This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Political Communication, 20:149–171, 2003
Copyright 2003 Taylor & Francis
1058-4609/03 $12.00 + .00
DOI: 10.1080/10584600390211172 Any Good News in Soft News? The Impact of Soft
News Preference on Political Knowledge
MARKUS PRIOR Several scholars, most notably Matt Baum, have recently argued that soft news formats contribute to democratic discourse, because they attract viewers who would
otherwise not be exposed to news at all. I extend Baum’s approach in two ways.
First, Baum’s theory postulates that people’s appreciation of entertainment is one of
the factors determining news exposure and, by extension, attention to politics, but he
does not analyze the underlying utility calculation directly. I create a measure of
entertainment preference and examine its impact on people’s preferred news formats. Second, while Baum’s analysis is restricted to attention paid to politics, I
assess the effect of soft news preference on political knowledge. If soft news leads
people to pay more attention to the “entertaining” aspect of politics, but does not
actually produce any learning effects, the suggested positive consequences of soft
news would have to be qualified. The main data source for this article is a survey of
2,358 randomly selected U.S. residents conducted by Knowledge Networks in February and March 2002. Results show that people like soft news for its entertainment
value but that soft news programs are still not very popular compared to hard news
and pure entertainment. More critically, there is only very limited evidence that
viewers actually learn from soft news. The positive consequences of soft news for
the political process remain to be demonstrated.
Keywords entertainment preference, hard news, news preference, political knowledge, soft news So-called “soft news” programs have proliferated in the last decade or so. According to
a definition by Tom Patterson (2000, p. 4), soft news is “typically more sensational,
more personality-centered, less time-bound, more practical, and more incident-based than
other news.”1 While critics have rushed to debate the normative implications of soft
news, empirical assessments of the consequences of greater diversity of news formats
are still rare. In the most important analysis to date, Matt Baum (2002b) shows that
some people who would otherwise not watch any news at all pay attention to soft news
I am grateful to David Brady, who made it possible for me to design the survey that provides the data for this article. The writing of the article was supported in part by a Pew Summer
Writing Fellowship from the Pew Internet & American Life Project. My thanks to Doug Arnold,
Larry Bartels, Matt Baum, Michael Delli Carpini, and Jay Hamilton for many helpful comments.
Markus Prior is a Ph.D. Candidate in the Department of Communication at Stanford University and a Visiting Fellow at the Center for the Study of Democratic Politics, Woodrow Wilson
School, Princeton University.
Address correspondence to Markus Prior, Woodrow Wilson School, Princeton University,
Princeton, NJ 08544-1013. E-mail: [email protected] 149 150 Markus Prior coverage of wars and foreign crises. Baum argues that some people do not watch hard
news programs because the opportunity costs from forfeiting payoffs from entertainment
are too high. By focusing on the more entertaining, shocking, or scandalous aspects of
politics, soft news offers these people an alternative that maximizes their utility because
it combines entertainment and information.
In this article, I extend Baum’s approach in two ways. First, I measure people’s
preferences for information and entertainment directly. Since Baum must rely on secondary data analysis, he simply examines how people’s self-reported soft news exposure
affects their attention to various news topics. He cannot test his theoretical argument
about the underlying utility calculation directly. The theory assumes, however, that people’s
appreciation of entertainment is one of the factors determining news exposure and, by
extension, attention to politics. My goal is to measure entertainment preference and gauge
its impact on people’s preferred news formats.
Second, Baum’s dependent variable is the amount of attention people pay to wars
and foreign policy crises. While getting people’s attention is an important prerequisite
for creating an informed public, the crucial step is learning. If soft news does not only
reach people who previously avoided the news, but also increases their knowledge of
some political matters, it could arguably play an important role in a democratic system.
If, on the other hand, soft news leads people to pay more attention to the “entertaining”
aspect of politics but does not actually produce any learning effects, the suggested positive consequences of soft news would have to be qualified. To address this question, I
measure people’s knowledge of several “hard” and “soft” political topics.
Data for this article come from a Web-based survey of 2,358 randomly selected
U.S. residents. The survey was conducted by Knowledge Networks (KN) in February
and March 2002. Knowledge Networks interviews national probability samples of the
U.S. population over the Internet by providing a large panel, selected through random
digit dialing (RDD), with WebTV units and free Internet connections in exchange for
taking surveys. The participants for this study constitute a randomly selected subset of
the KN panel, and hence a random sample of U.S. residents.2 How Popular Is Soft News?
Before delving into the analysis of why people watch soft news and how much they
learn from it, it is useful to assess the popularity of soft news programs compared to
more traditional formats. As the following analyses show, the audience for soft news,
far from rising to ever greater majorities of news users, is relatively small and has
grown slowly, if at all, in the past couple of years.
To measure the popularity of soft news, all respondents were asked to rank several
types of news. First, each respondent was presented with a list of eight different news
formats and asked to mark the format he or she liked most.3 Among the eight formats,
two can be classified as soft news: talk shows and so-called “infotainment” programs
which primarily report on lifestyle, wardrobe, and scandals of Hollywood stars and the
more glamorous politicians.4 The exact wording of the eight options is shown in Table
1. The column labeled “first choice” presents the distribution of most-liked news formats. It confirms past findings that television is the most popular news medium and that
more people prefer local news than network news. Very few people, no more than 5%
of the sample, like soft news programs best.
On the next screen, respondents were shown all options, except the one they just
selected, and asked to mark the format they liked second best. This was repeated twice 151 51.9
3.5 % who
Top 3 33.8
15.4 25.3 2.5
15.6 % who
21 12 25
26 % of TV
to news 35
34 30 62
on an average
3.1 2.8 4.3
3.2 Days in the
least one news
program on TV 2.3
2.6 1.1 2.7
3.9 Days in the
newspaper Of those who pick format as 1st choice Note: All respondents were asked to select the news format they like most from these 8 options. The question was repeated three times to create a ranking of
respondents’ four most-liked news formats. Finally, respondents were asked to mark any of the remaining formats that they dislike. The number of observations is
between 2,342 and 2,349. Reading daily newspapers
Reading current affairs magazines
(such as Time or Newsweek)
Watching local news on TV
Watching national news on TV
Watching late night or daytime talk
shows (such as Letterman or Oprah)
Watching “Infotainment” programs
(such as Hardcopy or E-Tonite)
Talk shows or infotainment
Listening to the news on the radio
Getting news from Internet sites % who
1st choice Of all respondents Table 1
Preferred news formats Any Good News in Soft News? 151 152 Markus Prior more, so that each respondent effectively provided a top-four ranking of news formats.
