If the news is so bad, why are presidential polls so high? Presidents, the news media, and the mass public in an era of new media.

Cohen, Jeffrey E.

By almost any standard, 1998 was a horrible year for any president. Bill Clinton's affair with Monica Lewinsky became public, leading to his impeachment. The Republican-controlled Congress heartily attacked him, and the news media, never easy on the administration (Kurtz 1998), escalated the degree to which it challenged and criticized the president. Figure 1 traces the percentage of news stories about the president and the administration from 1981 that were coded clearly negative or more negative than positive. (1) As the figure demonstrates, 1998 stood out in the degree of negative news reports. Only 1987, the year of the Iran-Contra scandal, produced a higher percentage of negative news stories on the president. (2) Even 1994, the year of Clinton's ill-fated health care initiative, itself a bad press year for Clinton at 58 percent, is still less negative than 1998 by nearly 10 percent. Not surprisingly, nearly 22 percent of all news stories in 1998 about the president and administration focused on the scandal and nearly all of those stories were negative. Without the scandal stories, only 47 percent of the news stories about Clinton would have been negative, which is similar to the amount of negative news that he received in any given year.

[FIGURE 1 OMITTED]

What is so remarkable about these figures is not that Clinton received so much bad press, but that his job approval polls rose that year. (3) Fifty-six percent of those polled in the last Gallup poll of 1997 approved of the job that Clinton was doing. By the end of July 1998, his job approval rating had risen to 65 percent. It remained at about that level throughout the remainder of the year, before spiking upward in very late 1998 and early 1999, reaching a peak of 73 percent in Gallup's poll of December 19-20, 1998.

That Clinton's polls did not plummet in the face of such bad press challenges many widely held assumptions about the role of news in shaping opinions. Brody's (1991) seminal book argues that the balance of positive and negative news about the president will affect public attitudes toward the president. When the news leans in a negative direction, presidential approval should drop (also Erikson, MacKuen, and Stimson 2002). Brody's perspective suggests that Clinton's approval should have declined in 1998, yet it rose! Our existing theories and understandings of the relationship among the news, the presidency, and public opinion cannot explain why Bill Clinton's approval rose in 1998.

Nor do our exiting theories explain the general decoupling of news from presidential approval. Figure 2 plots data on the annual tone of presidential news and Gallup approval from 1949 through 1992 (Ragsdale 1997). The two series diverge, as expected, until about the mid-1970s, and then begin to track together. Regression analysis reveals that the two series are negatively related (b = -.40; se = .16; t = 2.52; p = .02). However, if we bisect the series into subsets (1949-1976 and 1977-1992), we find the expected negative relationship for the 1949-1976 segment (b = -.68; p = .000) but no relationship for the 1977-1992 segment (b = .31; p = .28). In fact, if we drop 1987, the year of Iran-Contra, from the 1977-1992 segment, we find a strong positive relationship between negative news and presidential popularity (b = .90; p = .000)! In other words, bad news leads to higher presidential polls, barring major negative events, like Iran-Contra. I do not want to make such a claim, but obviously, our traditional understanding of the relationship between news and public regard for the president needs rethinking.

[FIGURE 2 OMITTED]

The presidential news subsystem, the set of interrelationships among the president, the news media, and the public, underwent a major transformation over the past 20 to 25 years. In this article, I identify the major aspects of this transformation and discuss the implications of this transformation for politics and presidential leadership in the late twentieth and early twenty-first century. Four changes are especially important. Briefly, they are, first, the news media have become increasingly competitive and decentralized. Second, reporting styles have changed as stories became softer and increasingly negative. Third, the public's consumption of news from traditional outlets has declined. Fourth, public regard toward the news media has plummeted. In the next section, I describe the presidential news subsystem during television's "golden age" of the 1960s and 1970s (Baum and Kernell 1999; Cook and Ragsdale 1998). Then I trace the four trends. I conclude by discussing the implications of these trends on modern American politics and presidential leadership and return to the paradox of the rise in Clinton's polls in 1998 despite a heavy barrage of bad news about the president.

