Content uploaded by Kim L. Fridkin
Author content
All content in this area was uploaded by Kim L. Fridkin on Sep 05, 2014
Content may be subject to copyright.
1
Citizens’ Understanding of their U.S. Senator:
How the Gender of the Senator and the Gender of Citizens Influence What People Know
Kim L. Fridkin
Patrick J. Kenney
Arizona State University
Paper presented at the Annual Meeting of the American Political Science Association, Seattle,
Washington, September 2011.
Note: This paper is draft chapter of a book in progress.
2
Citizens’ Understanding of their U.S. Senator:
How the Gender of the Senator and the Gender of Citizens Influence What People Know
The quality of representation depends on the messages that are communicated from
representatives to constituents. We have been systematically exploring the nature of the
representational messages that U.S. Senators disseminate to citizens on their websites and in
their press releases. We have traced these messages through the news media to determine the
amount, content and tone of the news messages that citizens see, read, and hear. We have
demonstrated that the representational messages emanating from senators vary by the gender of
the senator. Men and women senators emphasize different aspects of their jobs, different issue
themes, and different trait messages. The news media’s coverage also differs for men and
women senators, with reporters and editors allocating more coverage and more prominent
coverage to male senators. The resulting coverage resonates more closely with the messages of
male senators, but the tone of coverage is more positive for women senators.
In this chapter, we look at whether the messages presented by senators and transmitted by
the news media reach citizens. We explore whether citizens’ understanding of their senators
related to the actions and behaviors of their legislators. Similarly, we look at whether the news
media’s coverage of politicians and politics influence what constituents know about their
representatives in Washington. We know that levels of political knowledge can stimulate and
facilitate political participation (Palfrey and Poole, 1987; Junn, 1991). People with higher levels
of political knowledge feel more attached to the political system and are more willing to pay
attention to politics (Delli Carpini and Keeter, 1993). In addition, politically savvy citizens are
better able to process political information and may be more resistant to political manipulation
(e.g., Fiske, Lau, Smith, 1990; Lodge, McGraw, Stroh, 1989). Also, at election time, citizens
need rudimentary information about legislators in order to hold them accountable, such as
3
recognizing the senator’s name or the senator’s party affiliation. We are particularly interested
in determining if citizens’ knowledge of U.S. Senators varies by the gender of the senator.
Theory and Expectations
We know from decades of public opinion research that most people, most of the time,
have limited information about politics (e.g., Converse, 1964; Kahn and Kenney, 1999; Zaller,
1992). But, we do not know whether citizens’ basic levels of knowledge about senators are
systematically different for men and women senators. We have some theoretical expectations
regarding how the senator’s gender may influence what citizens know about their senators. To
begin, even though women gained the right to vote nearly one hundred years ago, politics
continues to be seen as a “man’s game” and people expect senators to be men (e.g., Verba,
Burns, Schlozman, 1997). While the number of women senators has increased almost ten-fold
since late the 1980s, male senators continue to outnumber women by a ratio of almost ten to one.
During the period of our study, the U.S. Senate included fourteen women senators among eighty-
six male senators.
Since norms about gender roles lead people to expect men to be senators, when people
encounter a senator (i.e., a woman senator) who violates this expectation, the senator stands out.
According to research in social psychology, salient stimuli are given “figural emphasis not
because of their own properties per se, but because of the contrast between them and the current
context or the perceiver’s temporary or long-term expectancies.” (Bargh, 1984: 18). In other
words, people pay attention to objects that are different. These objects may be noticed because
they are different from all the other objects in the setting (e.g., a few women among a large
group of men). Or, these objects may be noticed because they conflict with one’s expectations
(e.g., senators are expected to be men).
4
Information that is salient, because it is unexpected or novel, cannot be processed
automatically. Instead, unexpected information requires a greater degree of attention in order to
be understood. Since people are expending more effort when processing unexpected stimuli,
they are more likely to recall this information (Hastie and Kumar, 1979). Social psychologists
have consistently demonstrated that unique information is more memorable, producing the so-
called “Von Restorff Effect” (e.g., Hastie, 1981; Hunt, 1995).
Extrapolating to citizens’ knowledge about men and women senators, citizens cannot rely
on automatic processing to understand incoming information about women senators. Instead,
people need to rely on conscious processing and engage in more effortful behavior when trying
to interpret news about women senators. This behavior will produce greater recall of the
processed information, compared to information obtained about less salient male senators. Put
simply, we expect people will know more about women than men senators.
There is some empirical evidence suggesting that citizens may know more about women
than men politicians. In particular, an examination of senatorial candidates by Kahn and
Kenney (2004) found that potential voters were more likely to say they have been “exposed” to
the messages of incumbent women senators seeking reelection compared to incumbent men. In
addition, constituents were more willing to rate the performance of female incumbents.
Of course, the gender of the senator is only one force that may shape what citizens know
about their representatives. The characteristics of citizens play a key role in understanding
people’s knowledge about politics. For example, Delli Carpini and Keeter (1993), in their
classic book, What Americans Know about Politics and Why It Matters, identified a series of
demographic factors related to political knowledge, including education, age, and gender. In
5
particular, Delli Carpini and Keeter found that people with less education, younger people, and
women had lower levels of knowledge about politics.
The fact that women know less about politics than men is of particular interest to us (e.g.,
Burns, Schlozman and Verba, 2001; Kenski and Jamieson, 2000; Mondak and Anderson, 2004).
Research by Mondak and Anderson (2004) suggest that gender differences in political
knowledge are partially explained by a gender difference in the “propensity to guess” by survey
respondents answering questions. In particular, men are much less likely than women to choose
a “don’t know” response and are more likely than women to guess when offering an answer.
However, while gender differences in the willingness to answer survey questions
probably exaggerate the gender gap in political knowledge, it does not eliminate the gap. Burns,
Schlozman and Verba (2001) suggest that political socialization leads women to view politics as
a “man’s game,” leading to lower levels of political interest and political engagement. Since
women historically have been excluded from political life, politics and government is seen as an
arena dominated by men and, therefore, women are less motivated to seek out information about
politics and government (see also Delli Carpini and Keeter, 1996),
In certain circumstances, gender differences in political knowledge diminish or disappear.
For example, Delli Carpini and Keeter (1996) fail to find a gender gap in political knowledge
when looking at local political issues (e.g., people’s ability to name the head of the local school
board) and issues more directly relevant to women, such as health care and abortion policy (see
also Stolle and Gidengil, 2010; Kenski and Jamieson, 2000). In a more recent study, Dolan
(2011) compares traditional measures of political knowledge with a “gender-relevant” measure
of political knowledge. The gender-relevant measure of political knowledge includes questions
asking respondents to correctly identify the percent of women in the Congress or to correctly
6
identify the number of women on the Supreme Court. Dolan found that male respondents are
more likely to score higher than women on the traditional measure of political knowledge (i.e.,
“…do you happen to know which party holds a majority of the seats in the U.S. House of
Representatives?). In contrast, gender differences disappear or are even reversed for the gender
relevant measures of political knowledge. For example, significantly more women than men are
able to correctly identify the percentage of women serving in the U.S. Congress. Dolan’s
research also demonstrates the importance of political knowledge by showing that “gender-
relevant” political knowledge and “traditional” political knowledge influence levels of
participation, political interest, and political efficacy for both male and female respondents.
