News vs.

Entertainment: How Increasing Media

Choice Widens Gaps in Political Knowledge
and Turnout
Markus Prior Princeton University

Despite dramatic increases in available political information through cable television and the Internet, political knowledge
and turnout have not changed noticeably. To explain this seeming paradox, I argue that greater media choice makes it
easier for people to find their preferred content. People who like news take advantage of abundant political information to
become more knowledgeable and more likely to turn out. In contrast, people who prefer entertainment abandon the news
and become less likely to learn about politics and go to the polls. To test this proposition, I develop a measure of people’s
media content preference and include it in a representative opinion survey of 2,358 U.S. residents. Results show that content
preference indeed becomes a better predictor of political knowledge and turnout as media choice increases. Cable TV and
the Internet increase gaps in knowledge and turnout between people who prefer news and people who prefer entertainment.

T
he rise of new media has brought the question of up less political information than they used to. In a high-
audience fragmentation and selective exposure to choice environment, lack of motivation, not lack of skills
the forefront of scholarly and popular debate. In or resources, poses the main obstacle to a widely informed
one of the most widely discussed contributions to this de- electorate.
bate, Sunstein (2001) has proposed that people’s increas- As media choice increases, content preferences thus
ing ability to customize their political information will become the key to understanding political learning and
have a polarizing impact on democracy as media users be- participation. In a high-choice environment, politics con-
come less likely to encounter information that challenges stantly competes with entertainment. Until recently, the
their partisan viewpoints. While this debate is far from impact of content preferences was limited because media
settled, the issue which precedes it is equally important users did not enjoy much choice between different con-
and often sidestepped: as choice between different me- tent. Television quickly became the most popular mass
dia content increases, who continues to access any type of medium in history, but for decades the networks’ schedul-
political information? Cable television and the Internet ing ruled out situations in which viewers had to choose
have increased media choice so much in recent decades between entertainment and news. Largely unexposed to
that many Americans now live in a high-choice media en- entertainment competition, news had its place in the early
vironment. As media choice increases, the likelihood of evening and again before the late-night shows. Today, as
“chance encounters” (Sunstein) with any political content both entertainment and news are available around the
declines significantly for many people, as this study will clock on numerous cable channels and web sites, people’s
demonstrate. Greater choice allows politically interested content preferences determine more of what those with
people to access more information and increase their po- cable or Internet access watch, read, and hear.
litical knowledge. Yet those who prefer nonpolitical con- Distinguishing between people who like news and
tent can more easily escape the news and therefore pick take advantage of additional information and people who

Markus Prior is assistant professor of politics and public affairs, Woodrow Wilson School of Public and International Affairs, Princeton
University, 313 Robertson Hall, Princeton, NJ 08544-1013 (mprior@princeton.edu).
I am grateful to David Brady, the Center for the Study of Democratic Politics at Princeton, and Knowledge Networks for providing support
for this project. Larry Bartels, Matt Baum, Michael Delli-Carpini, John Geer, Marty Gilens, Jay Hamilton, Shanto Iyengar, Skip Lupia, Tali
Mendelberg, Diana Mutz, Sam Popkin, Wendy Rahn, and the anonymous referees deserve thanks for many helpful comments on previous
versions of this article.
American Journal of Political Science, Vol. 49, No. 3, July 2005, Pp. 577–592


C 2005 by the Midwest Political Science Association ISSN 0092-5853

577
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
578 MARKUS PRIOR

prefer other media content explains a puzzling empirical equalities have been overcome (e.g., National Telecom-
finding: despite the spectacular rise in available politi- munications and Information Administration 2002;
cal information, mean levels of political knowledge in Negroponte 1995). This ignores that greater media choice
the population have essentially remained constant (Delli leads to greater voluntary segmentation of the electorate.
Carpini and Keeter 1996; Gilens, Vavreck, and Cohen The present study suggests that gaps based on socioeco-
2004). Yet the fact that average knowledge levels did not nomic status will be eclipsed by preference-based gaps
change hides important trends: political knowledge has once access to new media becomes cheaper and more
risen in some segments of the electorate, but declined widely available. Gaps created by unequal distribution of
in others. Greater media choice thus widens the “knowl- resources and skills often emerged due to circumstances
edge gap” (Tichenor, Donohue, and Olien 1970). Follow- outside of people’s control. The preference-based gaps
ing Tichenor and colleagues’s formulation of the knowl- documented in this article are self-imposed as many peo-
edge gap hypothesis, numerous studies have examined ple abandon the news for entertainment simply because
the diffusion of information in the population and the they like it better. Inequality in political knowledge and
differences that emerge between more and less informed turnout increases as a result of voluntary, not circumstan-
individuals (for reviews see Gaziano 1997; Viswanath and tial, consumption decisions.
Finnegan 1996). According to some of these studies, tele- After presenting the theoretical framework that pre-
vision works as a “knowledge leveler” (Neuman 1976, dicts changes in knowledge and turnout, two studies test
122) because it presents information in less cognitively de- this theory empirically. The first study is based on a na-
manding ways (Eveland and Scheufele 2000; Kwak 1999). tionally representative survey I designed specifically to
To reconcile this effect with the hypothesis that more tele- measure content preferences and relate them to political
vision widens the knowledge gap, it is necessary to dis- knowledge and turnout. The second study relies on data
tinguish the effect of news exposure from the effect of collected by the National Election Studies and the Pew
the medium itself. In the low-choice broadcast environ- Research Center for the People and the Press. Although
ment, access to the medium and exposure to news were the preference measures are considerably weaker in the
practically one and the same, as less politically interested second study, they allow replication of the results for a
television viewers had no choice but to watch the news longer time period and different modes of (survey) data
from time to time (more on this below). As media choice collection.
increases, exposure to the news may continue to work
as a “knowledge leveler,” but the distribution of news
exposure itself has become more unequal. Access to the Theory
medium no longer implies exposure to the news. Televi-
sion news narrows the knowledge gap among its viewers. The basic premise of this analysis is that people’s media
For the population as a whole, more channels widen the environment determines the extent to which their me-
gap. dia use is governed by content preferences. According
The consequences of increasing media choice reach to theories of program choice, viewers have preferences
beyond a less equal distribution of political knowledge. over program characteristics (Bowman 1975; Lehmann
Since political knowledge is an important predictor of 1971) or program types (Youn 1994) and select the pro-
turnout and since exposure to political information moti- gram that promises to best satisfy these preferences. The
vates turnout, the shift from a low-choice to a high-choice simplest models distinguish between preferences for in-
media environment implies changes in electoral partici- formation and entertainment (Baum 2002; Becker and
pation as well. Those with a preference for news not only Schönbach 1989; Rubin 1984). In the low-choice broad-
become more knowledgeable, but also vote at higher rates. cast environment, most people watched news and learned
Those with a stronger interest in other media content vote about politics because they were reluctant to turn off the
less. set even if the programs offered at the time did not match
This study casts doubt on the view that the socioe- their preferences. One study conducted in the early 1970s
conomic dimension of the digital divide is the great- showed that 40% of the respondents reported watching
est obstacle to an informed and participating electorate. programs because they appeared on the channel they were
Many casual observers emphasize the great promise new already watching or because someone else wanted to see
technologies hold for democracy. They deplore current them (LoSciuto 1972). Audience research has proposed a
socioeconomic inequalities in access to new media, but two-stage model according to which people first decide to
predict increasing political knowledge and participation watch television and then pick the available program they
among currently disadvantaged people once these in- like best. Klein aptly called this model the “Theory of Least
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 579