The Presidential News System in the Golden Age of Television

Baum and Kernell (1999) colorfully call the period of the late 1950s through 1970s the "golden age of presidential television." In that era, the president enjoyed a number of major advantages. But that system also posed grave threats to presidential leadership. During the golden era, the public received the bulk of its news from the three nightly news broadcasts and the audience for the nightly broadcasts was large (Baum and Kernell 1999). Through television, presidents had easy access to the mass public, a major advantage for presidents bent on leading the public.

News content during this era also advantaged the president. The president was the dominant news story of the age, crowding rivals and other political leaders off of the news hole (Gilbert 1981, 1989). Additionally, news was reported objectively and tended to portray the president in a positive light (Grossman and Kumar 1981). While the seeds of interpretive and cynical news reporting were sown during this era, such a style was not yet the norm. Still, this news system could pose a danger to a president if it turned against him, as it occasionally did. Two administrations, Johnson's and Nixon's, were cut short in part because the press turned against the president.

The Public and Television News

In the 1950s, television emerged as the most popular entertainment and news medium for American citizens. For instance, in 1952, 79 percent of American National Election Studies respondents claimed to have read something about the presidential campaign in newspapers, where only 51 percent said that they watched a television program about the campaign. (4) By 1956 the pattern had shifted. Newspaper reading declined to 68 percent, while television watching rose to 73 percent. From then on, more people would claim to have received news about the campaign from television than from newspapers (see Figure 3).

[FIGURE 3 OMITTED]

Moreover, the percentage of people who only relied on television for news increased from about 8 percent in 1952 to 18 percent in 1956, ranging between 14 and 19 percent through 1968 (Figure 4). From one sixth to one fifth of the public claimed that television was its sole source of news. At the same time, the percentage of the public that only used newspapers dwindled precipitously from over 35 percent in 1952 to 12 percent four years later. Thereafter it never rose above 10 percent. As Figure 4 illustrates, the public had bifurcated into two groups by the late 1950s, those who only watched television for news about the campaign and those who used both television and newspapers.

[FIGURE 4 OMITTED]

Presidents and Reporters in the Golden Age

The three broadcast networks dominated the dissemination of both entertainment and news programming to the public by the late 1950s. It was not until John E Kennedy became president in 1961, however, that the presidency understood and attempted to harness the power and reach of television. While both Truman and Eisenhower occasionally appeared on television, they focused the bulk of their attention on print journalists. Broadcast journalists were decidedly second-class citizens among reporters. The status differences between print and broadcast reporters narrowed with the arrival of the Kennedy administration. Kennedy viewed the new medium as one to exploit rather than one to fear. It offered him access to the mass public unlike any previous means of communication. He innovated in its use, appearing to the public directly, over the heads of the news media, but also made it easier for broadcast journalists to cover him.

Unlike his predecessors, Kennedy offered live press conferences in prime time, a departure from past press conference practice. Over the course of the twentieth century, journalists had come to expect press conferences. Through press conferences, journalists received content for their news stories, which made them into key transmitters of information from government to the mass public. Kennedy's televised press conferences, however, altered the press-public connection. If people could watch the press conference, they no longer needed the newspaper story to fill them in on what transpired. Kennedy and subsequent White Houses began to favor the electronic over the print media, figuring that it could transmit its message directly to the public (Maltese 1992). They also figured that television would be satisfied with film showing the president and in this way would become relatively passive transmitters of news that presidents manufactured. Throughout the 1960s, because of the public's appetite for television and the growing stature of television journalists, the system worked to the advantage of the president and television journalists.

News Reporting Styles in the Golden Age

During the golden age, the president dominated the news. News tended to be reported objectively, although early signs of a new interpretive news style were emerging. The news also tended to portray the president in a positive light. The combination of these news-reporting attributes conferred advantages on the president, enhancing his ability to lead.

From the late 1800s on, the presidential image in the news had been expanding. Long before television was invented, presidents were beginning to receive more news than Congress, something unheard of previously, except during presidential election years (Balutis 1976, 1977; Cornwell 1959; Kernell and Jacobson 1987). Television enhanced this trend, garnering the president an even larger share of the news.