In a related vein, several researchers find that when the political landscape features
women politicians or women candidates, political interest, political knowledge, and political
engagement among women respondents increases (e.g., Atkeson, 2003; Campbell and
Wolbrecht, 2006; Hansen, 1997; Karp and Banducci, 2008; Koch, 1997). For example, Verba,
Burns, and Schlozman (1997) find that living in a state with a female senator or female senate
candidate increases a woman’s ability to identify the female politician. However, the presence of
a female politician in a state does not affect men’s ability to name their senator or identify the
senate candidates in their state. Similarly, Koch (1997) and Hansen (1997), looking at the 1992
election, also find that women’s political engagement increased significantly when a woman
senate candidate was on the ballot. In this chapter, we will contribute to this literature by
examining whether the gender gap in political knowledge is reduced when people are evaluating
women senators.
Beyond demographic characteristics, scholars have identified a series of additional
factors that are related to levels of political knowledge. For instance, media usage is related to
7
people’s understanding about politics, with people who pay attention to the news displaying
higher levels of knowledge about politics (e.g., Brians and Wattenberg, 1996; Zhao and Chaffee,
1995). Furthermore, people who are more interested in politics and are more strongly attached
to the political parties tend to be more informed about politics (e.g., Dolan, 2011; Dow, 2009;
Galston, 2001). Finally, researchers have shown that the political context can affect levels of
political information about the electorate. For instance, when campaigns are more competitive
and when the news media is covering the candidates more extensively, people’s understanding
about politics increases (Kahn and Kenney, 1999).
In the end, we expect people will have greater levels of information about female
senators compared to male senators because we expect the novelty of women senators will
encourage people to learn more about these senators. We also anticipate that gender differences
in political knowledge of senators will be reduced when people are exposed to women senators.
We turn now to an examination of what people know about their senators.
Data and Method
We need data on citizens’ understanding of U.S. Senators in order explore knowledge
levels about men and women senators. We rely on the 2006 Cooperative Congressional Election
Survey (CCES). As discussed in Chapter 2, the CCES was conducted by Polimetrix, Inc. during
the 2006 election. Respondents completed the survey on the Internet, answering a common
content section that all respondents received as well as a specific section of group content
constructed by individual research teams. The 2006 CCES contained a large sample of 36,500
respondents. In our analysis, we restrict our attention to the 32 senators examined in our study,
resulting in a survey sample of about 18,000 respondents in 17 states. We look at numerous
measures of citizens’ information about senators, including respondents’ ability to accurately
8
identify the senator’s party identification, to assess a senator’s job performance, to place senators
on an ideological scale, and to recall senators’ votes on a number of issues.
Gender and Political Knowledge
We develop a series of measures to assess how much respondents know about the
senators serving their states. We assess rudimentary information as well as more sophisticated
information about incumbent politicians. We begin by looking at citizens’ willingness to answer
simple questions about their senator. In particular, we examine whether people are willing to
answer three questions about their senator: (1) a question assessing the respondent’s approval of
the senator’s job performance; (2) a question asking respondents to volunteer the party
identification of the senator; and (3) a question asking respondents to identify the senator’s
ideological position on a scale ranging from extremely liberal to extremely conservative. For our
initial analysis, we create an index ranging from 0 to 3, measuring people’s willingness to
answer these three questions.1 Respondents answered, on average, two questions, with 10
percent of the sample answering none of the three questions and 37 percent of the respondents
answering all of the questions.
To explore whether the gender of the senator significantly influences people’s
willingness to answer questions about the senators, we rely on logistic ordinal regression since
the dependent variable can take on one of four values. In the regression equation, the primary
independent variable of interest is the senator’s gender. However, we need to control for rival
factors that may also influence people’s level of information about the senator before we can
draw judgments about the impact of a senator’s gender. At the state level, we include a control
variable for the election year, since people may be more aware of senators who are approaching
1 For each question, a respondent is given a score of 1 if the respondent answers the question and the respondent is
given a score of 0 if the respondent does not answer the question. After creating the binary variable (answered=1;
not answered=0) for each of the three questions, the three variables are summed, producing the three-point index.
9
election. Senators nearing reelection may be engaging in more media-oriented activities in
anticipation of their reelection campaign and people may be more motivated to seek out
information about these senators as Election Day nears.2
We also include a measure assessing the number of news paragraphs published about the
senator in the senator’s state newspaper. We expect that people will learn more about senators
and will be more willing to answer questions about their legislators when more information is
presented in the local press. The number of paragraphs in the state newspapers is a surrogate
measure for the level of information available in the respondent’s political environment. While
not every respondent in the survey is a daily reader of the newspaper examined in this study, the
amount of news attention in one newspaper (i.e., the largest circulating newspaper in the state) is
likely to be highly correlated with the amount of news attention in alternative outlets (Kahn and
Kenney, 1999).
In addition, it is important to control for individual level differences in people’s
predisposition to learn about politics in general, and their U.S. senators, in particular. The
literature on political knowledge points to a number of demographic factors, including the
respondent’s gender, age, and education level. We also include a series of political variables,
including how often a respondent pays attention to the national evening news since people who
watch the national news are likely to be more informed about politics. We include a measure of
political interest, where respondents are asked whether they are “very much interested,”
“somewhat interested,” or “not much interested” in “politics and current affairs.” We expect
people with high levels of political interest will be more informed about their senators.
2 We also examined whether the party of the senator and the seniority of the senator influenced citizens’ knowledge
of senators, but neither variable is influential; therefore, we do not include these variables in the present analysis.
10
We develop a measure of political sophistication by creating an index based on whether
respondents could correctly identify the party identification of their sitting governor as well as
correctly identify the political ideology of the Democratic Party and the Republican Party. For
the ideology questions, respondents received a “correct” score for the Democratic Party if they
placed the Democratic Party to the left of the middle and they received a “correct” score for the
Republican Party if they placed the Republican Party to the right of the middle. The political
sophistication scale ranged from 0 (no questions answered correctly) to 3 (each of the three
questions answered correctly).
Finally, we include a measure of strength of partisanship for respondents, with strong
Democrats and strong Republicans receiving a score of 3, weak Democrats or weak Republicans
receiving a score of 2, leaning Democrats and leaning Republicans scoring a 1 and pure
Independents receiving a score of 0. We expect strong identifiers to be more aware of their
political surroundings, including knowing pertinent facts about their sitting senators.