Objectionable Program” (1972, 77). If television viewers news audiences tend to be smaller when competing en-
are routinely “glued to the box” (Barwise, Ehrenberg, and tertainment programming is scheduled (Webster 1984;
Goodhardt 1982) and select the best available program, Webster and Newton 1988). Baum and Kernell (1999)
we can explain why so many Americans watched televi- show that cable subscribers, especially the less informed
sion news in the 1960s and 70s despite modest political among them, are less likely to watch the presidential de-
interest. Most television viewing in the broadcast era did bates than otherwise similar individuals who receive only
not stem from a deliberate choice of a program, but rather broadcast television. According to my first hypothesis, the
was determined by convenience, availability of spare time advent of cable TV increased the knowledge gap between
and the decision to spend that time in front of the TV people with a preference for news and people with a pref-
set. And since broadcast channels offered a solid block of erence for other media content.
news at the dinner hour and again after primetime, many Internet access should contribute to an increasing
viewers were routinely exposed to news even though they knowledge gap as well. Although the two media are un-
watched television primarily to be entertained: doubtedly different in many respects, access to the Inter-
net, like cable, makes media choice more efficient. Yet,
Those viewers who can be counted on to watch while they both increase media users’ content choice, ca-
a news program are not at all drawn to their set ble TV and the Internet are not perfect substitutes for each
from their various pursuits by the appeal of the other. Compared at least to dial-up Internet service, cable
program; for the main part they are already watch- offers greater immediacy and more visuals. The web of-
ing television at that hour, or disposed to watch fers more detailed information and can be customized to
it then, according to the audience-research stud- a greater extent. Both media, in other words, have unique
ies that networks have conducted over the years. features, and access to both of them offers users the great-
(Epstein 1973, 90) est flexibility. For instance, people with access to both
media can watch a campaign speech on cable and then
Once exposed to television news, people learn about compare online how different newspapers cover the event.
politics (e.g., Neuman, Just, and Crigler 1992; Zhao and Depending on their needs or the issue that interests them,
Chaffee 1995). Although a captive news audience does not they can actively search a wealth of political information
exhibit the same political interest as a self-selected one online or passively consume cable politics. Hence, the ef-
and therefore may not learn as much, research on pas- fects of cable TV and Internet access should be additive
sive learning (Krugman and Hartley 1970) suggests that and the knowledge gap largest among people with access
even unmotivated exposure can produce learning (Keeter to both new media.
and Wilson 1986; Zukin and Snyder 1984). According There are several reasons why exposure to political
to Graber, “[p]eople who are exposed to large amounts information increases the likelihood that an individual
of news will remember many stories despite lack of in- will cast a vote on election day. Exposure increases po-
terest because mere exposure produces learning” (1988, litical knowledge, which in turn increases turnout (e.g.,
114). Hence, even broadcast viewers who prefer enter- Delli Carpini and Keeter 1996; Verba, Schlozman, and
tainment programs absorb at least basic political knowl- Brady 1995) because people know where, how, and for
edge when they happen to tune in when only news is whom to vote. Furthermore, knowledgeable people are
on. more likely to perceive differences between candidates
I propose that such accidental exposure should be- and thus less likely to abstain due to indifference (Pal-
come less likely in a high-choice environment because frey and Poole 1987). Independent of learning effects,
greater horizontal diversity (the number of genres avail- exposure to political information on cable news and po-
able at any particular point in time) increases the chance litical web sites is likely to increase people’s campaign in-
that viewers will find content that matches their prefer- terest (e.g., Bartels and Rahn 2000). Interest, in turn, af-
ences. The impact of one’s preferences increases, and “in- fects turnout even when one controls for political knowl-
discriminate viewing” becomes less likely (Youn 1994). edge (Verba, Schlozman, and Brady 1995). Entertain-
Cable subscribers’ channel repertoire (the number of fre- ment fans with a cable box or Internet connection, on
quently viewed channels) is not dramatically higher than the other hand, will miss both the interest- and the
that of nonsubscribers (Heeter 1985), but their repertoire information-based effect of broadcast news on turnout.
reflects a set of channels that are more closely related to My second hypothesis thus predicts a widening turnout
their genre preferences. Two-stage viewing behavior thus gap in the current environment, as people who prefer
predicts that news audiences should decrease as more al- news vote at higher rates and those with other preferences
ternatives are offered on other channels. Indeed, local increasingly stay home from the polls.
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
580 MARKUS PRIOR

Study 1: The News and unclear if panel analyses of political learning should in-
Entertainment Survey clude the same items in both surveys. If different items
are used, the two knowledge scales might tap different di-
Data and Measures mensions of knowledge. If the items are identical, the first
interview might encourage panelists to find (or better re-
Demonstrating the assumed dynamic nature of widening
member) the answers. I exactly repeated a few items and
gaps in knowledge and turnout with survey data requires
used similar types of questions in the remaining cases. The
a panel design, because only a panel design can show that
appendix lists all knowledge questions and provides sum-
knowledge levels changed over time. According to my the-
mary statistics for the resulting knowledge scales. Data on
ory, people with both a preference for news and access to
turnout come from the second panel wave of the N&E Sur-
cable and/or the Internet become more knowledgeable
vey. Respondents were asked whether or not they voted
over time, while the opposite holds true for people who
in the 2002 House elections. Cable access is coded as a
prefer entertainment programming and can choose be-
dummy variable based on the question “Do you have ei-
tween many different channels. According to a possible
ther cable or satellite television?” Eighty-one percent of
alternative hypothesis, more knowledgeable respondents
the respondents have cable access. Respondents are de-
who prefer news are more likely to have access to cable
fined to have Internet access if they connect to the In-
or the Internet in the first place than those who know
ternet using their home computers.3 Thirty-eight per-
less but are equally interested. If the reverse happens for
cent fall in this category; 33% have access to both new
respondents who prefer entertainment, an interaction ef-
media.
fect of content preferences and media access emerges, but
At any given time, television viewers must commit
the knowledge gap would not in fact have widened as a
to one particular program. They can either watch enter-
result of greater media choice.
tainment or news, but not both. This notion implies that
To evaluate these competing hypotheses, I designed
viewers evaluate the different programming options rel-
the News & Entertainment (N&E) Survey, a panel sur-
ative to each other. To assess respondents’ preferences for
vey of 2,358 randomly selected U.S. residents. It was con-
news relative to other programming genres, they were
ducted by Knowledge Networks in two waves in 2002 and
shown a list of 10 genres and asked to select the one they
2003. Knowledge Networks interviews national probabil-
liked best:
ity samples over the Internet by providing a large panel, se-
lected through Random Digit Dialing, with WebTV units Science Fiction shows like X-Files or Star Trek
and free Internet connections in exchange for taking sur- Voyager
veys. The participants for this study constitute a randomly Comedy/Sitcoms like Friends or The Simpsons
selected subset of the KN panel and are thus close to a Drama shows like ER or Law and Order
random sample of the U.S. adult population.1 The first Soap Operas like General Hospital or One Life to
survey wave was conducted in February and March 2002, Live
the second wave in April 2003. Of the 2,358 first-wave Reality TV shows like Survivor or Cops
respondents, the 1,957 who were still part of the Knowl- Sports
edge Networks panel in April 20032 were contacted again. Game Shows like Jeopardy or Who Wants to be a
Of these, 1,650 panelists were reinterviewed, generating a Millionaire?
reinterview rate of 84%. News
Each panel wave includes a large set of political knowl- Documentary programs on channels like History
edge questions about current political events, political Channel or Discovery Channel
officeholders, as well as institutions and processes. It is Music Videos
1
The household cooperation rate during the period of the two sur- After respondents marked their favorite genre, the
veys was 53%. The survey completion rate for the first wave was next screen showed them the remaining nine genres and
85%. For details on the sampling mechanism used by Knowledge
Networks, see Krotki and Dennis (2001). In a comparison of KN again asked for the most-liked. This procedure was re-
data to an RDD telephone survey, Krosnick and Chang (2001) found peated twice more, yielding a ranking of respondents’ four
the KN sample to be representative of the U.S. population in terms most-preferred television genres. Then, respondents were
of demographics and political attitudes.
given a chance to mark all of the remaining genres that
2
Almost all of the remaining 400 respondents had voluntarily with-
drawn from the KN panel by the time of the second survey wave.
3
Since few of the surveys KN panelists complete cover politics, this As Knowledge Networks conducts interviews through WebTV, all
panel attrition should not be of great concern for my analysis. Em- respondents in the N&E Survey have basic Internet access on their
pirically, respondents who withdrew did not differ significantly on television. Study 2 uses samples that include respondents without
key variables from respondents who remained active (see below). any web access. I discuss the implications of this difference below.
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 581