Measuring the quantity of news is complicated because of the numerous news outlets and the sheer volume of news. Consequently, most studies of presidential news look at only the number of presidential news stories and perhaps collect similar figures for news about Congress. But rarely do such studies attempt to calculate the percentage of news about the president as a proportion of the total amount of news. Frank Baumgartner and Bryan Jones's agenda project provides one data source that can give us a glimpse at the comparative volume of presidential news.

Their agenda's project randomly sampled news stories from the index of the New York Times from 1946 to the mid-1990s. In Figure 5, I trace two trend lines using these data: presidential mentions as a percentage of all news stories and presidential mentions as a percentage of news stories about government. Both trend lines display an unmistakable and similar pattern, a growth of presidential news from the 1940s into the 1970s. The two are highly correlated in spite of their differing bases (Pearson's r = .87; p = .000). From 1946 to 1959, presidents received on average about 5 percent of all news and 12 percent of government/policy news. These figures jumped to 8 and 17 percent in the 1960s, respectively, growth rates of 60 and 42 percent. The growth in presidential news continued in the 1970s, with the president receiving 11 percent of all news and 23 percent of governmental/policy news during that decade.

[FIGURE 5 OMITTED]

Comparable data do not exist on television news coverage of the president, but indications exist that the president was an even more pronounced figure there than in newspapers. Still, these newspaper data probably reflect a similar trend in television news. Feeling competition from television, major national newspapers would likely follow suit and increase their coverage of the president. Plus, through interaction on the beat, training, and other processes, journalists tend to develop a consensual understanding of what is news. The president was the star news attraction across all news media during the golden age of the 1960s and 1970s.

News reporting on the president also tended to be objective and positive. The best data on these attributes come from Grossman and Kumar (1981). They ambitiously content analyzed news stories about the president from three news sources, Time magazine, the New York Times, and CBS News. The Time and the Times data span from 1953 through August 1978. The CBS data begin with August 1968, when the Vanderbilt Television Archives began collecting tapes of the broadcasts (p. 254).

All three news organizations reported more favorable than unfavorable news about the president. Approximately 60 percent of Time's stories were favorable, with about 11.8 percent neutral and 28.2 percent negative. The New York Times's breakdown is similar, with 48.7 percent positive, 24.1 percent neutral, and 27.2 percent negative. CBS, which spans a shorter time frame, displays a balance between positive (38.5 percent) and negative (38.6 percent) news, with 22.9 percent neutral.

The shorter data-gathering period for CBS allows Watergate to weigh heavily in CBS's totals, which accounts for the difference between CBS's tonality in news about the president and the two print outlets. From 1953 to 1965, both Time and the Times gave the president a ratio of favorable to unfavorable news of approximately 5 to 1. From 1966 to 1974, news from both print outlets was more negative than positive, as was the case for CBS. In the post-Watergate years, 1975-1978, news for all three news organizations shifted, with positive news again outweighing negative news, often by hefty margins (see Table 1). What is so remarkable is that even during the Watergate years, the news was not extraordinarily negative. Thus, other than when extraordinary events led the press to view the president in a negative light, presidents during the golden age could count on favorable news.

Summary

The presidential news subsystem of the golden age, roughly the 1960s and 1970s, was one in which people watched television for news and three major networks were their major news sources. The president was the dominant news story across all forms of news media, with non-television news copying television's emphasis on the president in order to compete with television. Furthermore, the news, except in extraordinary times, was generally positive toward the president. The large television news audience, the amount of presidential coverage, and its positive tone all should have enhanced the ability of the president to lead the public.

The great danger for the president occurred when events turned sour and the press began portraying the president in a negative light. In an age when negative news about the president was uncommon, a surge in bad press came as a shock and often would lead people to rethink their approval and support of the president. The centralization of the news delivery system, in which news providers offered the public essentially the same news, reinforced the opinion-shaping effects of news. In other words, the system of news delivery during this era was essentially a one-message system, to use Zaller's (1992) term. The public received essentially one message about the president no matter the news source. Two presidents, Johnson and Nixon, were arguably driven from office in part because of this dynamic. In the main, however, this news system worked to the president's advantage because the news tended to be favorably disposed toward him.