With the control variables in hand, we examine whether the gender of the senator
significantly influences people’s willingness to answer questions about their senator. In each of
the analyses in this chapter, it is necessary to examine “Senator 1” and “Senator 2” (the senior
and junior senator from each state) separately since each respondent is asked to make
assessments of both senators serving their state. The findings in Table 1 demonstrate the gender
of the senator significantly and powerfully influences people’s willingness to answer questions
about their senior and junior senator, holding key rival hypotheses constant. In both equations,
the coefficient for the gender of the senator is large and statistically significant, indicating that
people are much more likely to answer questions about women senators as opposed to male
senators.
11
Table 1 About Here
Compared to the gender of the senator, the senator’s proximity to reelection is less
important. Since both of these variables are binary, we can compare the size of the coefficients
directly. We see the coefficient for the gender of the senator is almost five times larger than the
election year coefficient in the equation for Senator 1 and more than ten times larger in the
equation for Senator 2. The final state level variable, paragraphs published about the senator, is
statistically significant and positive, indicating that people are more willing to answer questions
about their senators when these senators are covered more extensively in the press.
Turning to the respondent characteristics, our findings resonate with prior research
examining the determinants of political knowledge. For instance, people’s willingness to answer
questions about the senators is related to citizens’ level of education, age, and gender. In
particular, more educated respondents, older respondents, and male respondents are significantly
more willing to answer questions about their senators, compared to less educated, younger, and
women respondents. In addition, people who pay more attention to the news are more willing to
answer questions about their U.S. Senators. We also find that standard political variables
powerfully influence the respondents’ willingness to answer questions about sitting senators.
More specifically, we find that people’s level of political interest, their strength of partisanship,
and their level of political sophistication are important predictors in both of models presented in
Table 1.
We turn next to examining citizens’ willingness to answer more demanding questions
about their sitting senators. In particular, respondents are asked to identify each senator’s vote
on seven roll call votes on the following topics: (1) late-term abortion, (2) stem cell research, (3)
withdrawing of force from Iraq, (4) immigration reform, (5) raising the minimum wage, (6)
12
capital gains tax cut, and (7) the free trade agreement with Central America (CAFTA).3 We
sum up the number of roll call vote questions that respondents were willing to answer. On
average, respondents answered four of the seven vote questions, with about 18% of the people
unable to answer any of the questions and about 30% of the respondents willing to answer each
of the seven roll call questions. Relying on the same independent variables introduced earlier
and utilizing OLS regression, we look to see whether people are more willing to answer
challenging questions about women senators, compared to male senators.
The findings in Table 2 demonstrate respondents are significantly more likely to answer
questions about women senators’ roll call behavior compared to their male colleagues. In fact,
the gender of the senator is more important than election year or news coverage for predicting
the number of questions citizens are willing to answer about their senators’ voting records. In
both models, the gender of the senator produces almost a half-a-point advantage for women
senators on the eight-point index. These results, consistent with the findings presented in Table
1, suggest woman senators draw people’s attention and citizens are more likely to process
information about women senators compared to the male senators.
Table 2 About Here
We continue to find that the respondent’s demographic and political predispositions
influence powerfully their willingness to answer roll call questions. Political interest is the most
important political variable, followed by political sophistication. Among the demographic
characteristics, education and gender are the most consequential. We find as education
increases, people are more able to offer answers to questions about the senators’ voting record.
3 The series of questions began with the following preface: “As you know, Senators and Representatives in
Washington regularly have to decide how to vote on issues affecting the country. We’d like to ask you how you
would vote on some of these same issues as well as how you think your representative voted.” Please see the
Appendix for complete question wording for each of the ballot questions.
13
Women respondents are also significantly less likely than male respondents to offer answers to
questions about their senators. The tendency for women to respond less frequently to answering
questions (i.e., roll call questions as well as more generic questions about the senators) may
reflect women’s disinclination to guess, documented by Mondak and Anderson (2004).
Our results thus far show that people’s willingness to answer questions varies with
characteristics of their senator, their media environment, as well as by the demographic and
political characteristics of the respondents. We now turn to exploring whether these same factors
are important when we examine the accuracy of people’s perceptions of their senators. We start
with rudimentary information (i.e., the senator’s party identification) and move to progressively
more challenging information (i.e., ideology, roll call votes). We begin by exploring whether a
senator’s gender influences citizens’ ability to identify the party of the senator. We rely on
logistic regression to estimate people’s ability to correctly recall their senators’ party
identification since the dependent variable is dichotomous (i.e., correctly recall senator’s party
identification or not). We utilize the same independent variables introduced in our earlier
analyses. The findings in Table 3 indicate the gender of the senator powerfully influence
people’s ability to accurately identify the party of their sitting senator.
Table 3 About Here
In Figure 1, we illustrate graphically the impact of the gender of the senator by
converting the logit coefficients to probabilities, relying on a procedure described by King
(1989).4 The figure shows, all else equal, people have a higher probability of recalling a female
senator’s party correctly compared to a male senator’s party. More specifically, people have a
4 In calculating these probabilities, we vary the gender of the senator and hold all of the remaining variables in the
model at their means.
14
.85 probability of knowing the party identification of a female senator, but only a .73 probability
of correctly identifying the party of a male senator, representing a .12 change in probability.
Figure 1 About Here
The remaining variables in the model in Table 3 reveal similar patterns to those discussed
earlier. We again find that people learn from the news presented in their state; as news coverage
of the senator increase, people are significantly more likely to know the party identification of
their senator. We also continue to see that the respondents’ political and demographic
characteristics influence their knowledge about their senator. Again, people with more
education, more political interest, and more political sophistication about politics are more
educated about their senator. In addition, women and young people are less likely than men and
older people to correctly recall the party identification of their senator.
We turn next to a slightly more complicated task for citizens: can respondents correctly
identify the political ideology of their senator? We use a relatively easy standard to assess the
accuracy of the respondent’s answer. If a respondent places their Republican senator to the right
of moderate on the ideological scale, the respondent is coded as correctly identifying their
senator’s ideology. Similarly, if a respondent places their Democratic senator to the left of
moderate of the ideological scale, the respondent is coded as correctly identifying their senator’s
ideology. All other respondents are coded as unable to correctly identify their senator’s
ideology. 5
Placing senators on an ideological scale, even simply placing them to the left or the right
on a liberal-conservative continuum, is difficult for most respondents. While almost 80% of
respondents can correctly identify their senator’s party affiliation, only about half of the
5 Looking at ADA scores, no senator in our sample has a 50 (moderate) score on the ADA roll call index.
15
respondents can accurately place their senator on the correct side of the ideological scale.6 We
again rely on logistic regression to examine how the gender of the senator, holding rival forces
constant, influence people’s ability to place the senator’s ideological standings. The results of
this analysis are presented in Table 4. We continue to find that people give more accurate
answers about women senators, with respondents much more likely to place women senators
correctly on the ideological scale compared to male senators.