TABLE 1 The Effect of Content Preferences on Political Knowledge (Study 1)

(1) (2) (3)
Cable Access .43 (.44) .13 (.21) .40∗ (.19)
Internet Access .66∗ (.39) .42∗ (.19) .04 (.16)
Relative Entertainment Preference (REP) .03 (.12) −.05 (.06) −.02 (.05)
REP × Cable −.23∗ (.14) −.03 (.07) −.10∗ (.06)
REP × Internet −.22∗ (.11) −.15∗∗ (.05) −.03 (.05)
Wave 1 Political Knowledge .93∗∗ (.03) – –
∗∗ ∗∗
Civics Knowledge – .72 (.04) .62 (.03)
Political Efficacy .63∗ (.31) .57∗∗ (.15) .39∗∗ (.13)
∗ ∗∗
Education .15 (.06) .10 (.03) .17∗∗ (.03)
∗∗ ∗
Income .05 (.02) .02 (.01) .02∗∗ (.01)
R’s primary language is English 2.39∗∗ (.43) .22 (.25) −.06 (.18)
Gender −.34∗∗ (.14) −.27∗∗ (.07) −.33∗∗ (.06)
Age .03∗∗ (.005) .002 (.002) .003 (.002)
TV Households, Nielsen 2002-03 (in 1,000,000) .04 (.04) .03 (.02) .01 (.02)
R owns home computer .15 (.19) .01 (.10) .13 (.08)
Constant −3.98∗∗ (.71) 1.83∗∗ (.37) 2.01∗∗ (.30)
R2 .48 .33 .30
N 1,571 1,577 2,250
∗∗
p < .01, ∗ p < .05 (one-tailed).
Note: Cell entries are OLS coefficients and standard errors in parentheses.

they “really dislike.” For the purpose of this study, I am point per category) from the whole sample, indicating
only interested in the ranking of news vis-à-vis all other that respondents with a preference for entertainment were
genres. The appeal of news is greatest for respondents no more or less likely to complete the second survey.
who select news as their favorite genre. At the opposite Similarly, the difference in political knowledge between
end are respondents who specifically mark news as one first- and second-wave respondents is insignificant. With
they dislike.4 The measure of people’s relative preferences respect to the key variables, in other words, panel attrition
for entertainment over news (to which I will refer as Rel- is not a threat.
ative Entertainment Preference or REP) is thus coded 5 if
the respondent dislikes news, 4 if the respondent neither Analysis
dislikes it nor selects it as one of her top four genres, and
When television viewers have a choice between different
3, 2, 1, and 0 if the respondent selects news as her fourth-,
media content, their preferences should predict what pro-
third-, second- or most-liked genre, respectively.5 Exactly
grams they will select and, by extension, how much they
half of the respondents are indifferent to news, neither
will learn about politics. Relative Entertainment Prefer-
ranking it nor marking it as “disliked.” Three percent dis-
ence (REP) should thus impact political knowledge signif-
like news explicitly and the remaining respondents rank it
icantly more among cable subscribers than among non-
fourth (17%), third (14%), second (11%), or first (5%).
subscribers. Similarly, REP should have a stronger effect
The preference distribution among second-wave respon-
for Internet users. To test these hypotheses, political
dents differs minimally (by no more than 1 percentage
knowledge is regressed on cable and Internet access, REP,
and the interaction between REP and media access.6 The
4
This measure assumes that people either have preferences over dependent variable in the first model shown in Table 1
genres or can generalize from particular program examples for each
genre. I designed a second measure of content preferences that
asked about liking of news and entertainment in general without 6
mentioning genres or programs. The different measures of the same I include control variables to account for alternative explanations
concept are strongly correlated and yield essentially the same results. of political learning. As cable is not affordable for all Americans,
access is correlated with demographic variables, notably income and
5
I use a linear specification for simplicity. My theory does not nec- education. Also included are measures of the respondent’s media
essarily imply linear effects, and quadratic or logarithmic transfor- environment that do not directly affect media choice (e.g., the size
mations of REP indeed produce similar results. of the respondent’s media market).
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
582 MARKUS PRIOR

FIGURE 1 Political Knowledge and Relative Entertainment Preference

10.5

10
Predicted Political Knowledge Cable and Internet access

9.5

9
No new media access Internet only

Cable only
8.5

7.5
1st (5%) 2nd (11%) 3rd (14%) 4th (17%) Indifferent (50%) Disliked (3%)
Relative Entertainment Preference
(Labels indicate ranking of news)

Note: The graph plots the predicted values based on model 1 in Table 1.

is political knowledge as measured in the second wave of for respondents with a weak entertainment preference.
the N&E panel. All independent variables are from the (Individually, only one of them reaches conventional lev-
first wave and were measured a year before the depen- els of statistical significance, but their joint effect is dif-
dent variable. The joint effect of preference and media ferent from zero at p = .05.) To illustrate the interactions,
access as of spring 2002 on political knowledge a year Figure 1 plots the predicted values for different levels of
later is tested, while controlling for knowledge differ- media choice. REP has little effect on political knowledge
ences in 2002. Controlling for the lagged value of the for respondents with neither cable nor Internet access.
dependent variable reduces problems of selection bias In a media environment where you cannot choose be-
or reverse causation, because the change in knowledge tween entertainment and news, it does not matter very
between the two waves cannot affect first wave content much if you prefer one or the other. Among those with
preferences. access to both new media, on the other hand, the dif-
Model 1 provides clear support for the predicted ef- ference between highest and lowest REP amounts to a
fect of increased media choice. The two interactions of knowledge gain of about 27%. Compared to the modal
cable and Internet access with Relative Entertainment category—indifference toward the news—new media
Preference are statistically significant in the predicted di- users who rank news first or second are still 16% more
rection. The two main effects of media are sizable and knowledgeable.
positive, indicating that greater choice significantly in- The strong impact of content preferences is perhaps
creased political knowledge between the two panel waves most clearly demonstrated by a comparison with the ef-
fect of education, typically one of the strongest predictors
I excluded 23 respondents who revealed a lack of effort from of political knowledge (Delli Carpini and Keeter 1996).
all analyses of the N&E Survey. These respondents were identified Both variables have the same range (0–5 for REP, 1–6 for
by checking response set on two other measures of content prefer- education) and similar distributions (their standard de-
ences. Respondents answered five or six questions arranged on two
screens in grid form. The 23 respondents were excluded because viations are nearly identical at 1.26 for education and 1.27
they selected the same response for all items on a screen (even for REP). The size of each interaction effect exceeds the
though the orientation of questions varied). I also excluded six re- impact of education. The change in knowledge associated
spondents who answered all or all but one current events question
correctly in Wave 1, but only one or none at all in Wave 2. These with the difference between the most and the least ed-
dramatic differences probably arise because respondents lacked the ucated respondents is about 9%, compared to the 27%
motivation to complete the knowledge items in the second wave. change for the full range of REP among new media users.
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 583