Column 2 of Table 1 shows the percentage of people who ranked a format among their
top three. Thirty percent of the respondents listed at least one of the soft news options in
their top three. This part of the interview ended by giving respondents a chance to mark
any of the remaining formats they explicitly disliked. Column 3 shows that soft news
formats were mentioned most often as a particularly disliked news format. While people
either like traditional formats, notably local news, national news, and newspapers, or
feel indifferent toward them, soft news seems to be loved by some, hated by others.
Traditional hard news formats remain far more popular than soft news.5
At first glance, Nielsen ratings, such as those in Table 2 adopted in part from Baum
(2002b, p. 93), seem to contradict this conclusion. Soft news programs appear to be
almost as widely watched as news programs. Baum (2002b, p. 93), for example, writes
that “Entertainment Tonight and Oprah Winfrey are watched by about as many households as the evening newscasts of the major networks.” It is less than obvious how to
interpret the ratings, however. Since all hard news programs air at the same time in most
markets, the ratings in Table 2 imply a hard news audience in 1998 of at least 19.5
million households (not counting the Lehrer Newshour, CNN, Fox, and other cable channels).
The soft news programs on the list, in contrast, do not air simultaneously. Regis & Kelly
is on in the morning, Rosie O’Donnell (2 or 3 p.m., show ended in 2002) and Oprah in
the afternoon (4 p.m.), and Extra (7 p.m.) and Entertainment Tonight (7:30 p.m.) in the
evening, but not in the same time slot in many markets.6 It is possible, therefore, that
some of the same 6 million households that watch Oprah also tune in to Extra or Entertainment Tonight. Rather than parity between hard news and soft news, this scenario
would imply at least a 3-to-1 ratio of hard to soft news viewers.7
The soft news ratings in Baum’s (2002b) article are averages for the first six months
of 1999. Ratings for the same period for 2002 are not publicly available. To provide a
rough overtime comparison nonetheless, Table 2 provides ratings for two shorter periods
in the first half of 2002: The sweeps period in May as an upper bound (because networks make an extra effort during sweeps to attract viewers, as future advertising rates
are set) and a week in the summer as a lower bound (because vacationing and the
networks’ airing of re-runs generally depress television viewing). Audiences for these
particular soft news programs have declined somewhat between 1999 and 2002, but the
decline is modest when the first half of 1999 is compared to the sweeps period in May
2002. While systematic comparison is seriously hampered by the unavailability of published Nielsen data, Table 2 suggests that soft news has lost more of its audience than
Audiences for cable news still appear to be tiny in comparison with the nightly
audience for network news (Table 2). However, this is at least partly a function of the
measure. Comparing daily averages for cable news networks, such as those listed in
Table 2, with averages for half-hour network newscasts is an extremely skewed comparison. It is more meaningful to compare 9.4 million daily NBC Nightly News viewers
with 1.8 million who watched The O’Reilly Factor on an average night in May 2002,
1.3 million Larry King Live viewers, or 1.1 million who watched The Fox Report with
Shepard Smith. On the harder side, NewsNight with Aaron Brown (713,000 viewers),
The News With Brian Williams (combined 635,000 viewers on MSNBC and CNBC),
and Crossfire (559,000 viewers) also attracted a combined 1.9 million viewers.9 Even
hourly averages underestimate the number of people who tune in to cable news channels
several times a day for shorter news updates. CNN, Fox News, MSNBC, CNBC, and
CNN Headline News together had an estimated 11.9 million unique viewers during prime 153 4.9/5,170,000
2.8/2,954,000 0.4/376,000 (1998) 6.2/6,500,000
(see note) 5.9/5,864,000
3.6/3,624,000 Rating/households 2002
8,925,000 Viewers 501,000 6.7/7,053,000
3.5/3,693,000 Rating/households 2002
May sweeps 9,360,000
9,407,000 Viewers Note. 2002 data for hard news are based on Nielsen estimates of average viewers per week reported in the Washington Post and/or the
St. Petersburg Times. I estimated rating points and household estimates based on approximate conversion factors (1 rating point = 1,055,000
households = 1,400,000 viewers). Because of the complicated (and unpublished) way in which Nielsen calculates ratings points, households, and viewers, all estimated numbers are mere approximations. My estimates are shown in italics. Estimates for soft news are based
on rating points published in Electronic Media. For June/July 2002, E-Tonite and Extra ratings are for the week of July 8–14, Oprah
ratings for July 1–7, Regis and Kathy Lee and Rosie O’Donnell for June 17–23. Cable news data are also from Electronic Media. CNN
MSNBC Hard News
ABC nightly news
CBS nightly news
NBC nightly news Soft news
Live with Regis
and Kathy Lee
Rosie O’Donnell Rating/households 1999
(from Baum, 2002b) Table 2
Select Nielsen ratings of hard and soft news programs Any Good News in Soft News?
153 154 Markus Prior time (8–11 p.m.) and 31.2 million unique viewers per day in the 2002–2003 season.
Even during the network news (half) hour between 6:30 and 7 p.m., 3.2 million different viewers watched at least six minutes of cable news (compared to an average 34.7
million unique viewers for the three network newscasts.)10 According to these numbers,
cable news—some of it pretty soft, some arguably as hard as network news ever was—
is more popular than a 0.4 rating suggests, while the growth for soft news audiences has
slowed down, and in some instances even stopped.