Table 4 About Here
However, some of the factors that significantly influenced knowledge of senators in
previous analyses are less important here. For instance, the amount of news coverage devoted to
the senators in the state’s largest circulating newspaper does not encourage people to accurately
identify their senators’ ideological leanings. Similarly, people who pay attention to the news are
not more likely to know the ideological leanings of their senators. This is not surprising given
the results of our content analysis of news coverage. We know from our examination of the
press that reporters rarely discuss ideological information when covering senators. In particular,
less than 10% of all the articles examined in our study mention the senators’ ideology, with an
average of less than .10 paragraphs per article.
We continue to find that certain types of political and demographic qualities enhance
people’s likelihood of possessing ideological information about their senators. More
specifically, as age, education, political interest, and political sophistication increase, people are
significantly more aware of their senators’ political proclivities. And, controlling for all other
factors, men continue to outpace women in their ability to offer accurate ideological information
about their senators.
6 53% of the respondents correctly place the senior senator and 56% of respondents correctly place the junior
senator on the ideological scale.
16
We conclude our examination of the impact of the gender of the senator on people’s
knowledge about their senators with a difficult test for the respondents. We look at people’s
ability to accurately recall their senator’s vote on several different roll call votes. To measure
respondents’ knowledge of the senators’ voting records, we calculated an index by comparing
the senator’s actual vote with the respondent’s recollection of how the senator voted on the seven
ballot measures presented earlier. On average, respondents correctly reported three of the seven
roll call votes for their senator, with one out of five respondents not able to accurately recall any
of their senator’s votes and only seven percent of the respondents able to correctly answer each
of the seven roll call questions.
We examine how the gender of senator influences people’s familiarity with the senator’s
voting record with OLS regression, controlling for state level factors (e.g., the senator’s
proximity to reelection, the amount of news coverage about the senator) and individual level
factors (e.g., political and demographic factors). The findings in Table 5 once again demonstrate
the importance of the senator’s gender: the gender of the senator powerfully influences people’s
ability to recall their senator’s votes on important legislative matters. In the model predicting
Senator 1, the unstandardized OLS coefficient suggests that being a female senator, controlling
for all other factors, increases a respondent’s score on the eight-point scale by about half-a-
point. In the model predicting Senator 2, the unstandardized OLS coefficient indicates that
people’s scores for female senators are, on average, are about three-quarters of a point higher
than people’s scores for male senators, all else being equal. Furthermore, the impact of the
gender of the senator is substantial relative to the other variables in the model. The standardized
coefficients indicate that the senator’s gender is more important than the respondent’s age and
17
strength of partisanship when predicting people’s ability to accurately recall the senators’ voting
record.
Table 5 About Here
The remaining variables in the models perform as expected. For instance, we find that
newspapers that publish more paragraphs about the senators lead respondents to more accurately
recall their senators’ roll call votes. Similarly, people who pay more attention to the news are
more likely to score higher on the roll call index. And, the political and demographic
characteristics of the respondent continue to be important, with political interest and political
sophistication powerfully influenceing people’s recall of their senators’ actions in office. As an
illustration, the unstandardized coefficient for political interest in the model for Senator 1
suggests that a one-point change on the interest variable (e.g., moving from somewhat interested
to very interested in politics) produces almost a one-point increase in people’s score on the roll
call index, holding all rival hypotheses constant.
The results of our analysis thus far are strikingly consistent. People are more willing to
answer questions about women senators and people have much more accurate information about
women senators compared to their male counterparts. In addition, when people are given more
information about their senators in local news, they are more likely to answer questions more
accurately (with the exception of ideology). People who are more interested and informed about
politics have higher levels of information about their senators and are more willing to offer this
information when asked. Finally, older and educated respondents score higher on the
information measures examined in this chapter than younger and less educated individuals.
One of the most powerful and unwavering findings in this chapter is the fact that women
respondents score significantly lower on the various measures of senator knowledge. Given
18
findings from previous studies suggesting that women’s level of political knowledge is often
enhanced when they are exposed to women candidates and women officeholders (Atkeson, 2003;
Campbell and Wolbrecht, 2006; Hansen, 1997; Karp and Banducci, 2008; Koch, 1997), we
wanted to determine if women’s understanding of their senators increases when they are
represented by a woman senator. That is, we hypothesize that a woman’s level of information
about a senator depends on or is conditioned by the gender of the senator.
To explore the conditional relationship between women respondents’ awareness of their
senators and the gender of the senator, we examine the interaction between gender of the senator
and gender of the respondent in a series of OLS regression equations. In particular, we
developed two dependent variables. The first dependent variable is an index created by
summing people’s willingness to answer basic questions about their senator. We simply combine
responses to the first two dependent variables examined in this chapter (e.g., Tables 1-2),
creating an index that ranges from 0 to 10. The second dependent variable is an index created by
summing people’s accuracy in answering the following questions: (1) party identification of the
senator, (2) ideology of the senator, (3) roll call vote of the senator. We combine responses to
these last three dependent variables (e.g., Tables 3-5) to create an “accuracy” index ranging from
0 to 9.
We present the results of the multiplicative analyses in Table 6. We are particularly
interested in the interaction coefficients at the top of the table. As the positive coefficients
indicate, women respondents are more willing to answer questions about their senator and are
more accurate in these responses when they are evaluating a female senator compared to a male
senator. In three of the four cases, the interaction coefficient is large and statistically significant.
Table 6 About Here
19
To aid in the interpretation of the interaction effects, we rely on the OLS coefficients in
Table 6 to estimate male and female respondents’ willingness to answer questions as well as the
accuracy of their answers about men and women senators. In particular, we derive point
estimates by varying the gender of the respondent and the gender of the senator, while holding
all remaining variables at their means (Lewis-Beck, 1980).7 We begin by looking at people’s
willingness to answer questions about their senators. The data in Figure 2 reveal there is almost
a one-point increase in women’s willingness to answer questions about their senator when the
senator is a woman (5.99 vs 5.01). Furthermore, the difference between men and women
respondents’ readiness to answer questions is much more substantial for a male senator (i.e.,
more than one-point on the index, 6.19 vs 5.01) compared to a woman senator (i.e., less than .50
of a point on the index, 6.36 vs 5.99).
Figure 2 About Here
Turning to the accuracy of people’s answers about the senators, we find a similar pattern.
Women respondents give more accurate answers when assessing a female senator, compared to a
male senator (see Figure 3). We find, on average, almost a one-point difference in the accuracy
of answers when women are answering questions about a women senator than when women are
answering questions about a male senator (3.74 vs 2.95). For male respondents, the gender of
the senator is less consequential, creating less than a half of point change in male respondents’
accuracy scores (4.19 vs 3.86). Similarly, we also find that the gender difference in respondents’
answers is more dramatic when these respondents are targeting a male senator (i.e., almost a one-
point difference in scores on the accuracy index, 3.86 vs 2.95) than when respondents are
answering questions about a female senator (less than a .50 point difference on the accuracy
index, 4.19 vs 3.74).