In a high-choice environment, people’s content prefer- TABLE 2 The Effect of Content Preferences on
ences become better predictors of political learning than Turnout (Study 1)
even their level of education.
By assessing the change in political knowledge com- (1) (2)
pared to a baseline set by performance in the first panel Cable access .51 (.41) .38 (.43)
wave, the panel design provides a very rigorous test of my Internet access 1.09∗∗ (.42) 1.09∗∗ (.43)
hypotheses. A cross-sectional design still makes it possi- Relative Entertainment .09 (.11) .06 (.12)
ble to show static gaps in knowledge (rather than learning Preference (REP)
over time) among people with greater media access. In REP × Cable −.18# (.12) −.17 (.13)
the cross-sectional case, however, the “baseline” cannot REP × Internet −.30∗∗ (.12) −.26∗ (.12)
be knowledge assessed at an earlier point in time. Instead, Turnout in 2000 House – 1.05∗∗ (.16)
I use knowledge that respondents presumably acquired Election
earlier. Knowledge about governmental processes and in- Wave 1 Political .16∗∗ (.03) .16∗∗ (.03)
stitutions probably fits that description both because it Knowledge
is emphasized in civics education and because the correct Sense of Civic Duty .22 (.30) .002 (.002)
answers do not change as frequently as for questions about Education .23∗∗ (.06) .22∗∗ (.06)
current events. The dependent variable in columns 2 and Gender .22∗ (.13) .15 (.14)
3 of Table 1 is thus Wave 1 knowledge of current affairs, Age .03∗∗ (.005) .03∗∗ (.005)
while civics knowledge, also measured in Wave 1, serves as Income .04∗ (.02) .04∗ (.02)
a control variable.7 Column 2 includes only respondents R is married ∗∗
.41 (.14) .36∗∗ (.14)
who were also interviewed in Wave 2, while column 3 R’s primary language is .41 (.005) .50 (.48)
shows the same model for all Wave 1 respondents. English
The results for the cross-sectional analysis are weaker Racial Minority .28 (.17) .24 (.18)
than for the panel setup. Individually, only one of the TV Households, Nielsen −.01 (.04) .00 (.04)
two interaction effects is significant in each model. In 2002-3 (in 1,000,000)
both models, however, the addition of cable access, In- R owns home computer .004 (.17) −.06 (.17)
ternet access, and their interactions with REP increases R owns VCR .39∗∗ (.14) .39∗∗ (.15)
the model fit significantly [F(1,1562) = 3.5, p = .06 for ∗∗
R reports being registered 3.38 (.33) 2.77∗∗ (.34)
column 2, F(1,2235) = 3.1, p = .08 for column 3]. Bear- to vote
ing in mind later analyses in this article, it is useful to Constant −8.33∗∗ (.83) −8.06∗∗ (.85)
know that even cross-sectional data analysis can demon- Pseudo R2 .26 .29
strate the presence of wider knowledge gaps between news Log Likelihood −762.8 −734.4
and entertainment fans in a high-choice media environ- N 1,534 1,523
ment. Below, I take advantage of the greater availability
∗∗
of cross-sectional data to replicate the results obtained in p < .01, ∗ p < .05, # p < .07 (one-tailed).
Note: Cell entries are logit coefficients and standard errors in
this section. parentheses.

Turnout
models in Table 2 which regress turnout in the 2002 con-
Increasing penetration of new media technologies should gressional election on Relative Entertainment Preference,
increase the difference in turnout rates between those who access to cable and Internet, and their interactions (plus
like news and those who prefer entertainment, both be- demographic controls). Consistent with the hypothesis,
cause political knowledge and turnout are tightly related the logit coefficients for the interaction terms are neg-
and because exposure to political information motivates ative and statistically significant (at p < .01 for Inter-
people to vote. People who do not like news should be- net access and p < .07 for cable access). Including the
come less likely to go to the polls, while the reverse should lagged dependent variable (self-reported turnout in the
happen among politically interested citizens with access 2000 House election, as measured in the first wave) in
to cable or the Internet. This proposition is tested by the the model of 2002 turnout leaves the result largely un-
changed (column 2). This is strong evidence against the
7
The civics knowledge index is the number of correct responses to alternative hypothesis that people with extreme content
the questions about judicial review, presidential veto, maximum
number of presidential terms, and the more conservative party (see preferences were more likely (news-seekers) or less likely
appendix). (entertainment-seekers) to turn out all along and just
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
584 MARKUS PRIOR

FIGURE 2 Turnout and Relative Entertainment Preference

Predicted Turnout Probability 0.9 Cable and Internet access

0.8 Internet only

0.7 Cable only

No new media access

0.6

0.5

0.4
1st (5%) 2nd (11%) 3rd (14%) 4th (17%) Indifferent (50%) Disliked (3%)
Relative Entertainment Preference
(Labels indicate ranking of news)

Note: This graph plots predicted values from the model in column 2 of Table 3

obtained access to cable or the Internet earlier or at higher makes content preference a major influence on turnout.
rates than people with more moderate preferences. As in the case of political knowledge, the effect of prefer-
Once people obtain access to new media, content ence exceeds the effect of education, typically one of the
preference becomes a powerful predictor of turnout, as strongest predictors.
shown by the predicted values in Figure 2 (derived from
the second model in Table 2). Among respondents without
either cable or Internet access, the relationship between
Study 2: NES and Pew
content preference and turnout is not statistically differ-
The N&E Survey clearly supports the hypothesis that
ent from zero. For cable subscribers and Internet users,
greater media choice increased inequality in the distri-
in contrast, the effects of REP are sizable. Slightly more
butions of political knowledge and turnout. At the time
than half of the sample are indifferent towards news or
of the survey, most Americans had enjoyed access to cable
dislike the genre. With access to cable and Internet, their
television for many years, and the Internet had reached
probability of voting drops from about .69 to .64. Roughly
about half of all households. The main purpose of repli-
a third of the sample ranks news among their top three
cating the analysis using additional (secondary) data—in
genres. For them, cable and Internet access increases the
addition to providing a general robustness check—is to
probability of voting from about .65 to .82. Among new
determine if the results hold for different points in the
media users, in other words, moving from indifference
diffusion process of cable TV and the Internet.
to a fairly strong preference for news increases predicted
turnout by almost 30%. Including Wave 2 political knowl- Data and Measures
edge in the models reduces the magnitude of the effects by
less than a third. This result could arise for two reasons: ei- Survey data for the replications come from the National
ther because my measure of political knowledge does not Election Studies (NES) and the Media Consumption Sur-
capture the type of knowledge most relevant to encourag- veys (MCS) conducted biannually by the Pew Center for
ing turnout or because exposure to political information the People and the Press. The NES conducts face-to-face
motivates people to vote even when it does not affect their (and, in 2000, telephone) interviews; Pew uses phone in-
political knowledge. Regardless of which reason is more terviews. Unfortunately, the main drawback of secondary
important, this analysis shows that greater media choice data is a big one. Neither the NES nor Pew’s MCS include
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 585