Survey data collected by Pew over the last decade and shown in Figure 1 confirm the
continued popularity advantage of hard news over soft news. The percentage of Americans who say that they follow network news regularly is about 3 to 4 times higher than the
percentage of regular viewers of any particular talk show or entertainment news program. It is even more interesting to observe that the popularity of soft news, at least of
Pew’s selection of it, did not increase noticeably over time. The percentage of selfdeclared E-Tonite regulars remained at about 10% over the last decade. Viewing of
daytime talk shows, be they hosted by Ricki Lake, Jerry Springer, or Oprah Winfrey, even
appears to have decreased in the past couple of years, confirming the conclusion drawn
from the over-time comparison of Nielsen ratings. The unreliability of self reports, social
desirability bias, and the substitution of old soft news favorites by new soft news favorites
caution against reading too much into these trends. But the main story in Figure 1 is one
of the diversification of new formats, not of the decline of the hard news audience. To
be sure, audiences for network news and even local news have declined, but at the same
time new cable news channels have emerged and attracted increasing numbers of regular
viewers. Unless one declares cable news universally soft—and a recent content analysis
commissioned by the Lehrer Newshour (ADT Research, 2002) does not support that
proposition—both Nielsen ratings and Pew survey data point to a diversification of the
news audience, not an across-the-board decrease in hard news viewing.
If anything, the ranking task described earlier exaggerates the popularity of soft
news. Since everyone in the sample was asked to rank the different news formats, the
results in Table 1 are influenced by respondents who do not watch a lot of news. In
order to relate preferences to actual news viewing, respondents were asked to report
how much television they watch on “an average day” and, on a subsequent interview
screen, to list what percentage of their viewing time they usually devote to news.11
Columns 4 and 5 of Table 1 list the means as a function of preferred news formats.
Results show that people who prefer soft news devote considerably less time to news
than all other groups, both in relative and in absolute terms. For people who rank infotainment
or talk shows as their most-liked news format, news makes up only 12% of their daily
television viewing, compared to 34% for people who like national TV news best. In
absolute terms, they spend an estimated 31 minutes every day watching news, compared
to 77 minutes for people who rank national TV news highest.12 And while newspaper
reading substitutes for TV news in some subgroups according to columns 6 and 7, people
with a preference for soft news rank lowest in both watching TV news programs and
reading daily newspapers. Clearly, soft news is popular among light news viewers who
devote very little time to hard news and spend most of their viewing hours following
programming other than news. Why People Like Soft News
While not as popular as hard news, soft news still attracts millions of viewers and ranks
among the top 3 news formats for about a third of the population. This section attempts Any Good News in Soft News? 155 (a) NetworkNetw ork and Cable New s
(a) and Cable News 100
Percent of Regular Viewers
Percentageof Regular Viewers 80 Local New s 70
50 Netw ork 40
30 CNN 20 Fox New s 10
1992 1994 1996 1998 2000 2002 Year (b) News and Soft Soft News
(b) Cable Cable News and News 35 Percentage of Regular Viewers
Percent of Regular Viewers CNN
10 Ricki Lake or
Jerry Springer MSNBC
Rosie O'Donnell or
Oprah Winfrey E-Tonite or Access 5
1992 1994 1996 1998 2000 2002 Year Figure 1. Regular viewers of network news, cable news, and soft news. (Source: Various Pew
surveys. Cell entries are percentages of respondents who say they watch program “regularly.”) to explain who these viewers are and why they like soft news. Many scholars and observers see soft news as a way to reach audiences that would otherwise not watch, read,
or hear news. Soft news informs and entertains at the same time and, according to this
argument, appeals to people unwilling to give up on entertainment, even while they are
watching news. Liking of news formats, then, is a function of two different individual 156 Markus Prior traits: preference for news and preference for entertainment. People with a preference
for news are unlikely to enjoy formats whose mission is to entertain, shock, and scare,
as well as inform—unless they also happen to have a strong preference for entertainment. Conversely, people with a marked preference for entertainment will likely pick
the soft news options when asked to rank news formats—unless their news preference
outweighs the appeal of entertainment.
To test these hypotheses about the underlying preference structure of news consumption, I draw on two sets of items aimed at measuring respondents’ fondness of
news and (television) entertainment. The items are listed in Table 3. Both scales have
reliability levels of greater than .7, which is generally considered appropriate. Substantively more important is the fact that the two scales are not correlated at all (Pearson’s
r = .07). It is clearly not the case, then, that the two scales are simply reversed measures
of the same dimension. High entertainment preference does not imply low news preference (or political interest, which correlates with entertainment preference at r = –.08).13
To the contrary, a person who likes news is empirically not more likely to dislike entertainment than to enjoy entertainment.
The measures of entertainment and news preference are reliable and almost
orthogonal, but do they help us understand why people prefer one news format or
another? For each news format, Figure 2 graphs the mean values on both scales among
people who rank the format highest. Infotainment and talk shows are to the right of the
other formats, indicating that people who favor them have a higher entertainment preference. Ranking talk shows highest is associated with the lowest news preference of all
formats. Newspapers, news magazines, and national TV news are relatively close together at higher levels of news preferences than the other formats.
In order to assess how the likelihood of picking a particular news format varies
with entertainment and news preference, several of people’s preferred news formats are
regressed on the two scales and a variety of control variables. The dependent variable is
Entertainment preference and news preference
Entertainment preference (Cronbach’s α = .73)
It bothers me to miss my favorite TV prime time show.
I watch television to be entertained.
I enjoy spending an entire evening watching movies and TV shows.
Watching movies or prime time shows on television is unsatisfying. (reversed)
After watching my favorite TV show, I often can’t wait to see the next episode.
News preferences (Cronbach’s α = .72)
Following the news is not my idea of fun. (reversed)
I like complex news stories even if they require my full attention.
I prefer detailed news coverage to news headlines even if it requires more time.
I hate to miss the news.