7 In both Figure 2 and Figure 3, we calculate point estimates for Senator 1 only.
20
Figure 3 About Here
These results indicate that the gender of the senator is more consequential for women
respondents than for male respondents. While both men and women respondents are more likely
to answer questions about women senators and are more accurate in their answers to questions
about women senators, women respondents are more affected by the gender of the senator. In
other words, the difference in men and women’s responses diminish when respondents are
evaluating a woman senator compared to a male senator. The presence of woman senator may
stimulate women citizens to pay more attention to their political environment and learn more
about their representatives.
Conclusion
In this chapter we examined the question: does the gender of the senator influence
citizens’ levels of knowledge about their senators? The answer is a resounding yes. First
things first, citizens are more willing to answer questions about women senators compared to
men senators. In addition, constituents are more likely to know a woman senator’s name, party
affiliation, ideological leanings and specific votes on key issues, compared to an equivalent male
senator. These tasks, ranging from rudimentary questions to quite sophisticated and demanding
queries, yield unwavering results. People know more about women senators than male senators,
even controlling for a range of rival factors.
These clear, consistent, and powerful findings support our contention that people pay
more attention to the actions and information associated with women senators because they are
more novel and unusual. Women senators remain unique, especially across the arc of U.S.
history. Consequently, citizens need to expend more energy to processing information about
these senators. This additional cognitive work allows citizens to hold onto and more readily
recall information about women senators. This finding is especially intriguing given the
21
findings in the previous chapter demonstrating that male senators receive more media coverage
than their female counterparts.
We also find that women senators reduce the “gender gap” in men and women’s
understanding about politics. Across each of the dependent variables examined in this chapter,
we find that female respondents are less knowledgeable than male respondents. However, the
gender gap in political knowledge is reduced significantly when female respondents are
assessing a woman senator, instead of a male senator. Consistent with earlier research on this
topic, we find that women’s interest in politics is piqued by the presence of women leaders.
Women in leadership positions seem to signify to female constituents that politics is relevant to
them, encouraging women to pay more attention to political life.
22
References
Atkeson, Lonna Rae. 2003. Not All Cues are Created Equal: The Conditional Impact of Female
Candidates on Political Engagement. Journal of Politics, Volume 65 (4): 1040–61
Bargh, John A. 1984. Automatic and conscious processing of social information. In
Robert S. Wyer and Thomas K. Srull (eds)., Handbook of Social Cognition, Volume 3, 1-
44. Hillsdale, NJ: Erlbaum.
Brians, Craig Leonard and Martin P. Wattenberg. 1996. Campaign Issue Knowledge and
Salience: Comparing Reception from TV Commercials, TV News and Newspapers.
American Journal of Political Science. Volume 40(1), 172-193.
Burns, Nancy, Kay, Schlozman, and Verba, Sidney. 2001. The Private Roots of Public Action.
Cambridge, MA: Harvard University Press.
Campbell, David E., and Christina Wolbrecht. 2006. See Jane run: Women politicians as role
models for adolescents. Journal of Politics, Volume 68(2), 233–245.
Converse, Philip E. 1964. Nature of Belief Systems in Mass Publics. In Davie Apter (ed.)
Ideology and Discontent. New York: Free Press.
Delli Carpini, Michael X., and Scott Keeter. 1993. Measuring Political Knowledge: Putting First
Things First. American Journal of Political Science, Volume 37: 179-206.
Dolan, Kathleen. 2011. Do Women and Men Know Different Things? Measuring Gender
Differences in Political Knowledge. Journal of Politics, Volume 73: 97-107.
Dow, Jay. 2009. Gender Differences in Political Knowledge: Distinguishing Characteristics-
Based and Returns-Based Differences. Political Behavior, Volume 31 (1): 117–36.
Fiske, Susan T., Richard R. Lau, and Richard A. Smith. 1990. On the Varieties and Utilities
of Political Expertise. Social Cognition Volume 8:31-48.
23
Galston, William. 2001. Political Knowledge, Political Engagement, and Civic Education.
Annual Review of Political Science, 4, 217-234.
Hansen, Susan. 1997. Talking About Politics: Gender and Contextual Effects on Political
Proselytizing. Journal of Politics, Volume 59 (1): 73–103.
Hastie, Reid. 1981. Schematic principles in human memory. In E. Tony Higgins, Charles
Herman & Mark P. Zanna (Eds.), Social Cognition: The Ontario Symposium, Volume
1, 39-88. Hillsdale, NJ: Erlbaum.
Hastie, Reid and Amand Kumar. 1979. Person memory: Personality traits as
organizing principles in memory for behaviors. Journal of Personality
and Social Psychology, Volume 37, 25-38.
Hunt, R. Reed. 1995. The subtlety of distinctiveness: What von Restorff really did
Psychonomic Bulletin & Review, Volume 2(1), 105-112.
Junn, Jane. 1991. Participation and Political Knowledge. In William Crotty (ed.) Political
Participation and American Democracy, New York: Greenwood Press, 193-212.
Kahn, Kim E, and Patrick J. Kenney. 1999. The Spectacle of U.S. Senate Campaigns. Princeton,
NJ: Princeton University Press.
Kahn, Kim Fridkin, and Patrick J. Kenney. 2004. No Holds Barred: Negativity in U.S. Senate
Campaigns. Upper Saddle River, NJ: Pearson Prentice Hall.
Karp, Jeffrey A. and Susan A. Banducci. 2008. When Politics in Not Just a Man’s Game:
Women’s Representation and Political Engagement. Electoral Studies, Volume 27: 105-
115.
Kenski, Kate, and Jamieson, Kathleen Hall. 2000. The Gender Gap in Political Knowledge: Are
24
Women Less Knowledgeable Than Men about Politics? In Kathleen Hall Jamieson (ed.)
Everything You Think You Know About Politics And Why You’re Wrong. New York:
Basic Books, 83–92.
Koch, Jeffrey. 1997. Candidate Gender and Women’s Psychological Engagement in Politics.
American Politics Quarterly 25 (1): 118–33.
Lewis-Beck, Michael. 1980. Applied Regression. Quantitative Applications in the Social
Science, Newbury Park, California: Sage Publications.
Lodge, Milton G., Kathleen McGraw, and Patrick Stroh. 1989. An Impression-Driven
Model of Candidate Evaluation. American Political Science Review
Volume 83: 399-420.
Mondak, Jeffery, and Anderson, Mary. 2004. The Knowledge Gap: A Reexamination of
Gender–Based Differences in Political Knowledge. Journal of Politics Volume 66 (2):
492–512.
Palfrey, Thomas R. and Keith T. Poole. 1987. The Relationship between Information, Ideology,
and Voting Behavior. American Journal of Political Science, Volume 31:511-530.
Stolle, Dietlind, and Gidengil, Elisabeth. 2010. What Do Women Really Know? A Gendered
Analysis of Varieties of Political Knowledge. Perspectives on Politics Volume 8 (1): 93–
109.