direct measures of the key concept, entertainment pref- Jerry Springer, and Oprah (MCS 2000), watching Enter-
erence. Instead, only a few questions about exposure to tainment Tonight and MTV and reading People Magazine
entertainment shows are available. Hence, no measure de- (MCS 1998), and watching Hardcopy, Jerry Springer, and
rived from these data comes even close in construct valid- MTV (MCS 1996). All items used four-point response for-
ity to the measure in the N&E Survey. The two NES data mats (“regularly,” “sometimes,” “hardly ever,” “never”).
sets that ask about cable and Internet access—the NES REP cannot be computed for respondents who reported
1996 and 2000—include a variety of knowledge items. no information and entertainment viewing at all. These
For each data set, a knowledge measure is created by sum- respondents are excluded from the analysis (2.4% in the
ming the number of correct responses to 14 knowledge MCS 1996, 2% in the MCS 1998, 4.5% in the MCS 2000,
questions (listed in the appendix). Turnout in the NES 5.8% in the NES 1996, and 9.9 in the NES 2000).
1996 is assessed by respondent’s self-reported vote in the The resulting REP measures range from 0 to 1 and
1996 House election. Vote measures in the Pew surveys are have means of between .21 (NES 1996) and .28 (MCS
based on questions about voting in the last presidential 2000) with standard deviations between .19 and .28. Evi-
(and, in the MCS 1996, House) election. dently, the ratio measures are skewed towards news pref-
erence. This is not surprising because the surveys asked
Relative Entertainment Preference about very few entertainment programs and—with the
exception of the NES 2000 talk show item—only about
Although secondary data sources provide no direct pref- specific programs rather than genres. Many respondents
erence measures, it is possible to approximately infer re- surely reported low entertainment exposure not because
spondents’ relative preferences from the type of content they do not like entertainment, but because they were not
they report watching. High entertainment exposure per se asked about their favorite shows. Yet even though the REP
does not necessarily indicate a greater preference for enter- measures built from NES and Pew data underestimate
tainment. Respondents with high entertainment exposure people’s preference for entertainment, they still identify a
may watch a lot of television in general, without neces- set of respondents to whom entertainment programming
sarily preferring entertainment strongly to other genres. is relatively more attractive. To verify that measures of
Hence, the mix of exposure to different genres, rather REP are roughly comparable for the different data sets,
than absolute exposure to entertainment alone, should I examined the relationship between demographics and
be a (very rough) indicator of underlying relative pref- REP for each data source. Demographics have very simi-
erence. Respondents who watch entertainment programs lar effects on the more precise REP measure in the N&E
but little or no news probably do so out of a preference Survey and on the somewhat noisier ones in NES and Pew
for entertainment. High news exposure and low entertain- surveys.
ment exposure would reflect a preference for news. The
ratio of entertainment viewing to overall viewing thus
yields the best possible measure of relative entertainment Analysis
preference in NES and Pew data. This ratio is defined as As before, political knowledge (Table 3) and turnout
follows: (Table 4) are regressed on access to cable TV and the Inter-
net, Relative Entertainment Preference, and the respective
Relative Entertainment Preference (REP) interactions. Overall, the replications clearly support the
Entertainment Viewing hypothesis that greater media choice makes content pref-
= erences better predictors of people’s political knowledge
Entertainment Viewing + News Viewing
and turnout. All interactions of cable access and REP are
News Viewing is the average number of days per week in the predicted direction and six of the seven are statisti-
the respondent watched national and local news. The par- cally significant. The conditioning effect of Internet access
ticular operationalization of entertainment viewing de- is evident for turnout only. To summarize the joint effect
pends on the items in the surveys. For the NES 2000, of preferences and media access in these replications, I use
entertainment viewing is average exposure to Jeopardy, the average of the coefficients across the two knowledge
Wheel of Fortune, and “television talk shows such as Oprah models and the five turnout models, respectively, to graph
Winfrey, Rosie O’Donnell, or Jerry Springer.” The NES predicted values for the range of REP in Figure 3. Political
1996 measures entertainment viewing as the average daily knowledge of respondents without access to cable or In-
viewing of “Jeopardy or Wheel of Fortune” and Dr Quinn, ternet is unrelated to their content preference. For those
Medicine Woman. In the Pew data, entertainment view- with access to cable TV, on the other hand, moving from
ing is operationalized as watching Entertainment Tonight, low to high entertainment preference corresponds to a
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
586 MARKUS PRIOR

TABLE 3 The Effect of Content Preferences on a competing hypothesis, respondents with strong con-
Political Knowledge (Study 2) tent preferences may have had access to these media ear-
lier. In that case, the finding that political knowledge and
NES 1996 NES 2000 turnout are higher among news-seekers and lower among
Cable Access .26 (.22) .04 (.23) entertainment-seekers with greater media choice would
Internet Access .15 (.22) .39∗ (.21) not imply any changes in knowledge and turnout lev-
Relative Entertainment −.17 (.60) .06 (.59) els, only that these two segments were the most and least
Preference (REP) knowledgeable in the first place. The fact that Relative
REP × Cable −1.33∗ (.68) −1.38∗ (.64) Entertainment Preference consistently had a sizable effect
REP × Internet .80 (.68) .09 (.52) among cable viewers and Internet users between 1996 and
Education .44∗∗ (.05) .58∗∗ (.05) 2003 makes this claim hard to sustain. During this period,
Gender −.64∗∗ (.15) −.88∗∗ (.14) the percentage of cable subscribers rose from 69% in 1996
Age .03∗∗ (.01) .04∗∗ (.005) (MCS 1996) to 81% in 2003 (N&E Survey), and the per-
Income (MV imputed) .07∗∗ (.02) .08∗∗ (.02) centage of Americans with Internet access increased from
Income was imputed −.22 (.26) −.14 (.19) a mere 21% in April of 1996 (MCS 1996) to 62% in 2000
African American −1.35∗∗ (.25) −.78∗∗ (.24) (NES 2000). At some point in this diffusion process, even
Hispanic or Latino – −.41 (.31) people with moderate content preferences would have ob-
Other Minority −.15 (.45) −.07 (.30) tained access to new media. Yet this analysis has shown
R works 20+ hours per .10 (.18) −.38∗ (.17) significant effects both in 1996, when only 15% of pop-
week ulation had access to both cable and the Internet, and in
R lives in the South −.31∗ (.16) −.74∗∗ (.15) 2000, when 53% did. It thus becomes impossible to ar-
Party ID .02 (.03) −.02 (.04) gue that over the period of this analysis, respondents with
Strength of Party ID .37∗∗ (.08) .43∗∗ (.07) strong content preferences were always more likely to have
Frequency of political 1.60∗∗ (.24) 1.29∗∗ (.17) greater media choice. The effect of increased choice ap-
discussion with pears to be quite stable over the past decade, even though
friends or family more and more people gained access to cable TV and the
Constant 3.52∗∗ (.46) −.25 (.44) Internet.
R2 .29 .40 Successful replication minimizes concerns about one
N 1,284 1,334 other methodological issue: Knowledge Networks, the
company which conducted the N&E Survey, provides
∗∗
p < .01, ∗ p < .05 (one-tailed). its respondents with WebTV to conduct periodic inter-
Note: Cell entries are unstandardized OLS coefficients and
standard errors in parentheses. views. Strictly speaking, the N&E Survey therefore does
not permit inferences about the effect of Internet access.
Although WebTV is much less convenient than using a
regular browser and a mouse, even N&E respondents with
20% drop in political knowledge. The effect of REP among only WebTV access are not a genuine ‘no Internet’ group.
Internet users is insignificant and not even in the predicted This is not a major concern here because if anything it
direction. This result is the only instance in this article in would produce conservative estimates of the true effect.
which a hypothesis is not supported. The turnout effects If respondents with WebTV as their only means of using
mirror the N&E analysis quite precisely. The likelihood of the Internet did indeed behave just as Internet users with
turnout among people without access to cable or Inter- dial-up or broadband connections, then Internet access as
net is just above chance regardless of their entertainment I define it in the N&E Survey should not condition the ef-
preference. For those with cable and the Internet, the like- fect of content preference at all. That Internet access does
lihood of casting a vote drops from a three-quarter chance significantly boost the impact of content preference sug-
among people with the least interest in entertainment to gests that accessing the Internet on a television set is not
less than .4 among those with the strongest preference for quite the same thing as even a dial-up connection. Still,
entertainment. replicating results on data sets that include respondents
Beyond replicating the N&E results, the analysis of with no Internet connection at all bolsters the robust-
NES and Pew data adds an important element by demon- ness of my findings. Substantively, the more important
strating that access to cable and Internet conditioned the point is that differences in the efficiency of media access
effect of preferences as early as 1996 (the first year for probably matter as much as the difference between no ac-
which we have data to test the hypothesis). According to cess and presently typical dial-up connections. As Internet
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 587