All this would begin to change in the late 1970s. The news media decentralized. Competition and other forces led to a new style of news reporting that was more interpretive and negative than the style of the golden age. The audience for news also shrank and people, perhaps because of the new style of news, became more cynical toward the news media itself, which likely muted the impact of news on public opinion (Patterson 2000). Together, these trends blunted the impact of the news on public attitudes toward the president. They also lessened the ability of the president to lead the public.

The Presidential News Subsystem in the New Media Age

Competition among the news media, the rise of soft news, increasing negativity in hard news, a shrinking news audience, and declining public regard for the news media characterize the presidential news subsystem in the new media age (Baum 2003; Davis and Owen 1998). Such a news subsystem ironically limits the damage that the news media can do to a president, while also limiting the president's ability to lead the public. In response to the changes in the structure of the presidential news subsystem, presidents have changed their style of "going public." Instead of focusing attention on leading the broad mass public, much presidential activity now targets select constituencies, which are often already presidential allies.

The Decline of Network Monopoly and the Rise of Competition

In the golden age, three national networks and a handful of other national news organizations dominated the production, definition, and dissemination of the news. All news outlets offered essentially the same basic portrait of the president, except that television could not portray the president in as much depth as the print media. Technological and economic forces came together in the late 1970s and 1980s to crack the control that these elite news organizations had over the news (Hamilton 2003). By the 1990s, the news production system had decentralized and splintered. Many news organizations competed for a shrinking news audience: To create a market niche, news programs and producers tried to differentiate themselves from their competitors by presenting distinctive voices and perspectives to the news)

Cable television, the Internet, new printing technologies, handheld cameras, and satellite systems all worked together to break the monopoly of the elite press of the golden age. Lightweight, handheld cameras plus satellite technology allowed local broadcast stations to produce their own news from almost anyplace. Although they tended to center their activities within their localities, some local broadcasters would send production teams to Washington to follow events of local interest in the nation's capitol, as well as to give a local spin to a national story. Similarly, new printing technologies, desktop publishing, and computer-assisted customer lists allowed smaller, specialized magazines, newspapers, and so on to proliferate and prosper during this decade.

Cable television had perhaps the greatest impact on the presidential news subsystem. It spawned dedicated news networks, such as CNN, Fox, and MSNBC, offering news around the clock. No longer did news-hungry viewers have to wait until the nightly broadcast to learn of events. Each cable network also tried to differentiate itself. CNN prided itself on international coverage. Fox offers viewers news with a conservative tilt, or so it proclaims. All have increased the amount of political commentary, with pro grams such as "Crossfire" that mix theater and political opinion. Importantly, the big three networks now faced competition from the 24/7 cable news outlets.

Cable had another, perhaps more profound effect. Cable programming offered viewers a choice of shows to watch beside what the three major networks offered. By and large, the public of the golden age had little programming choice. In the early evening, all three networks broadcast their national nightly news programs, sandwiched between local news broadcasts. If one was to watch television during these hours, one had to tune into a news program. The structure of television during the golden age effectively captured the public. People without an interest in news had little choice bur watch such programs, unless they decided to turn off the television.

With cable, people's tastes were better served. Reluctant news viewers had a myriad of entertainment offerings to watch in lieu of the network nightly news (and network entertainment fare). Across the board, the ratings of network programs, news and entertainment alike, eroded. VCRs allowed people to tape a show and watch it when they wanted, as well as rent or buy a copy of a film to watch whenever they felt like it. It is highly unlikely that many people taped news programs for later viewing. The Internet may similarly peel away the television audience, as some early data indicate it is doing. All traditional news organizations, from television stations to newspapers and news magazines, have a presence on the Internet. The news hungry may peruse the web pages of these established news media outlets, which some seem to be doing.

Less traditional news providers are also springing up on the Internet, challenging the dominance of the traditional media, in the process redefining news and affecting public regard for news organizations. The prime example of this phenomenon is Matt Drudge, who came to national prominence in late 1997, when he published on his web page the allegations that President Clinton was having an affair with Monica Lewinsky. Newsweek purportedly also had the story, but refused to publish because the news magazine could not find a second source to confirm it. Drudge turned the hand of the traditional news media, much as print tabloids had been doing for the past 10 to 15 years, forcing the news media en masse to cover the scandal.