Verba, Sidney, Burns, Nancy, and Schlozman, Kay. 1997. Knowing and Caring about Politics:
Gender and Political Engagement. Journal of Politics, Volume 59 (4): 1051–72
Zaller, John R. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge
University Press.
Zhao, Xinshu, and Steven H. Chaffee. 1995. Campaign Advertisements Versus Television News
25
as Sources of Political Issue Information. Public Opinion Quarterly Volume 59:41-65.
26
73
85
50
55
60
65
70
75
80
85
90
95
100
Percentage
Figure 1. Correctly Recalling Senator's Party
by Gender of the Senator and Gender of the Respondent
Male Senator
Female senator
Note: This figure shows the impact of the gender of the senator on the probability of correctly recalling the senator's party
identification. These probabilities are based on the unstandardized estiamtes presented in Table 3. We calculate the probabilities by
varying the gender of the senator whileholding all remaining variables at their means (See King, 1989: 105). For ease of presentation,
we average the probabilities for Senator 1 and Senator 2.
27
0
1
2
3
4
5
6
7
Male Senator 1 Female Senator
6.91
6.36
5.01
5.99
Willingness to Answer
Figure 2. The Interaction Effect of Senator's Gender and Respondent's
Gender on Willingness to Answer Questions about Senators
Male Respondent
Female Respondent
Note: These estimates are based on the OLS coefficients presented in Table 6. The point estimates are derived
by varying the gender of the respondent and the gender of the senator while holding all remaining variables
at their means (Lewis-Beck, 1980).
28
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Male Senator 1 Female Senator
3.86
4.19
2.95
3.74
Accuracy of Answe r
Figure 3. The Interaction Effect of Senator's Gender and
Respondent's Gender on Accuracy of Answers about Senator
Male Respondent
Female Responde nt
Note: These estimates are based on the OLS coefficients presented in Table 6. The point estimates are derived by
varying the gender of the respondent and the gender of the senator while holding all remaining variables
at their means (Lewis-Beck, 1980).
29
Table 1. Logistic Ordinal Regression Predicting
Respondents’ Willingness to Answer Questions about Their Senators1
Senator 1 Senator 2
State Level Variables
Gender of Senator .67 (.07)*** .89 (.07)***
Election Year -.14 (.7)** -.06 (.10)
News about Senator .002 (.000)*** .14 (.03)***
Respondent Characteristics
Strength of Party Identification .17 (.03)*** .17 (.03)***
Political Interest .91 (.05)*** .80 (.05)***
Political Sophistication .60 (.03)*** .59 (.03)***
Education .15 (.02)*** .14 (.02)***
Age .04 (.003)*** .03 (.002)***
Gender of Respondent -.61 (.07)*** -.45 (.07)***
Attention to News .16 (.03)*** .14 (.03)***
Threshold 0 3.91 (.18)*** 3.64 (.17)***
Threshold 1 4.96 (.19)*** 4.59 (.18)***
Threshold 2 6.13 (.19)*** 5.71 (.18)***
Model Chi2 2524.47*** 2427.81***
-2 Log Likelihood 10524.93 10662.20
DF 10 10
Pseudo R-Square (Cox and Snell) .38 .36
N 5,329 5,434
1 Unstandardized logit coefficients are followed by standard errors in parentheses.
Note: The dependent variable is citizens’ willingness to answer three questions about their senator: a
question assessing the respondent’s approval of the senator’s job performance, a question asking
respondents to volunteer the party identification of the senator, and a question asking respondents to
identify the senator’s ideological position. Gender of the Senator is coded 1 for female senators, 0 for
male senators. Election Year is coded 1 for senators up for reelection in 2006, 0 for other senators. News
about the senator is the number of paragraphs about the senator published in the largest circulating
newspaper in the state. Strength of party ranges from Independent (0) to strong Republican or Democrat
(3). Political interest ranged from “not at all” interested (1) to “very interested (3). Political
sophistication ranges from 0 to 3. Education is coded on a six-point scale ranging from no high school to
post graduate education and Age is coded in years. Gender of respondent is coded 1 for female and 0 for
male. Attention to news is based on a four-point scale indicating that the respondent watched the
national evening news: not at all (1); once or twice a week (2); a few times a week (3); almost every day
(4).
*** p<.01
** p<.05
* p<.10
30
Table 2. OLS Regression Predicting
Respondents’ Willingness to Answer Roll Call Questions1
Senator 1 Senator 2
State Level Variables
Gender of Senator .44 (.07)*** .05 .41 (.08)*** .08
Election Year -.14 (.7)* -.02 -.02 (.11) -.004
News about Senator .001 (.000)** .03 .001 (.000)*** .06
Respondent Characteristics
Strength of Party Identification .12 (.03)*** .05 .13 (.03) .05
Political Interest 1.1 (.06)*** .28 1.03 (.05)*** .26
Political Sophistication .38 (.03)*** .15 .37 (.03)*** .15
Education .20 (.02)*** .10 .21 (.02)*** .11
Age .01 (.003)*** .05 .009 (.003)*** .04
Gender of Respondent -.59 (.07)*** .10 -.58 (.07)*** -.10
Attention to News .13 (.03)*** .06 .13 (.03)*** .05
Constant -1.57 (.20)*** -1.35 (.18)***
R-Squared .25 .23
N 5,339 5,450
1 Unstandardized OLS coefficients, with standard errors in parentheses, followed by standardized
coefficients.
Note: The dependent variable is citizens’ willingness to answer questions about the senator’s roll call vote
on seven different votes. Gender of the Senator is coded 1 for female senators, 0 for male senators.
Election Year is coded 1 for senators up for reelection in 2006, 0 for other senators. News about the
senator is the number of paragraphs about the senator published in the largest circulating newspaper in the
state. Strength of party ranges from Independent (0) to strong Republican or Democrat (3). Political
interest ranged from “not at all” interested (1) to “very interested (3). Political sophistication ranges from
0 to 3. Education is coded on a six-point scale ranging from no high school to post graduate education
and Age is coded in years. Gender of respondent is coded 1 for female and 0 for male. Attention to news
is based on a four-point scale indicating that the respondent watched the national evening news: not at all
(1); once or twice a week (2); a few times a week (3); almost every day (4).
