
(0.006)
[24125]
0.011
(0.005)
[36101]
-0.026
(0.005)
[38031]
-0.014
(0.005)
[29649]
-0.011
(0.004)
[37896]
-0.016
(0.004)
[41355]
-0.025
(0.005)
[33391]
-0.041
(0.010)
[9142]
-0.041
(0.009)
[10893]
-0.041
(0.008)
[14066]
-0.001
(0.014)
[3733]
0.005
(0.012)
[6116]
-0.043
(0.014)
[5114]
-0.001
(0.014)
[5172]
0.003
(0.014)
[4097]
0.014
(0.012)
[5126]
-0.004
(0.012)
[5344]
-0.017
(0.013)
[4911]
0.028
(0.024)
[1510]
0.002
(0.022)
[1740]
-0.012
(0.022)
[1614]
0.000
(0.008)
[11629]
0.031
(0.009)
[9744]
0.008
(0.008)
[16732]
-0.015
(0.008)
[17252]
0.004
(0.008)
[14207]
-0.006
(0.007)
[17960]
-0.027
(0.007)
[19495]
-0.041
(0.007)
[17114]
-0.067
(0.014)
[4584]
-0.051
(0.014)
[4840]
-0.062
(0.012)
[6284]
Size of City of Residence
Under
10k to
Over
10k
100k
100k
Survey Year:
1953-1956
1957-1960
1961-1964
1965-1968
1969-1972
1973-1976
1977-1980
1981-1984
1985-1988
1989-1992
1993-1996
0.006
(0.004)
[56369]
0.029
(0.004)
[54213]
-0.001
(0.003)
[82343]
0.003
(0.003)
[78233]
0.001
(0.004)
[61360]
0.011
(0.003)
[68246]
-0.012
(0.003)
[71934]
-0.023
(0.004)
[65403]
-0.045
(0.008)
[15050]
-0.052
(0.007)
[20510]
-0.058
(0.006)
[28746]
0.010
(0.007)
[15401]
0.046
(0.009)
[11088]
-0.008
(0.007)
[19561]
-0.007
(0.007)
[19496]
-0.005
(0.007)
[17824]
-0.004
(0.006)
[24999]
-0.007
(0.005)
[27499]
-0.038
(0.007)
[19280]
-0.032
(0.019)
[2557]
-0.045
(0.015)
[4170]
-0.035
(0.012)
[5337]
0.004
(0.004)
[40470]
0.001
(0.005)
[39420]
-0.005
(0.004)
[66880]
-0.017
(0.004)
[67274]
-0.014
(0.004)
[56357]
-0.008
(0.004)
[67404]
-0.013
(0.003)
[74153]
-0.028
(0.004)
[64824]
-0.023
(0.010)
[12892]
-0.015
(0.008)
[18481]
-0.008
(0.007)
[25687]
Notes: Each cell presents the coefficient, standard error, and the number of observations with
a given characteristic over the specified time period. The coefficient and standard error are
on the interaction between the given characteristic and a female indicator from a regression
of partisanship on gender, a set of characteristics that partition the sample, the interaction
between gender and a set of characteristics that partition the sample, and a survey fixed
effects. Robust standard errors reported in parentheses and the sample size is reported in
brackets.