Correlation between the two indices r = .07
Note. The two indices are created by averaging agreement with the respective items on a
four-point scale (disagree strongly to agree strongly). Any Good News in Soft News? 157 1 News Preference National TV New s
New s Magazines New spapers Radio New s Local New s
Infotainment Internet New s Talk Show s 0
0 1 Entertainment Preference Figure 2. Underlying preference structure of respondents’ preferred news formats. coded one if the respondent listed the particular format among his or her top 3 formats,
zero if not. As is evident in Table 4, a distinct user profile emerges for each of the four
news formats examined here—soft news, national news, newspapers, and local news.
Soft news programs are selected by people with a high preference for entertainment and
a low appreciation of news (column 1). Results for local news, interestingly, resemble
results for soft news very closely. Local news, too, is preferred by people who like to be
entertained and lack a strong preference for news (column 2). The opposite pattern occurs
for newspapers. Here, a preference for entertainment makes it less likely that a respondent selects newspapers as a source of news (column 4). Entertainment preference does
not affect at all how much people like national TV news (column 3), but news preference is an even stronger predictor than it was for newspapers.
Figures 3 and 4 show graphically how the likelihood of ranking different news
formats among the top three14 varies with entertainment preference (Figure 3) and news
preference (Figure 4), while holding all other variables at their means. Figure 3 illustrates
the importance of entertainment preference in selecting soft news. People with no appreciation of television as entertainment have a .1 probability of selecting soft news. Among
those with the greatest fondness of entertainment, this probability exceeds .5. Entertainment preference also affects liking of newspapers, albeit in the opposite direction and at
a slightly lower magnitude than soft news. Local news and in particular national news,
on the other hand, are well liked independent of one’s entertainment preference. Figure
4 emphasizes the large impact of news preference on liking national TV news and soft
news. At a probability of .5, both are equally likely to be picked among the top 3 news
formats by people with the lowest news preference. At maximum news preference, people
select national news with a probability of .9 and soft news with a probability of .2. The
declining popularity of local news at increasing levels of news preference is statistically
significant, but smaller in magnitude. In sum, a two-dimensional structure of underlying
preferences allows us to distinguish with impressive precision people who like hard
news from those who prefer soft news. 158 Markus Prior
Who likes what? Predicting preferred news types
Soft News Entertainment preference
News preference Local News 2.65***
(.26) National News Newspaper Civic duty –.31
(.21) Education –.16***
(.04) Respondent is
African American .61***
(.14) Respondent is not White
or African American .61**
(.19) Age –.02
(.02) Age2 .00
(.00) Income –.04**
(.01) Female .31**
(.09) Constant .12
2,290 % correctly classified
2,291 Note. Cell entries are logit coefficients and standard errors in parentheses.
*p < .05; **p < .01; ***p < .001. Soft News and Political Knowledge
Up to this point, results complement Baum’s analysis. The criteria to establish the “success” of soft news are fairly weak, however. I have shown that people who like television for its entertainment value are more likely to rank soft news among their preferred
news formats. Baum demonstrates that people who watch a lot of soft news report following news stories about foreign crises and domestic scandals just as much as regular
hard news consumers. A more rigorous criterion to assess the “success” of soft news is
whether those who like it or say they watch it also learn something from the experience. Any Good News in Soft News? 159 Predicted Probability of Selecting News Format Among Top 3 1 Local News
0.8 National News 0.6 Newspapers 0.4 Soft News 0.2 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Entertainment Preference Figure 3. Effect of entertainment preference on preferred news format. Predicted Probability of Selecting News Format Among Top 3 1 Local News
0.8 0.6 National News
Newspapers 0.4 0.2 Soft News 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 News Preference Figure 4. Effect of entertainment preference on preferred news format. 0.9 1 160 Markus Prior According to a weak learning criterion, soft news is a “success” if its viewers learn
about the types of issues prominent on soft news, roughly sex, drugs, and celebrities. A
tougher learning criterion declares soft news a “success” only if it increases knowledge
of hard news topics as well.15 To assess soft news against these criteria, I draw on a set
of knowledge questions summarized in Table 5.
Table 5 contains the question wording, response options, and marginals for four
knowledge questions about domestic political scandals, three questions about the “War
on Terrorism,”16 and four standard hard news questions. It is immediately evident that
the American public is fairly well informed about the scandals of their politicians. Four
in five Americans know that Jesse Jackson confessed to having had an illegitimate child
and that Jenna Bush was arrested for underage drinking. Still 60% know that a flight
attendant claimed to have had an affair with Gary Condit. It is obviously difficult to
compare performance on soft and hard news questions, but it is telling that more Americans know about Condit’s (alleged) affairs, Jenna’s drinking, and Jackson’s child than
about the current majority party in the U.S. House. That this is more than an artifact of
question selection is clear from a quick look at the news stories respondents recalled
best in the 1989 NES Pilot Study. The Jim Bakker trial, former President Reagan’s
surgery, and the trial of actress Zsa Zsa Gabor were recalled by more than 70% of the
respondents, while less than 20% remembered any details from stories about a Supreme
Court decision on the death penalty or a congressional debate on health care (Price &
Zaller, 1993). The amount of soft news clearly increased between 1989 and 2002; the
public’s knowledge level of scandals and courtroom drama has apparently always been
To examine the effect of people’s preferred news formats on their political knowledge, I regress each knowledge item on the rankings of the eight news formats17 as well
as various demographic and attitudinal control variables. It is of course true that people
who prefer hard news tend to be more knowledgeable in the first place. I control for this
selection effect by including education, political interest, sense of civic duty, and internal efficacy in the models.18 Logit estimates are shown in Tables 6–8. Table 6 shows the
results for the four soft news questions. The weak knowledge criterion is clearly not met
in any systematic way. People who like talk shows are more likely to know that Jenna
Bush was arrested for underage drinking.19 The ranking of infotainment programs has a
positive effect on knowing that a flight attendant claimed to have had an affair with
Gary Condit. The two results are the only indication that a preference for soft news
formats increases knowledge in the domain of domestic political scandals. Price and
Zaller (1993) also found that readers of People magazine were more likely to recall
details about the Zsa Zsa Gabor story. The fact that the only significant effect of a
preference for infotainment in the entire study occurs for the question about Condit’s
alleged affair with a flight attendant might suggest that the other soft news questions
were not soft enough. Yet, if the effect of soft news is restricted to the most scandalous
of the soft news topics, it becomes hard to defend it on the grounds that it informs
people about politics.