The Rise of Negative News and the Decline of Political News

The decentralization of the news media and the ability of almost anyone to become a "journalist" and publish a Web page (e.g., Matt Drudge) threatened the traditional norms of news publishing, such as source protection and confirmation of information, norms that had evolved over the course of the twentieth century. Whereas news in the golden age was primarily objective and positive toward the president, news in the age of new media became softer and more sensational (Patterson 2000; Sabato 1991). The boundary between entertainment and news blurred, and journalist voices began to appear in the news in greater quantity than the voices of politicians. Here I will focus on two trends with special relevance to the president and his relationship with the mass public: the decline in the volume of news and the rise of negativity in news reporting.

Many commentators have noted the decline of traditional hard news, such as reporting on government and public policy, as crime stories, entertainment and celebrity profiles, and personal health and so on replaced traditional hard news (Patterson 2000; Hess 2000). Many factors are alleged to have stimulated these shifts, including the rise of market-driven journalism, greater competition among news organizations, and the rise of cable television. Here we need not detain ourselves with the causes of these trends, however interesting they may be. Instead, I just want to try to review some relevant evidence on this trend.

Declining Volume of Hard News. The Project for Excellence in Journalism conducted a study that content analyzed the news of seven major media outlets (the three networks, Time, Newsweek, the New York Times, and the Los Angeles Times) for three time points, 1977, 1987, and 1997. These data reveal a dramatic decline in traditional hard news across almost all media (see Table 2). The Project coded entire network broadcasts, the front pages of the newspapers, and cover stories of the news weeklies. (6)

Here I use the Project's definition of traditional hard news--news about the government, military, domestic policy, and foreign policy. Their other subject categories are entertainment/celebrities, lifestyle, celebrity crime, personal health, other crime, business and commerce, science, technology, arts, religion, sports, weather and disasters, and science fiction and the supernatural. As Table 2 displays, only the major newspapers seemed to have resisted the trend of replacing government- and policy-related news with other types of content. The decline among the three networks is striking, from about two thirds or more of broadcast new stories dealing with government and policy in 1977 to about 40 percent in 1997. Similarly, the weekly news magazines Time and Newsweek, which in the 1970s ran government- and policy-related figures and stories on about half of their covers, ran such stories on their covers only about 20 to 25 percent of the time in 1997, a reduction rate of approximately 50 percent. The Los Angeles Times displays a minor drop in government and policy news on its front page, from about three quarters to two thirds of stores. Still, the bulk of the Los Angeles Times's front page is given to hard news. Only the New York Times seems to have resisted the trend of declining news coverage on its front page, with about two thirds of front-page stories given to the traditional news items for all three time points.

The Project's data are limited in several regards. They provide us with only three time points. They code only one month per year. Either of these data collection decisions may have skewed results. For the news magazines, exclusive attention to the cover story may overstate the decline in news coverage between the covers, as covers are used primarily to attract newsstand purchasers.

The Baumgartner and Jones Agendas Project data give us another vantage point to look at trends in news coverage in the New York Times. Figure 5 presents the percentage of presidential news as a fraction of all news and governmental news. Unlike the Project for Excellence data, these data span a much longer time flame (1946-1994), but also contain more than just front-page stories. Unfortunately, the sampling system of Baumgartner and Jones just does not provide enough front-page stories to render reliable temporal comparisons.

In these data, we see the rise of presidential news through the mid- to late 1970s, followed by a steady decline thereafter, which comports well with Project for Excellence data. Thus, while the New York Times, according to Project data, may have resisted the pressures to replace "hard" news on its front page, the Times may not have been so able to resist such trends in the rest of the publication. Perhaps the most important point from these data is that from several different vantage pints, coverage of traditional government and policy news has declined across a variety of news media. Moreover, the networks, still the medium with the largest news audience, show the steepest decline in government and policy news.