A28 A29
-0.025
(0.004)
-0.004
(0.005)
-0.009
(0.004)
-0.007
(0.005)
-0.036
(0.007)
0.041
(0.008)
0.004
(0.007)
-0.006
(0.007)
-0.004
(0.010)
-0.005
(0.012)
-0.002
(0.009)
0.009
(0.010)
Female
-0.042
(0.008)
0.016
(0.009)
-0.010
(0.006)
0.003
(0.006)
0.041
(0.012)
-0.026
(0.014)
0.032
(0.009)
-0.004
(0.009)
-0.003
(0.018)
-0.004
(0.021)
-0.014
(0.013)
0.009
(0.013)
169393
(2)
Social
Welfare
-0.058
(0.010)
0.023
(0.012)
0.009
(0.008)
-0.035
(0.014)
-0.096
(0.019)
0.029
(0.021)
0.001
(0.011)
-0.004
(0.019)
0.029
(0.028)
-0.034
(0.030)
0.000
(0.016)
0.036
(0.026)
(3)
Foreign
Relations
74,
78, 02
43809
-0.019
(0.006)
-0.011
(0.009)
-0.025
(0.010)
0.019
(0.013)
-0.108
(0.010)
0.042
(0.014)
0.023
(0.014)
0.068
(0.016)
0.002
(0.015)
0.001
(0.021)
0.011
(0.020)
-0.016
(0.023)
88493
Race
02
(4)
0.037
(0.010)
-0.031
(0.013)
0.005
(0.012)
-0.042
(0.014)
-0.116
(0.015)
0.007
(0.019)
0.041
(0.015)
0.014
(0.016)
-0.081
(0.020)
0.042
(0.026)
0.001
(0.021)
0.033
(0.022)
(5)
Gender
Roles
70
86, 02
25459
-0.014
(0.009)
0.003
(0.012)
-0.014
(0.012)
0.033
(0.017)
0.142
(0.016)
-0.010
(0.021)
-0.042
(0.017)
0.031
(0.022)
0.007
(0.023)
0.022
(0.029)
0.067
(0.025)
-0.075
(0.030)
Abortion
70
74, 02
30024
(6)
-0.132
(0.035)
0.094
(0.025)
0.037
(0.015)
ERA
70-74,
82-08
4589
(7)
-0.006
(0.005)
-0.009
(0.006)
0.002
(0.006)
-0.006
(0.006)
0.000
(0.009)
-0.001
(0.012)
0.023
(0.010)
-0.035
(0.010)
-0.028
(0.014)
0.010
(0.017)
-0.033
(0.014)
0.041
(0.014)
27353
Ideology
78
(8)
-0.044
(0.011)
-0.004
(0.014)
0.006
(0.014)
-0.003
(0.015)
0.056
(0.021)
0.066
(0.026)
-0.030
(0.020)
-0.081
(0.021)
0.004
(0.029)
-0.016
(0.036)
-0.013
(0.029)
0.043
(0.029)
33114
-0.026
(0.010)
-0.003
(0.013)
-0.011
(0.011)
0.003
(0.013)
0.031
(0.019)
0.039
(0.023)
-0.052
(0.018)
-0.015
(0.018)
0.024
(0.025)
-0.057
(0.032)
0.035
(0.025)
-0.018
(0.025)
28998
(9)
(10)
Personal Finances
Last Year Next Year
02
Notes: Issue positions and ideology are recoded so that they range from 0 (“Most Liberal”) to 1 (“Most Conservative”). Economic evaluations
recoded so that they range from 0 (“Least Favorable”) to 1 (”Most Favorable”). “No Questions Asked” indicates years in which no questions
were asked on that specific issue dimension. All regressions also include year fixed effects and observations are weighted by their sample weight.
Robust standard errors that are clustered by respondent are reported in parentheses.
Female X
College Graduate
Female X
College Graduate X Year ≥ 1980
Female X
College Graduate X Year ≥ 1990
Female X
College Graduate X Year ≥ 2000
College Graduate X Year ≥ 2000
College Graduate X Year ≥ 1990
College Graduate X Year ≥ 1980
College Graduate
Female X Year ≥ 2000
Female X Year ≥ 1990
Female X Year ≥ 1980
514313
(1)
All Issue
Positions
N
No Questions Asked
Dependent Variable
Table A.17: Gender Differences in Issue Positions, Ideology, and Economic Evaluations by Education Attainment (ANES, 1970 1998) A30
0.012
(0.028)
-0.016
(0.019)
0.014
(0.015)
0.005
(0.013)
0.007
(0.011)
0.002
(0.020)
-0.041
(0.038)
0.009
(0.022)
-0.030
(0.016)
-0.031
(0.013)
-0.035
(0.011)
-0.032
(0.010)
-0.014
(0.015)
0.024
(0.031)
-0.039
(0.017)
-0.050
(0.014)
-0.049
(0.011)
-0.023
(0.010)
-0.051
(0.007)
-0.050
(0.010)
-0.015
(0.033)
-0.094
(0.021)
-0.080
(0.017)
-0.085
(0.016)
-0.076
(0.011)
-0.071
(0.010)
-0.090
(0.026)
College Graduates
(2)
(3)
(4)
63-72
73-82
83-92
-0.022
(0.043)
-0.050
(0.029)
-0.117
(0.027)
-0.065