*** p<.01
** p<.05
* p<.10
31
Table 3. Logistic Regression Predicting Ability to Identify Senator’s Party Affiliation1
Senator 1 Senator 2
State Level Variables
Gender of Senator .85 (.08)*** .97 (.09)***
Election Year -.07 (.08) .09 (.12)
News about Senator .002 (.000)*** .003 (.000)***
Respondent Characteristics
Strength of Party Identification .27 (.03)*** .30 (.03)***
Political Interest .91 (.06)*** .87 (.06)***
Political Sophistication .69 (.03)*** .74 (.04)***
Education .25 (.03)*** .34 (.03)***
Age .04 (.003)*** .03 (.003)***
Gender of Respondent -.45 (.08)*** -.31 (.08)***
Attention to News .04 (.04) .05 (.04)
Constant -6.45 (.25)*** -6.904 (.25)***
% of Cases Correctly Predicted 77% 78%
N 4288 4378
1 Unstandardized logit coefficients are followed by standard errors in parentheses.
Note: The dependent variable is the citizens’ ability to identify correctly the party of the senator. Gender
of the Senator is coded 1 for female senators, 0 for male senators. Election Year is coded 1 for senators
up for reelection in 2006, 0 for other senators. News about the senator is the number of paragraphs about
the senator published in the largest circulating newspaper in the state. Strength of party ranges from
Independent (0) to strong Republican or Democrat (3). Political interest ranged from “not at all”
interested (1) to “very interested (3). Political sophistication ranges from 0 to 3. Education is coded on a
six-point scale ranging from no high school to post graduate education and Age is coded in years. Gender
of respondent is coded 1 for female and 0 for male. Attention to news is based on a four-point scale
indicating that the respondent watched the national evening news: not at all (1); once or twice a week (2);
a few times a week (3); almost every day (4).
*** p<.01
** p<.05
* p<.10
32
Table 4. Logistic Regression Predicting Ability to Correctly Identify the Senator’s Ideology1
Senator 1 Senator 2
State Level Variables
Gender of Senator .75(.08)*** .41 (.08)***
Election Year -.16 (.08)** -.15 (.11)
News about Senator .00 (.000) .000 (.000)
Respondent Characteristics
Strength of Party Identification .12 (.03)*** .03 (.03)
Political Interest .56 (.06)*** .46 (.05)***
Political Sophistication .86 (.04)*** .92 (.04)***
Education .06 (.02)*** .08 (.04)***
Age .01 (.003)*** .01 (.003)***
Gender of Respondent -.31 (.07)*** -.26 (.07)***
Attention to News -.01 (.03) -.003 (.03)
Constant -4.72 (.23)*** -4.43 (.21)***
% of Cases Correctly Predicted 74% 72%
N 4288 4378
1 Unstandardized logit coefficients are followed by standard errors in parentheses.
Note: The dependent variable is the citizens’ ability to identify correctly the ideology of the senator. If
the respondent identified the ideology of a Democratic senator as left of moderate on the ideological
scale, the respondent was coded as correctly identifying the senator’s ideology. If the respondent
identified the ideology of a Republican senator as right of moderate on the ideological scale, the
respondent was coded as correctly identifying the senator’s ideology. All other respondents were coded
as incorrectly indentifying the ideology of the senator. Gender of the Senator is coded 1 for female
senators, 0 for male senators. Election Year is coded 1 for senators up for reelection in 2006, 0 for other
senators. News about the senator is the number of paragraphs about the senator published in the largest
circulating newspaper in the state. Strength of party ranges from Independent (0) to strong Republican or
Democrat (3). Political interest ranged from “not at all” interested (1) to “very interested (3). Political
sophistication ranges from 0 to 3. Education is coded on a six-point scale ranging from no high school to
post graduate education and Age is coded in years. Gender of respondent is coded 1 for female and 0 for
male. Attention to news is based on a four-point scale indicating that the respondent watched the
national evening news: not at all (1); once or twice a week (2); a few times a week (3); almost every day
(4).
*** p<.01
** p<.05
* p<.10
33
Table 5. OLS Regression Predicting
Accuracy of Respondents’ Knowledge of Senators’ Roll Call Votes1
Senator 1 Senator 2
State Level Variables
Gender of Senator .42 (.06)*** .09 .74 (.07)*** .17
Election Year -.31 (.6)*** -.02 -.02 (.09) -.005
News about Senator .001 (.000)*** .05 .001 (.000)*** .05
Respondent Characteristics
Strength of Party Identification .10 (.02)*** .05 .10 (.02)*** .05
Political Interest .94 (.04)*** .30 .83 (.04)*** .26
Political Sophistication .37 (.03)*** .19 .36 (.03)*** .18
Education .20 (.02)*** .13 .21 (.02)*** .13
Age .01 (.002)*** .08 .01(.002)*** .05
Gender of Respondent -.58 (.06)*** .12 -.49(.06)*** -.10
Attention to News .04(.02)** .02 .06 (.02)** .03
Constant -1.85 (.15)*** -1.74 (.14)***
R-Squared .30 .29
N 5,339 5,450
1 Unstandardized OLS coefficients, with standard errors in parentheses, followed by standardized
coefficients.
Note: The dependent variable is the number of roll call votes that citizens’ correctly answer about the
senator across seven different roll call votes. Gender of the Senator is coded 1 for female senators, 0 for
male senators. Election Year is coded 1 for senators up for reelection in 2006, 0 for other senators. News
about the senator is the number of paragraphs about the senator published in the largest circulating
newspaper in the state. Strength of party ranges from Independent (0) to strong Republican or Democrat
(3). Political interest ranged from “not at all” interested (1) to “very interested (3). Political
sophistication ranges from 0 to 3. Education is coded on a six-point scale ranging from no high school to
post graduate education and Age is coded in years. Gender of respondent is coded 1 for female and 0 for
male. Attention to news is based on a four-point scale indicating that the respondent watched the
national evening news: not at all (1); once or twice a week (2); a few times a week (3); almost every day
(4).
*** p<.01
** p<.05
* p<.10
34
Table 6. OLS Regression Estimating the Interaction Effect of Senator Gender and Respondent
Gender on Willingness and Accuracy in Answering Questions about the
Willingness to Answer Accuracy in Answering
Senator 1 Senator 2 Senator 1 Senator 2
Interaction
Senator Gender * Respondent Gender .81 (.17)*** .53 (.17)*** .46 (.13)*** .06 (.12)
State Level Variables
Gender of Senator .17 (.15) .52 (.16)*** .33 (.11)*** .56 (.11)***
Election Year -.20 (.09)** -.15 (.14) -.33 (.07)*** -.18 (.10)*
News about Senator .002 (.00)*** .002 (.00)*** .001 (.00)*** .001 (.00)***
Respondent Characteristics
Strength of Party Identification .20 (.04)*** .22 (.04)*** .16 (.03)*** .15 (.02)***