TABLE 4 The Effect of Content Preferences on Turnout (Study 2)

Pew 1996
NES 1996 Pew 2000 Pew 1998
1996 House 1996 Pres. 1996 Pres. 1992 Pres. 1994 House
∗∗ ∗∗ ∗ ∗∗
Cable .49 (.21) .25 (.11) .18 (.10) .59 (.15) .36∗ (.14)
Internet .21 (.21) .62∗∗ (.11) .54∗∗ (.12) .58∗∗ (.18) .34∗ (.17)
∗
REP .44 (.54) .08 (.27) .05 (.24) −.84 (.38) −.64 (.40)
REP × Cable −1.28∗ (.61) −.57∗ (.29) −.26 (.28) −.99∗ (.44) −1.15∗ (.46)
REP × Internet .39 (.66) −.67∗∗ (.28) −.79∗∗ (.30) −1.11∗ (.52) −.10 (.53)
Education .31∗∗ (.05) .22∗∗ (.02) .26∗∗ (.02) .23∗∗ (.03) .22∗∗ (.03)
Income .07∗∗ (.01) .10∗∗ (.02) .11∗∗ (.02) .08∗∗ (.02) .15∗∗ (.02)
(Imputations for MV)
Income Missing .21 (.27) −.91∗∗ (.09) −.79∗∗ (.08) −.60∗∗ (.14) .07 (.14)
Gender .19 (.14) −.10 (.06) .13∗ (.06) −.21∗∗ (.08) −.13∗ (.08)
Age .04∗∗ (.005) .04∗∗ (.002) .04∗∗ (.002) .04∗∗ (.003) .05∗∗ (.003)
Employment Status .09 (.17) .19∗∗ (.07) .17∗∗ (.07) .26∗∗ (.08) −.19∗ (.08)
∗
Size of Town – −.07 (.03) −.02 (.03) −.02 (.04) .05 (.04)
Party ID .11∗∗ (.04) – – – –
∗∗
Strength of Party ID .61 (.08) – – – –
Constant −4.87∗∗ (.46) −3.25∗∗ (.21) −3.19∗∗ (.18) −2.45∗∗ (.26) −3.70∗∗ (.27)
−2 Log-likelihood 1,354.9 6,219.8 7,121.8 4,180.4 4,179.3
N 1,410 2,910 2,856 1,674 1,620
∗∗
p < .01, ∗ p < .05 (one-tailed).
Note: Cell entries are logit coefficients and standard errors in parentheses.

connections become faster, the impact of content prefer- spiral downwards; the circle is either virtuous or vicious.
ences is likely to increase further. The analyses presented here show that both are true. New
In sum, replication using a number of different data media do indeed increase political knowledge and involve-
sets was successful. The hypothesized conditioning ef- ment in the electoral process among some people, just as
fect of media choice has been shown for very different the optimists predict. Yet, the evidence supports the pes-
points in the diffusion process of cable TV and the Inter- simists’ scenario as well. Other people take advantage of
net (1996–2003), for different interview modes (phone, greater choice and tune out of politics completely. Those
face-to-face, and web-based) and political contexts (dur- with a preference for entertainment, once they gain access
ing election campaigns and mid-term), as well as for to new media, become less knowledgeable about politics
different measures of Relative Entertainment Preference. and less likely to vote. People’s media content preferences
become the key to understanding the political implica-
tions of new media.
Conclusion Analog cable systems and dial-up Internet
connections—currently the most common ways of
When speculating about the political implications of new new media access—are only the first technological steps
media, pundits and scholars tend to either praise the likely towards greater choice. Digital technology will multiply
benefits for democracy in the digital age or dwell on the the number of choices and the efficiency of choosing,
dangers. The optimists claim that the greater availability of thereby further increasing the impact of content prefer-
political information will lead more people to learn more ences on users’ choices. This study has begun to show
about politics and increase their involvement in the polit- this impact, but it has also raised new questions: How are
ical process. The pessimists fear that new media will make content preferences formed? How easily do they change?
people apolitical and provide mind-numbing entertain- What changes them? A measure of content preferences
ment that keeps citizens from fulfilling their democratic not used in this study was included in both panel waves
responsibilities. These two predictions are often presented of the N&E Survey. Overtime correlations of .6 (Pearson’s
as mutually exclusive. Things will either spiral upwards or r) suggest considerable preference stability. In light of the
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
588 MARKUS PRIOR

FIGURE 3 Pew and NES Replications, Average Effects on Political

Knowledge and Turnout

a) Political Knowledge
8

7.5
Internet only
Predicted Political Knowledge

7
No New Media

6.5
Cable and Internet access

6 Cable only

5.5

5
0 1
Relative Entertainment Preference

b) Turnout
0.8

Cable and Internet access

0.7
Predicted Vote Likelihood

Cable only
0.6
Internet only

No New Media
0.5

0.4

0.3
0 Relative Entertainment Preference 1

Note: This graph plots predicted values based on the average of the coefficients in Table 4

powerful effects content preferences had in this study, political interest or sense of civic duty) remain constant.
the next logical step is to examine if and how political By this logic, the decreasing size of the news audience is
socialization, education, or the content and style of not necessarily an indication of reduced political interest.
political news affect people’s preferences. Interest in politics may simply never have been as high
The decline in the size of news audiences over the as audience shares for evening news suggested. A com-
last three decades has been identified as cause for con- bined market share for the three network newscasts of
cern by many observers who have generally interpreted it almost 90% takes on a different meaning if one consid-
as a sign of waning political interest and a disappearing ers that people had hardly any viewing alternatives. It was
sense of civic duty. Yet changes in available content can “politics by default” (Neuman 1996, 19), not politics by
affect news consumption and learning even in the absence choice. Even the mediocre levels of political knowledge
of preference changes. People’s media use may change in a during the broadcast era (e.g., Delli Carpini and Keeter
modified media environment, even if their preferences (or 1996), in other words, were partly a result of de facto
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 589