(0.019)
-0.076
(0.016)
-0.063
(0.018)
(5)
93-97
0.037
(0.007)
0.032
(0.005)
0.034
(0.005)
0.023
(0.004)
0.029
(0.004)
0.004
(0.007)
(6)
53-62
0.003
(0.011)
0.007
(0.006)
0.008
(0.005)
0.013
(0.004)
0.006
(0.004)
-0.008
(0.004)
-0.008
(0.006)
0.024
(0.011)
0.012
(0.006)
0.011
(0.005)
0.012
(0.004)
0.005
(0.004)
-0.006
(0.004)
-0.019
(0.006)
0.036
(0.014)
-0.009
(0.008)
0.006
(0.007)
0.003
(0.008)
-0.014
(0.007)
-0.032
(0.006)
-0.067
(0.014)
Non-College Graduates
(7)
(8)
(9)
63-72
73-82
83-92
0.010
(0.018)
-0.012
(0.012)
-0.008
(0.013)
-0.016
(0.013)
-0.025
(0.010)
-0.013
(0.011)
(10)
93-97
Notes: Each cell presents an estimate and standard error of the partisan gender gap for individuals of the specified age
with the specified educational attainment over the specified time period. Sample excludes respondents between ages 18
and 24. Observations are weighted by their sample weight. Robust standard errors are reported in parentheses.
1960s
1950s
1940s
1930s
1920s
1910s
1900s
1890s
Decade of Birth:
1880s
Year of Survey
(1)
53-62
Table A.18: Partisan Gender Gap by Birth Cohort and Education Attainment Across Time (Gallup) .04
Partisan Gender Gap
−.08
−.04
0
78−82
83−87
88−92
93−97
98−02
Survey Year
Col. Grad. (In−Person)
Not Col. Grad. (In−Person)
03−07
−.12
−.12
Partisan Gender Gap
−.08
−.04
0
.04
Figure A.8: Comparing Partisan Gender Gap by Birth Cohort in In Person and Phone Surveys
(Gallup)
08−12
78−82
College Grad. (Phone)
Not College Grad. (Phone)
83−87
93−97
98−02
Survey Year
Col. Grad. (In−Person)
Not Col. Grad. (In−Person)
03−07
08−12
College Grad. (Phone)
Not College Grad. (Phone)
Partisan Gender Gap
−.08
−.04
0
78−82
83−87
88−92
93−97
98−02
Survey Year
Col. Grad. (In−Person)
Not Col. Grad. (In−Person)
03−07
−.12
Partisan Gender Gap
−.08
−.04
0
.04
Born 1930s
.04
Born 1920s
−.12
88−92
08−12
College Grad. (Phone)
Not College Grad. (Phone)
78−82
83−87
88−92
93−97
98−02
Survey Year
Col. Grad. (In−Person)
Not Col. Grad. (In−Person)
Born 1940s
Born 1950s
A31
03−07
08−12
College Grad. (Phone)
Not College Grad. (Phone)
Related Documents
- Critical Reflection on Tate's 'Life Between Islands'
- Inclusive Education Support System
- Inclusive Education Management
- Guidance and Counseling in Education
- Basic Concept of Graduation for Learners
- Digitally Leadership Educational Materials
- Nevada History and Constitution Assignment
- The History of International
- An Example of International Relations
- Lab 0: Reflecting on Movement
- Exploring Perspectives in Policing and Government Power
- Local Politics Term Paper Example
- A Day In The Life Of The President
- Supranational Organizations Project
- The Concept of Political Ideology and Its Significance
- The Evolution of Democratic Governance
- Global Leadership Study Resources
- School Financial Planning
- Unapologetic Authenticity: Decoding Barbara Walker's Self Portrait
- Response to “Wrong Place at Wrong Time”
Party Polarization, Ideological Sorting and the Emergence - Supplemental Appendix
of 31
Recommended Documents
⏰ Deadline pressure?
Get your assignment done in just 3 hours. Quick, easy, and available 24/7.
Report
Tell us what’s wrong with it:
Thanks, got it!
We will moderate it soon!
Free up your schedule!
Our EduBirdie Experts Are Here for You 24/7! Just fill out a form and let us know how we can assist you.
Take 5 seconds to unlock
Enter your email below and get instant access to your document