Political Interest 1.52 (.07)*** 1.41 (.07)*** 1.19 (.05)*** 1.16 (.05)***
Political Sophistication .68 (.04)*** .65 (.04)*** .63 (.03)*** .63 (.03)***
Education .25 (.03)*** .27 (.03)*** .24 (.02)*** .24 (.02)***
Age .02 (.003)*** .02 (.003)*** .02 (.003)*** .02 (.002)***
Gender of Respondent -1.18 (.12)*** -1.01 (.12)*** -.91 (.08)*** -.72 (.09)***
Attention to News .22 (.04)*** .19 (.04)*** .05 (.03)* .05 (.03)*
Constant -2.11 (.24)*** -1.90 (.23)*** -2.42 (.18)*** -2.34 (.17)***
R-Squared .34 .32 .37 .38
N 5328 5322 5329 5339
1 Unstandardized OLS coefficients, with standard errors in parentheses, followed by standardized coefficients.
Note: “Willingness to Answer” is an index composed of the dependent variables utilized in Tables 1-3 and
“Accuracy in Answering” is an index composed of the dependent variables utilized in Tables 4-5 (see text for more
information). Gender of the Senator is coded 1 for female senators, 0 for male senators. Election Year is coded 1 for
senators up for reelection in 2006, 0 for other senators. News about the senator is the number of paragraphs about
the senator published in the largest circulating newspaper in the state. Strength of party ranges from Independent (0)
to strong Republican or Democrat (3). Political interest ranged from “not at all” interested (1) to “very interested
(3). Political sophistication ranges from 0 to 3. Education is coded on a six-point scale ranging from no high school
to post graduate education and Age is coded in years. Gender of respondent is coded 1 for female and 0 for male.
Attention to news is based on a four-point scale indicating that the respondent watched the national evening news:
not at all (1); once or twice a week (2); a few times a week (3); almost every day (4).
*** p<.01
** p<.05
* p<.10
35
Appendix.
Survey Questions from Cooperative Congressional Election Study (CCES/Common Content)
Approval of Senator
Do you approve or disapprove of the way <Senator’s Name> is handling his/her job as U. S. Senator for
<state>?
<1> Strongly approve
<2> Somewhat approve
<3> Somewhat disapprove
<4> Strongly disapprove
<5> Not sure
Party Identification of Senator
Do you happen to remember the party affiliation of <Senator’s Name>?
<1> Democrat
<2> Republican
<3> Independent
<4> Don't know
Ideology of Senators
One way that people talk about politics in the United States is in terms of left, right, and center, or liberal,
conservative, and moderate. We would like to know how you view the parties and candidates using these
terms. The scale below represents the ideological spectrum from very liberal (0) to very conservative
(100). The most centrist American is exactly at the middle (50).
Where would you place <Senator’s Name>?
If you are not sure, or don't know, please check here ___.
Roll Call Questions
Late-Term Abortion Ban
First, we'd like to ask about a proposal in Congress to ban a type of late-term abortion sometimes called
"partial-birth abortion." Some argue that late-term abortion is a barbaric procedure and should be banned.
Others argue that late-term abortions are extremely uncommon and used only in exceptional
circumstances best determined by a doctor, not the Congress. The proposed legislation could also be the
opening to a broader ban on abortion.
How about <Senator’s Name>? Do you think <he/she> voted for or against banning late-term abortion?
<1> For (that is, to ban late-term abortion)
<2> Against (that is, not to ban late-term abortion)
<3> Don't Know
Funding for Stem Cell Research
Now we’d like to ask you about whether the federal government should fund stem cell research.
Some in Congress argue that this research may lead to cures for diseases and disabilities affecting large
numbers of Americans, and should be funded. Others argue that a potential human life has to be destroyed
in order to use these cells, and funding it would be unethical.
36
How about <Senator’s Name>? Do you think <he/she> voted for or against funding the research?
<1> For (that is, funding the research)
<2> Against
<3> Don't know
Phased Redeployment of U.S. Troops in Iraq
Congress also debated a proposal that the president begin phased redeployment of U.S. troops from Iraq
starting this year and submit to Congress by the end of 2006 a plan with estimated dates for continued
phased withdrawal.
Some politicians argue that setting out a plan to withdraw would make Iraqis take responsibility for their
country and become more independent of the U.S. Others argue that it is too early to start withdrawing,
and that doing so would make terrorists grow bolder.
How about <Senator’s Name>? Do you think <he/she> voted for or against this plan?
<1> For (setting a timetable to withdraw from Iraq)
<2> Against
<3> Don't know
Illegal Immigration
Another issue is illegal immigration. One plan considered by the Senate would offer illegal immigrants
who already live in the U.S. more opportunities to become legal citizens.
Some politicians argue that people who have worked hard in jobs that the economy depends should be
offered the chance to live here legally. Other politicians argue that the plan is an amnesty that rewards
people who have broken the law.
How about <Senator’s Name>? Do you think <he/she> voted for or against this proposal?
<1> For (offering illegal immigrants an opportunity to become citizens)
<2> Against
<3> Don't know
Raising the Minimum Wage
Congress considered a proposal to increase the federal minimum wage from $5.15 to $6.25 within the
next year and a half.
Some politicians argue that the wage should be increased because it hasn't changed since 1997 and many
workers still live in poverty. Other politicians argue that raising the wage might force small businesses to
cut jobs and would hurt the economy.
How about <Senator’s Name> Do you think <he/she> voted for or against increasing the minimum wage?
<1> For (in favor of increasing the federal minimum wage)
<2> Against
<3> Don't know
Extension of Capital Gains Tax Cuts
We'd like to ask about cutting taxes on the money people make from selling investments, also referred to
as capital gains. This past year the Senate considered a bill to extend capital gains tax cuts passed in
2001.
37
Some politicians argue that these tax reductions make the economy strong and encourage people to invest
more. Others argue that the plan would mostly benefit people who are already rich and that any tax cuts
should be shared more fairly among all taxpayers.
How about <Senator’s Name>? Do you think <he/she> voted for or against increasing these tax cuts?
<1> For (that is to extend the capital gains tax cuts)
<2> Against
<3> Don't know
CAFTA
This year Congress also debated a new free trade agreement that reduces barriers to trade between the
U.S. and countries in Central America.
Some politicians argue that the agreement allows America to better compete in the global economy and
would create more stable democracies in Central America. Other politicians argue that it helps businesses
to move jobs abroad where labor is cheaper and does not protect American producers.
How about <Senator’s Name>? Do you think <he/she> voted for or against the trade agreement?
<1> For (that is to ratify the trade agreement)
<2> Against
<3> Don't know
Party Identification of Respondent
Generally speaking, do you think of yourself as a Democrat, Republican, Independent or what?
[If Democrat or Republican],would you call yourself a strong [Democrat/Republican] or a not very strong
[Democrat/Republican]? If Independent, Do you think of yourself as closer to the Democratic or the
Republican Party?
<1> Strong Democrat
<2> Weak Democrat
<3> Democratic Leaner
<4> Independent
<5> Republican Leaner
<6> Weak Republican
<7> Strong Republican
<8> Other
<9> Don’t Know
Interest in Politics
How interested are you in politics and current affairs?
<1> Very much interested
<2> Somewhat interested
<3> Not much interested
Pay Attention to the News
During the past week, how many times did you watch the national evening news?
<1> Not at all (0 times)
<2> Once or Twice (1-2 times)
<3> A few times (3-4 times)
<4> Almost every day (5-7 times)