While liking talk shows or infotainment programs affects soft news knowledge only
weakly at best, people who prefer traditional hard news formats are consistently better
informed about questions of sex, drugs, and missing interns. People’s rankings of newspapers and national TV news have positive and clearly significant effects for all four
soft news questions. This pattern continues in Table 7, which shows results for knowledge of the War on Terrorism. People who prefer to get their news from newspapers
or national TV news are consistently more knowledgeable about the war. The logit Any Good News in Soft News? 161 Table 5
Political knowledge questions
in boldface) Question wording
Soft News Knowledge
Who of these four politicians confessed to
having had an illegitimate child last year? %
• Ralph Nader
Rudy Giuliani 86.4 On what charges was President Bush’s
daughter Jenna arrested last year? •
• driving under the influence
possession of illegal drugs
indecent exposure 79.9 During the investigation of Congressman
Gary Condit’s affair with a missing intern,
which other woman claimed to have had
an affair with Condit? •
• a fashion model
a flight attendant
a clerk in Condit’s
Washington office 60.0 Do you happen to know who the person
shown by this photograph is?
(photo of Chandra Levy on screen) open-endeda 25.3 •
Kazakhstan 85.4 In the war in Afghanistan, which of the
following groups fought on the side of the
coalition led by the United States and Britain? •
• The Islamic Jihad
The Northern Alliance
Al-Qaeda 83.5 Which of the following agencies was founded
in the wake of September 11? •
• Office for Homeland Security
National Security Agency
Department of Civilian Defense 77.3 Knowledge of the “War on Terrorism”
Which of the following countries shares
a border with Afghanistan? Hard news knowledge
Which party currently has the most
members in the House of Representatives
In general, thinking about the political
parties in Washington, would you say that
Democrats are more conservative than
Republicans, or Republicans are more
conservative than Democrats? • Democrats
• Republicans 55.8 • Democrats
• Republicans 69.8 (Table continues on next page) 162 Markus Prior
Political knowledge questions (Continued) Question wording Response options
in boldface) %
correct For every dollar spent by the federal
government in Washington, how much of
each dollar do you think goes for foreign
aid to help other countries? open-endedb 20.1 Would you say there is more, less, or about
the same amount of crime in the United
States today as compared to 10 years ago? • more
• about the same 21.2 a Some people (7.3%) simply responded “yes” to this question. Because of an ambiguity in
the question wording—people were not explicitly asked to provide Levy’s name—it is unclear
whether “yes” answers may reflect knowledge of the name. To fix this problem, I impute “yes”
answers by using predicted values from a logistic regression of responses on all other knowledge
questions in the survey. The cut point is adjusted to maximize correct classification. This procedure introduces less bias than deleting the “yes” answers, or counting them as either correct or
Following Gilens (2001), I count answers of 5% or less as correct. estimates do not offer much support for any positive effect of soft news preference on
knowledge in the domain of war and foreign policy crises either. The only significant
soft news coefficient indicates that those who prefer talk shows are more likely to know
that Pakistan shares a border with Afghanistan. Matt Baum has convincingly documented
(for a far larger array of foreign policy topics than I examine here) that the entertainment value with which such crises are presented in soft news programs may get people
to follow news topics they would otherwise ignore. The analysis of knowledge about
domestic scandals and the War on Terrorism has shown that this attention does not
translate reliably into a learning effect. People who prefer soft news formats are significantly less knowledgeable than those who prefer hard news, even in the political domains most easily presented as entertainment.
After failing the weak knowledge criterion, it is no surprise that soft news preference does not have consistent effects on knowledge of hard news either. As Table 8
demonstrates, there are two notable exceptions, however. People who rank talk shows
highly among their favorite news formats are more likely to know that crime rates have
decreased over the past ten years and that Republicans are more conservative than Democrats.
Both findings are not completely counterintuitive. Crime is clearly one of the news
topics most easily and frequently featured in local and soft news (Gilliam & Iyengar,
2000; Hamilton, 1998).20 And to the extent that political talk shows are shouting matches
between Democrats and Republicans, they make it easy to figure out who is more conservative. Entertainment talk shows, however, usually present politics in less political,
less partisan terms (Baum, 2002a).21 All told, a limited positive effect of liking the talk
show format is evident. People’s ranking of infotainment, on the other hand, did not
produce any significant effects.22
Table 8 shows that people who like to get their news from magazines such as Time Any Good News in Soft News? 163 Table 6
The effect of preferred news format on soft news knowledge Jesse Jackson
child Jenna Bush
(openended) Ranking of talk shows .082
(.056) Ranking of infotainment .063
(.058) Ranking of newspapers .155**
(.053) Ranking of magazines .112
(.065) Ranking of national
TV news .325***
(.055) Ranking of local TV news .092
(.053) Ranking of Internet news .159*
(.058) Ranking of radio news .146**
(.054) Education .197**
(.047) Respondent is African
(.206) Respondent is not White
or African American –.169 )
(.226) Age .023
(.019) Age2 .000
(.000) Income .042*
(.015) Female .063
(.107) Respondent’s primary
language is English .853*
(.449) (Table continues on next page) 164 Markus Prior
The effect of preferred news format on soft news knowledge (Continued) Jesse Jackson
child Jenna Bush
(open–ended) Political interest .833**
(.210) Civic duty –.153
(.244) Internal efficacy .010
% correctly classified 87.6 80.9 65.7 76.4 N 2,248 2,248 2,248 2,246 Note. Cell entries are logistic regression coefficients and standard errors in parentheses.