restrictions of people’s freedom to choose their preferred Appendix

media content.
Ironically, we might have to pin our hopes of cre- Description of Knowledge Measures
ating a reasonably evenly informed electorate on that N&E Survey
reviled form of communication, political advertising.
Large segments of the electorate in a high-choice en- Political Knowledge, Wave 1, 12-item index
vironment do not voluntarily watch, read, or listen to “Which of the following countries shares a border with
political information. Their greatest chance for encoun- Afghanistan?” (Russia/Pakistan/Iraq Kazakhstan)
ters with the political world occurs when commercials
“In the war in Afghanistan, which of the following groups
are inserted into their regular entertainment diet. And
fought on the side of the coalition led by the United States
exposure to political ads can increase viewers’ polit-
and Britain?” (The Islamic Jihad/The Taliban/The North-
ical knowledge (Ansolabehere and Iyengar 1995). At
ern Alliance/Al-Qaeda)
least for the time being, before recording services like
TiVo, which automatically skip the commercial breaks, “Which of the following agencies was founded in the
or subscriber-financed premium cable channels with- wake of the terrorist attacks on September 11?” (Of-
out advertising become more widespread, political ad- fice for Homeland Security/Delta Force/National Secu-
vertising is more likely than news coverage to reach these rity Agency/Department of Civilian Defense)
viewers.
“Would you say there is more, less, or about the same
It might seem counterintuitive that political knowl-
amount of crime in the United States today as compared
edge has decreased for a substantial portion of the elec-
to 10 years ago?” (more/less/same)
torate even though the amount of political information
has multiplied and is more readily available than ever be- “Please give me your best guess for this next question. For
fore. The share of politically uninformed people has risen every dollar spent by the federal government in Wash-
since we entered the so-called “information age.” Televi- ington, how much of each dollar do you think goes for
sion as a medium has often been denigrated as “dumb,” foreign aid to help other countries?” (following Gilens
but, helped by the features of the broadcast environment, (2001), 5% or less is coded as correct)
it may have been more successful in reaching less inter- “Do you happen to know which party currently has the
ested segments of the population than the “encyclopedic” most members in the House of Representatives in Wash-
Internet. In contrast to the view that politics is simply too ington?” (Democrats/Republicans)
difficult and complex to understand, this study shows that
motivation, not ability, is the main obstacle that stands be- “Whose responsibility is it to determine if a law is consti-
tween an abundance of political information and a well- tutional or not?” (President/Congress/Supreme Court)
and evenly informed public. “How much of a majority is required for the US Senate
When differences in political knowledge and turnout and House to override a presidential veto?” (one-half plus
arise from inequality in the distribution of resources and one vote/three-fifths/two-thirds/three quarters)
skills, recommendations for how to help the information
“How many four-year terms can the president of the
have-nots are generally uncontroversial. To the extent that
United States serve?” (1/2/3/unlimited number of terms)
knowledge and turnout gaps in the new media environ-
ment arise from voluntary consumption decisions, rec- “In general, thinking about the political parties in Wash-
ommendations for how to narrow them, or whether to ington, would you say that Democrats are more conser-
narrow them at all, become more contestable on norma- vative than Republicans, or Republicans are more con-
tive grounds. As Downs remarked a long time ago, “[t]he servative than Democrats?” (Democrats more conserva-
loss of freedom involved in forcing people to acquire in- tive/Republicans more conservative)
formation would probably far outweigh the benefits to “On this page, you see four photographs. Do you happen
be gained from a better-informed electorate” (1957, 247). to know which of the photographs shows John McCain?”
Even if a consensus emerged to reduce media choice for the
public good, it would still be technically impossible, even “On this page, you see four photographs. Do you hap-
temporarily, to put the genie back in the bottle. Avoid- pen to know which of the photographs shows Vladimir
ing politics will never again be as difficult as it was in the Putin?”
“golden age” of television. Cronbach’s alpha = .70, mean = 8.0, s.d. = 2.4
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
590 MARKUS PRIOR

Political Knowledge, Wave 2, 15-item index NES 2000

“Who is the current secretary of defense?” (DonaldRums- Political Knowledge, 14-item index
feld/John Ashcroft/George Tenet/Colin Powell) Knows which party had majority in the House before
“Who is the current Senate majority leader?” (Bill election
Frist/Trent Lott/Dick Gephardt/John Kerry) Knows which party had majority in the Senate before
“Who is the Chief Justice on the U.S. Supreme election
Court?” (William Rehnquist/Clarence Thomas/Antonin Correctly responds that Trent Lott’s current job is Senate
Scalia/Anthony Kennedy) majority leader (open-ended)
“What office is currently held by Condoleezza (“Condi”) Correctly responds that William Rehnquist’s current job
Rice?” (U.S. Attorney General/National Security Ad- is chief justice of the Supreme Court (open-ended)
viser/Secretary of Defense/White House Chief of Staff)
Correctly responds that Tony Blair’s current job is prime
“What position is currently held by Ari Fleischer?” minister of England/Great Britain (open-ended)
(White House Chief of Staff/White House Press Secre-
Correctly responds that Janet Reno’s current job is attor-
tary/Education Secretary/Senior Presidential Campaign
ney general (open-ended)
Advisor)
Correct recall of incumbent/challenger
“What position is currently held by Alan Greenspan?” (Di-
rector of the Central Intelligence Agency/Treasury Secre- Correct recognition of incumbent/challenger (rated on
tary/Chairman of the Federal Reserve/Commerce Secre- feeling thermometer)
tary) Correctly identifies incumbent in the district
“For each of the following politicians, please indicate to Remembers something about incumbent in district
which party they belong: Tom Daschle, Christine Todd
Gives response other than ‘Don’t Know’ to “Do you hap-
Whitman, Howard Dean, Ralph Nader.” (Republican,
pen to know about how many years [incumbent] has been
Democrat, Green Party, Reform Party)
in the House of Representatives?”
“Do you happen to know which party currently has the
Gives response other than ‘Don’t Know’ to “How good a
most members in the House of Representatives in Wash-
job would you say U.S. Representative [NAME] does of
ington?” (Democrats/Republicans)
keeping in touch with the people in your district?”
“Do you happen to know which party currently has the
mean = 4.97, s.d. = 3.18
most members in the Senate?” (Democrats/Republicans)
NES 1996
“Whose responsibility is it to determine if a law is consti-
tutional or not?” (President/Congress/Supreme Court) Political Knowledge, 14-item index

“How much of a majority is required for the US Senate Knows which party had majority in the House before
and House to override a presidential veto?” (one-half plus election
one vote/three-fifths/two-thirds/three quarters) Knows which party had majority in the Senate before
“In general, thinking about the political parties in election
Washington, would you say that Democrats are more con- Correctly responds that Al Gore’s current job is vice-
servative than Republicans, or Republicans are more con- president (open-ended)
servative than Democrats?” (Democrats more conserva-
Correctly responds that William Rehnquist’s current job
tive/Republicans more conservative)
is chief justice of the Supreme Court (open-ended)
Note: One half of the respondents were randomly assigned Correctly responds that Boris Yelsin’s current job is pres-
to a visual condition in which photographs of the politi- ident (leader) of Russia (open-ended)
cians instead of their names were shown on screen. For the
purpose of this study, I simply average across this (ran- Correctly responds that Newt Gingrich’s current job is
dom) variation. Speaker of the House (open-ended)

Cronbach’s alpha = .83, mean = 8.3, s.d. = 3.8 Correct recall of incumbent/challenger
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NEWS VS. ENTERTAINMENT 591