*p < .05; **p < .01; ***p < .001. or Newsweek are consistently more knowledgeable than the average respondent on the
hard news questions. The effects of national TV news and newspapers are more mixed
than before, perhaps partly reflecting the difficulty of two of the hard news items (crime
and foreign aid; see Table 5). Interestingly, the effect of Internet news ranking is consistently positive and at least marginally significant for all items. The real significance of
the new media environment, this finding suggests, may not lie in the opportunities for
news hybrids such as soft news, but in media that give users the greatest control over
timing, pace, and content of their news.23
Table 8 also makes it clear that local news, not soft news, is the real villain of our
story. While a preference for soft news, and for infotainment programs in particular,
failed to produce more than occasional learning effects, the effect of liking local news is
actually negative for most of the hard news items. The first half of this article showed
that people turn to soft and local news because they watch television mostly for its
entertainment value. The major difference between the two formats is that local news is
much more popular than soft news. Because of this great popularity, the negative effect
of a preference for local news on hard news knowledge is a much more serious obstacle
to creating an informed electorate than the various soft news formats. Conclusion
If we accept the market logic—that media assess what people want and satisfy those
needs to maximize their profits—then the appearance of soft news programs is a predictable consequence of the proliferation of media outlets. Cable TV and Internet have Any Good News in Soft News? 165 Table 7
The effect of preferred news format on war knowledge
Ranking of talk shows
Ranking of infotainment
Ranking of newspapers
Ranking of magazines
Ranking of national
Ranking of local TV news –.004
Ranking of Internet news
Ranking of radio news
Respondent is African
Respondent is not White
or African American
language is English
% correctly classified
2,248 Which country
shares a border
with Afghanistan? Who fought
with U.S. in
Afghanistan war? .179** (.058)
.270*** (.052) .015
.318*** (.052) .038
–.886*** (.177) –.018
–1.295*** (.174) –.188 (.266) –.628* .005
(.813) .560* (.231)
2,248 (.249) 84.6
2,248 Note. Cell entries are logistic regression coefficients and standard errors in parentheses.
*p < .05; **p < .01; ***p < .001. multiplied the supply of media content. The result is greater diversity of content (see,
e.g., Steiner, 1952). The fact that more people watch soft news now than thirty years
ago is not surprising given that none of the channels offered soft news at the time. But it
must be mentioned in the same breath that more people watch CNN or C-SPAN today
than thirty years ago as well. In a similar vein, comparing the size of network news
audiences over time can tell us little about the changing demand for news in general. In
the seventies, they were the only show in town; now they have to compete with hundreds 166 Markus Prior
The effect of preferred news format on hard news knowledge
foreign aid Change in
ten years House
than Democrats Ranking of talk shows –.032
(.050) Ranking of infotainment .061
(.049) Ranking of newspapers –.014
(.045) Ranking of magazines .168**
(.069) Ranking of national
TV news –.011
(.046) Ranking of local TV news –.104*
(.048) Ranking of Internet news .163**
(.054) Ranking of radio news .042
(.047) Education –.003
(.048) Respondent is African
(.153) Respondent is not White
or African American –.118
(.210) Age –.027
(.019) Age2 .000
(.000) Income .023
(.014) Female –.300**
(.105) Any Good News in Soft News? 167 Table 8
The effect of preferred news format on hard news knowledge (Continued)
language is English –1.006***
(.280) Change in
ten years House
than Democrats –.248
(.323) Political interest –.139
(.196) Civic duty –.074
(.240) Internal efficacy .396*
(.172) Constant –.049
(.737) % correctly classified 80.7 78.7 56.4 63.6 N 2,248 2,248 2,248 2,248 Note. Cell entries are logistic regression coefficients and standard errors in parentheses.
*p < .05; **p < .01; ***p < .001. of cable channels and thousands of Web sites. It is a mistake to confuse the effect of
market differentiation (or fragmentation) with the “dumbing down” of audiences or media
content. Now that they can choose from a greater variety of news formats than before,
some people pick soft news. It is important to point out, as Matt Baum (2002b) does,
that not all of these people are former hard news users.
The data presented in this article show quite clearly that the demand for soft news
is limited. Most Americans still prefer hard news to soft news. And as far as the people
who are looking for entertainment are concerned, it is hard to see how too many of
them could get stuck with soft news. As Tom Patterson (2000, p. 15) notes, “In the long
run, entertainment programming is more entertaining than news for those who desire to
be entertained.” Few people want exactly the mix of news and entertainment that soft
news has to offer. When compared to the demand for hard news and pure entertainment,
demand for soft news appears to be quite small.
Matt Baum (2002b) has shown that heavy soft news viewers are just as likely as
hard news viewers to report that they follow news on foreign policy crises. I have
argued that this criterion is too weak, since people might say that they follow particular
stories but still not learn anything about the political matter at hand. For soft news to
contribute to the goal of an informed electorate, people must learn from the exposure.
Yet, those who clearly prefer soft news are not more knowledgeable than others about
typical soft news topics or other political issues. A preference for talk shows had significant positive effects (at p < .05) on 1 out of 4 soft news items, 1 out of 3 “War on 168 Markus Prior Terrorism” items, and 2 out of 4 hard news items. People who like infotainment programs are more knowledgeable than those who do not about only one item: Gary Condit’s
alleged affair with a flight attendant.24
This article has focused on the effects of soft news in one particular area, factual
knowledge. In some theories, however, high knowledge scores are not a prerequisite for
informed judgments. According to the online model of political information processing
(Lodge, Steenbergen, & Braun, 1995), people store periodically updated political evaluations in memory, but not the information used in making those evaluations. The notion
of low-information rationality (Popkin, 1994) suggests that heuristics can compensate
for lack of knowledge. Hence, even in the absence of consistent learning effects, soft
news may still affect other elements of political reasoning, most notably political attitudes and voting behavior (Baum, 2002a). While the positive consequences of soft news
for the political process remain to be demonstrated, it would be premature to conclude
that soft news is bad news. The benefits of a well-informed electorate, however, that so
many recent studies have documented (e.g., Althaus, 1998, 2001; Bartels, 1996; Delli
Carpini, 2000; Delli Carpini & Keeter, 1996; Gilens, 2001) are unlikely to emerge as a
result of greater soft news consumption. The diversification of news formats helps some
viewers find news programs they like. The hope that a more efficient market for news
might coincidentally promote democracy by simultaneously entertaining and informing
people does not receive much empirical support in this study. Notes
1. Soft news refers to both news stories and entire programs. Using Patterson’s definition,
any given news story can be classified as hard or soft, regardless of the news program in which it
appears. Alternatively, several new news formats, blends of information and entertainment, that have
emerged in the past two decades are labeled “soft news” and thereby contrasted to more traditional
hard news formats, most notably the national network news (e.g., Baum, 2002b). I use the second
defi-nition, as my focus here is on the impact of soft news programs on political knowledge.
2. For details on the sampling mechanism used by Knowledge Networks, see Krotki and
Dennis (2001); for a comparison to other Web survey methods, see Couper (2000); for the validity of Knowledge Networks data, see Krosnick and Chang (2001).
3. Question wording was as follows: “Which of the following types of news do you like
4. The choice of these two formats follows the two examples emphasized by Baum (2002b,
p. 93). Shows like Entertainment Tonight, Hardcopy, and Extra are also referred to as “tabloid
news shows” (e.g., Baum, 2002b, p. 93; Delli Carpini & Williams 2000, p. 171), but this label is
likely to worsen the problem of social desirability bias. “Infotainment shows” was selected as a
more neutral genre description.
5. These results are similar to those obtained by Patterson (2000, pp. 6–7) in a phone
survey that used different operationalizations of hard and soft news.
6. Ratings for network news magazines such as 60 Minutes or Dateline are higher than
ratings for the talk shows and entertainment news programs. In Baum’s (2002b, p. 93) example
week in 1999, 60 Minutes was viewed by a little over half as many viewers as the three network
news programs combined. These programs have become softer over time and now focus predominantly on soft news topics (Kalb, 1998; Zaller, 1999).
7. Limited by the Nielsen ratings data published at the time, Baum compares soft news
ratings for the first half of the year with hard news ratings for a week in the summer. Since few
people watch TV in the summer, this comparison is probably biased against hard news.
8. Even if ratings were available for all programs that we consider soft news, it would still
be impossible to determine the number of different soft news viewers per day. Any Good News in Soft News? 169 9. Audience sizes for particular cable news shows are Nielsen estimates published in Electronic Media, June 3, 2002, p. 30.
10. Numbers are from a CBS press release from August 1, 2002, based on Nielsen data.
According to CBS Research, a viewer has to watch a minimum of 6 minutes to be counted as a
unique viewer for a particular channel. The 2002–2003 season started in November 2002 and
should therefore be only mildly influenced by unusually high news viewing in the immediate
aftermath of 9/11 (Althaus, 2002). The slight increase of news audiences that persists in the
summer of 2002 (compared to previous years), on the other hand, may well be due to 9/11.
11. Specifically, respondents were asked what percentage of their viewing time was made
up of news, entertainment, sports, and “other programs.” If the entries did not add up to 100%,
respondents were asked to correct their answers. These questions were asked a few minutes after
respondents had completed the ranking of news formats. This was done to give people an idea of
the programs that could be defined as “news,” without constraining them by mentioning actual
exemplars in the question.
12. The numbers are generated by multiplying respondents’ total viewing time by the percentage of viewing devoted to news.
13. The correlation between political interest and news preference is r = .34. The advantage
of news preference is that it is an approximately continuous (21-point) scale.
14. Results are highly similar if only the first choice is used instead.
15. I am agnostic as to the process by which soft news could increase knowledge of hard
news. One possibility is that soft news programs do cover hard news, perhaps marginally or in a
“soft” way. Another, more interesting explanation would be that soft news programs increase
interest in “hard” politics and motivate viewers to watch hard news programs.
16. As Table 5 indicates, knowledge about the “War on Terrorism” was quite high, probably as a consequence of unusually intense media coverage and very high attentiveness (Althaus,
2002; Prior, 2002).
17. Rankings are scored 5 for first choice, 4 for second choice, 3 for third choice, 2 for
fourth choice, 1 for indifference, and 0 if the format was marked as disliked. I use preference
measures instead of exposure measures because it has been shown repeatedly that people’s selfreports of media use are extremely unreliable (e.g., Ansolabehere & Iyengar, 1998; Price & Zaller,
1993; Zaller, 1996) and biased (Ansolabehere, Iyengar, & Simon, 1999). These measurement
problems make is very difficult to estimate the effect of soft news exposure on political knowledge directly using survey data. While using preference measures as a proxy for exposure requires the assumption that people watch the types of news they like, it is an open empirical
question whether this assumption is more or less justified than the (false) assumption of accurate
self-reports. In Baum’s (2002b) theoretical model, people’s media consumption is derived from a
comparison of utilities, a process that assumes the existence of preferences as the basis for viewing decisions. Empirically, Baum’s soft news index combines self-reported exposure items and
measures of attention (to particular types of soft news topics). It is thus not clear whether the
positive effect of soft news on attention to foreign crises is conditional on a certain level of
attention to soft news in general.
18. Following Prince and Zaller’s (1993) approach, I also estimated the models with direct controls
for political knowledge. The coefficients for the news rankings changed marginally at the most.
19. The term “talk show” may be open to different interpretations. Respondents were asked
to rank “daytime and late-night talk shows (such as Letterman or Oprah).” Some respondents
may have included more “highbrow” talk shows such as Charlie Rose or the Sunday morning
interview shows. It thus remains somewhat unclear if the occasional positive effect of talk shows
is really an effect of soft news. (But recall from Figure 2 that people who prefer talk shows have
the lowest news preference of all groups.)
20. It is usually assumed, however, that the heavy focus on crime in local news leads
viewers to overestimate the amount of crime. The coefficient for local news reflects this expectation, even though it only approaches significance. High rankings of talk shows, in contrast, are
associated with a greater likelihood to know that crime rates decreased. 170 Markus Prior 21. It is plausible that respondents who are familiar with ideological terms are more likely
to prefer talk shows in the first place. Controlling for strength of ideology and/or partisanship did
not change the results, however.