Correct recognition of incumbent/challenger (rated on Gilens, Martin, Lynn Vavreck, and Martin Cohen. 2004.” See
feeling thermometer) Spot Run: The Rise of Advertising, the Decline of News,
and the American Public’s Perceptions of Presidential Can-
Correctly identifies incumbent in the district didates, 1952–2000.“ Presented at the Annual Meeting of the
Midwest Political Science Association.
Gives response other than ‘Don’t Know’ to “Did Repre- Graber, Doris A. 1988. Processing the News: How People Tame
sentative [NAME] vote for or against the welfare reform the Information Tide. 2nd ed. New York: Longman.
bill?” Heeter, Carrie. 1985. “Program Selection with Abundance of
Choice: A Process Model.” Human Communication Research
Gives response other than ‘Don’t Know’ to “How often 12(1):126–52.
has Representative [NAME] supported President Clin- Keeter, Scott, and Harry Wilson. 1986. “Natural Treatment and
ton’s legislative proposals?” Control Settings for Research on the Effects of Television.”
Communication Research 13(1):37–53.
Gives response other than ‘Don’t Know’ to “How good a
Klein, Paul. 1972. “The Television Audience and Program Medi-
job would you say U.S. Representative [NAME] does of ocrity.” In Mass Media and Society, ed. Alan Wells. Palo Alto:
keeping in touch with the people in your district?” National Press Books, pp. 76–79.
mean = 8.43, s.d. = 3.07 Krosnick, Jon A., and Lin Chiat Chang. 2001. “A Comparison
of the Random Digit Dialing Telephone Survey Methodol-
ogy with Internet Survey Methodology as Implemented by
Knowledge Networks and Harris Interactive.” Unpublished
References manuscript. The Ohio State University.
Krotki, Karol, and J. Michael Dennis. 2001. “Probability-Based
Ansolabehere, Stephen, and Shanto Iyengar. 1995. Going Nega- Survey Research on the Internet.” Paper presented at the 53rd
tive: How Attack Ads Shrink and Polarize the Electorate. New Conference of the International Statistical Institute, Seoul,
York: Free Press. South Korea.
Bartels, Larry M., and Wendy M. Rahn. 2000. “Political Attitudes Krugman, Herbert E., and Eugene L. Hartley. 1970. “Pas-
in the Post-Network Era.” Presented at the Annual Meeting sive Learning from Television.” Public Opinion Quarterly
of the American Political Science Association, Washington. 34(2):184–90.
Barwise, T. P., A. S. C. Ehrenberg, and G. J. Goodhardt. 1982. Kwak, Nojin. 1999. “Revisiting the Knowledge Gap Hypothe-
“Glued to the Box. Patterns of Tv Repeat-Viewing.” Journal sis: Education, Motivation, and Media Use.” Communication
of Communication 32(4):22–29. Research 26(4):385–413.
Baum, Matthew A. 2002. “Sex, Lies, and War: How Soft News Lehmann, Donald R. 1971. “Television Show Preference: Ap-
Brings Foreign Policy to the Inattentive Public.” American plication of a Choice Model.” Journal of Marketing Research
Political Science Review 96(1):91–110. 8(1):47–55.
Baum, Matthew A., and Samuel Kernell. 1999. “Has Cable Ended LoSciuto, Leonard A. 1972. “A National Inventory of Television
the Golden Age of Presidential Television?” American Politi- Viewing Behavior.” In Television and Social Behavior. Televi-
cal Science Review 93(1):99–114. sion in Day-to-Day Life: Patterns of Use, ed. Eli A. Rubinstein,
Becker, Lee B., and Klaus Schönbach. 1989. “When Media Con- George A. Comstock and John P. Murray. Washington: U.S.
tent Diversifies: Anticipating Audience Behaviors.” In Audi- Government Printing Office, pp. 33–86.
ence Responses to Media Diversification: Coping with Plenty, National Telecommunications and Information Administra-
ed. Lee B. Becker and Klaus Schönbach. Hillsdale: Lawrence tion. 2002. A Nation Online: How Americans Are Expanding
Erlbaum Associates Inc., pp. 1–27. Their Use of the Internet. Washington: U.S. Department of
Bowman, Gary. 1975. “Consumer Choice and Television.” Ap- Commerce.
plied Economics 7(3):175–84. Negroponte, Nicholas. 1995. Being Digital. New York: Knopf.
Delli Carpini, Michael X., and Scott Keeter. 1996. What Amer- Neuman, W. Russell. 1976. “Patterns of Recall among Television
icans Know About Politics and Why It Matters. New Haven: News Viewers.” Public Opinion Quarterly 40(1):115–23.
Yale University Press.
Neuman, W. Russell. 1996. “Political Communication Infras-
Downs, Anthony. 1957. An Economic Theory of Democracy. New tructure.” The Annals of the American Academy of Political
York: Harper. and Social Science 546(July):9–21.
Epstein, Jay Edward. 1973. News from Nowhere. New York: Ran- Neuman, W. Russell, Marion R. Just, and Ann N. Crigler. 1992.
dom House. Common Knowledge: News and the Construction of Political
Eveland, William P., Jr., and Dietram A. Scheufele. 2000. “Con- Meaning. Chicago: University of Chicago Press.
necting News Media Use with Gaps in Knowledge and Par- Palfrey, Thomas R., and Keith T. Poole. 1987. “The Relation-
ticipation.” Political Communication 17(3):215–37. ship between Information, Ideology, and Voting Behavior.”
Gaziano, Cecilie. 1997. “Forecast 2000: Widening Knowl- American Journal of Political Science 31(3):511–30.
edge Gaps.” Journalism and Mass Communication Quarterly Rubin, Alan M. 1984. “Ritualized and Instrumental Television
74(2):237–64. Viewing.” Journal of Communication 34(3):67–77.
Gilens, Martin. 2001. “Political Ignorance and Collective Policy Sunstein, Cass R. 2001. Republic.Com. Princeton: Princeton
Preferences.” American Political Science Review 95(2):379– University Press.
96.
 15405907, 2005, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1540-5907.2005.00143.x by The Chinese University of Hong Kong, Wiley Online Library on [19/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
592 MARKUS PRIOR

Tichenor, Philip J., George A. Donohue, and Calice A. Olien. Webster, James G., and Gregory D. Newton. 1988. “Structural
1970. “Mass Flow and Differential Growth in Knowledge.” Determinants of the Television News Audience.” Journal of
Public Opinion Quarterly 34(2):149–70. Broadcasting & Electronic Media 32(4):381–89.
Verba, Sidney, Kay Lehman Schlozman, and Henry E. Brady. Youn, Sug-Min. 1994. “Program Type Preference and Program
1995. Voice and Equality: Civic Voluntarism in American Pol- Choice in a Multichannel Situation.” Journal of Broadcasting
itics. Cambridge: Harvard University Press. & Electronic Media 38(4):465–75.
Viswanath, Kasisomayajula, and John R. Finnegan, Jr. 1996. Zhao, Xinshu, and Steven H. Chaffee. 1995. “Campaign Adver-
“The Knowledge Gap Hypothesis: Twenty-Five Years Later.” tisements Versus Television News as Sources of Political Issue
In Communication Yearbook, ed. Brant Burleson. Thousand Information.” Public Opinion Quarterly 59(1):41–65.
Oaks: Sage Publications, pp. 187–227. Zukin, Cliff, and Robin Snyder. 1984. “Passive Learning: When
Webster, James G. 1984. “Cable Television’s Impact on Audience the Media Environment Is the Message.” Public Opinion
for Local News.” Journalism Quarterly 61(2):419–22. Quarterly 48(3):629–38.