22. Since all respondents were asked to rank-order the different news formats, those
who ranked soft news highly probably include people who would rather watch pure entertainment than soft news. One might therefore expect to find a stronger relationship between
ranking of soft news and knowledge for those respondents who report at least some news viewing. If anything, however, the effect is larger among those who report spending no time watching
news at all. Conditioning the effect of soft news ranking on news viewing, political interest, or
news preference did not yield statistically significant results for the majority of the knowledge
23. Statements about causality derived from cross-sectional analysis are often suspect. Plausibly, individuals who were more knowledgeable to begin with could have developed a preference for Internet news subsequently, indicating preference change rather than learning. This is
unlikely, however, because the knowledge items in the study refer mostly to relatively recent
events, and the effect of news did not change significantly when a measure of civics knowledge
(derived from questions about presidential veto, judicial review, and number of presidential terms)
was included in the model to control for higher initial knowledge.
24. Many scholars (e.g., Kalb, 1998; Patterson, 2000; Zaller, 1999) have argued that hard
news programs themselves have become softer in recent years. The present article focuses only
on the differences between hard news and soft news formats, and does not address the possibility
that proliferation of soft news stories in hard news programs affects political knowledge as well.
Moreover, hard news may enjoy the popularity advantage over soft news documented here in part
because it has already become softer. References
ADT Research. Cable news wars. Content analysis commissioned by the Newshour. (2002). Available
Althaus, S. L. (1998). Information effects in collective preferences. American Political Science
Review, 92, 545–558.
Althaus, S. L. (2001). Who’s voted in when the people tune out? Information effects in congressional elections. In R. P. Hart & D. Shaw (Ed.), Communication in U.S. Elections: New
Agendas (pp. 33–53). Lanham, MD: Rowman & Littlefield.
Althaus, S. L. (2002). American news consumption during times of national crisis. PS: Political
Science and Politics, 35, 517–521.
Ansolabehere, S, & Iyengar, S. (1998). Message forgotten: Misreporting in surveys and the bias
toward minimal effects. Unpublished manuscript.
Ansolabehere, S., Iyengar, S., & Simon, A. (1999). Replicating experiments using aggregate and
survey data: The case of negative advertising and yurnout. American Political Science Review, 93, 901–910.
Bartels, L. M. (1996). Uninformed votes: Information effects in presidential elections. American
Journal of Political Science, 40, 177–207.
Baum, M. A. (2002a). Making politics fun: What happens when presidential candidates hit the
talk show circuit? Paper presented at the 98th Annual Meeting of the American Political
Science Association, Boston, MA.
Baum, M. A. (2002b). Sex, lies, and war: How soft news brings foreign policy to the inattentive
public. American Political Science Review, 96, 91–110.
Couper, M. P. (2000). Web surveys: A review of issues and approaches. Public Opinion Quarterly, 64, 464–494.
Delli Carpini, M. X. (2000). In search of the informed citizen: What Americans know about
politics and why it matters. Communication Review, 4, 129–164. Any Good News in Soft News? 171 Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it
matters. New Haven, CT: Yale University Press.
Delli Carpini, M. X., & Williams, B. A. (2000). Let us infotain you: Politics in the new media
environment. In W. L. Bennett & R. M. Entman (Eds.), Mediated politics: Communication
in the future of democracy (pp. 160–181). Cambridge, England: Cambridge University Press.
Gilens, M. (2001). Political ignorance and collective policy preferences. American Political Science Review, 95, 379–396.
Gilliam, F. D., & Iyengar, S. (2000). Prime suspects: The influence of local television news on
the viewing public. American Journal of Political Science, 44, 560–573.
Hamilton, J. T. (1998). Channeling violence: The economic market for violent television programming. Princeton, NJ: Princeton University Press.
Kalb, M. (1998). The rise of the ‘new news.’ A case study of two root causes of the modern
scandal coverage (Discussion Paper D-34). Cambridge, MA: Joan Shorenstein Center, Harvard
Krosnick, J. A., & Chang, L. C. (2001). A comparison of the random digit dialing telephone
survey methodology with internet survey methodology as implemented by Knowledge Networks and Harris Interactive. Unpublished manuscript. Ohio State University.
Krotki, K., & Dennis, J. M. (2001). Probability-based survey research on the Internet. Paper
presented at the 53rd Conference of the International Statistical Institute, Seoul, South Korea.
Lodge, M., Steenbergen, M., & Braun, S. (1995). The responsive voter: Campaign information
and the dynamics of candidate evaluation. American Political Science Review, 89, 309–326.
Patterson, T. E. (2000). Doing well and doing good: How soft news and critical journalism are
shrinking the new audience and weakening democracy—And what news outlets can do about
it (Faculty Research Working Paper Series, RWP01-001). Cambridge, MA: John F. Kennedy
School of Government, Harvard University.
Popkin, S. L. (1994). The reasoning voter: Communication and persuasion in presidential campaigns (2nd ed.). Chicago: University of Chicago Press.
Price, V., & Zaller, J. (1993). Who gets the news? Alternative measures of news reception and
their implications for research. Public Opinion Quarterly, 57, 133–164.
Prior, M. (2002). Political knowledge after September 11. PS: Political Science and Politics, 35,
Steiner, P. O. (1952). Program patterns and preferences and the workability of competition in
radio broadcasting. Quarterly Journal of Economics, 66, 194–223.
Zaller, J. (1996). The myth of massive media impact revived. New support for a discredited idea.
In D. C. Mutz, P. M. Sniderman, & R. A. Brody (Eds.), Political persuasion and attitude
change (pp. 17–78). Ann Arbor: University of Michigan Press.
Zaller, J. (1999). Market competition and news quality. Paper presented at the 95th Annual Meeting of the American Political Science Association, Atlanta, GA. ...
View Full Document
This note was uploaded on 07/21/2011 for the course BUS 10001 taught by Professor All during the Spring '11 term at Shaheed Zulfiqar Ali Bhutto Institute of Science and Technology.
- Spring '11
- The Land