Increasing Negativity in Presidential News. At the same time that the volume of governmental news, much of it about the presidency, has been ebbing, the lion's share of presidential news is no longer positive (Groeling and Kernell 1998). Again, the Project for Excellence provides some useful data. They compared nightly network news coverage of Presidents Bill Clinton and George W. Bush during their first 100 days. (7) In his first 100 days, Bill Clinton received positive coverage 22 percent of the time, compared to 28 percent negative and 49 percent neutral coverage. George W. Bush news was similar: 27 percent positive, 28 percent negative, and 44 percent neutral. While not extraordinarily negative, that the figures are not lopsidedly positive is notable given that this is the president's traditional honeymoon period. If the honeymoon represents a time period when presidents can expect their best overall news coverage, these figures do not bode well for the rest of the administration's time in office. These numbers differ markedly from the ones reported above that Grossman and Kumar collected for the 1953 to 1977 period, and they differ dramatically from the high degree of positive news that presidents received on average in the 1950s and early 1960s. Yet it is hard to simply compare the Grossman-Kumar data with the Project for Excellence data.

Fortunately, Lyn Ragsdale has taken Grossman and Kumar's New York Times data and extended the series forward to 1949 and through 1992. Her efforts provide us with a long annual time series, which I have displayed in Figure 2. The figure plots the percentage of negative news, which shows a clear upward trend. (8) From 1949 through 1959, about 12 percent of presidential news was negative. In the 1960s, this percentage rose to 17.5, climbing to nearly 32 percent in the 1970s and 28 percent in the 1980s. Obviously, Watergate had a huge impact on the average amount of negative news in the 1970s, with over half of news about the president negative in 1973 and 1974. If we remove 1973 and 1974 from the average of the 1970s, we find about 26 percent of presidential news being negative in the remaining years of that decade. In the 1980s and 1990s, presidents could expect news to be more negative than it was in the 1960s and 1970s, and on average presidential news in the 1980s and 1990s tended to be more negative than positive. (9)

The Declining Audience for News

At the same time that the amount of political and presidential news declined and negative news rose, the news audience was shrinking. Commentators offer several explanations for the declining size of the news audience. Patterson (2000) attributes the decline to changing reporting styles. The rise of sensationalistic and soft news, according to Patterson, alienated some of the news audience, especially those who preferred hard news. Baum and Kernell (1999) point to the effects of changes in the structure of the mass media. The arrival of cable television offered viewing choices besides the network nightly news. Instead of watching the network broadcasts, television viewers with cable access could watch entertainment, sports, and other non-news flare. Indications are that large numbers of viewers left the networks for cable offerings, as well as other media, such as VCRs and the Internet.

Whatever the source of the decline in the audience for news, we see it in some of the data presented earlier in Figure 4. Figure 6 presents the average number of days that respondents used different news media from the mid-1980s forward, when ANES began collecting such data. All news media show declines in average usage from the 1980s to the 1990s. In the late 1980s and early 1990s, people on average watched the nightly network news broadcast from 4.6 to 5 days a week. Their news viewing dropped to less than 4 days a week from 1996 through 2000. On average, people now claim to watch the nightly network news about one day a week less than they did a decade ago. Coupling this decline with the decline in hard news content may lead to the conclusion that the public overall, which relies mainly on these broadcast for news, is much less informed about public affairs than it was two decades ago.

[FIGURE 6 OMITTED]

Newspaper consumption has also declined, although not as steeply. On average, people claimed to read a daily newspaper nearly 4 days a week from 1984 through 1994. From 1996 through 2000, they decreased their newspaper reading about one half a day per week, to about 3.5 days per week or about every other day, although we cannot say whether the mix of stories that they read during the time frame changed or not. Given that the Baumgartner and Jones New York Times data and the Project for Excellence data on the Los Angeles Times front page suggest a decline of hard news as a percentage of overall newspaper content across the last two decades or so, it would not be surprising if newspaper readers also were Less well informed than two decades ago, given these changes in news content and reader habits.

The ANES data are displayed in a different way in Figure 7. Here the number of people who are heavy media users, that is, those whose combined use of television news and newspapers totals at least 11 days per week over the past week, (10) declined as a percentage of the population from about 35 to 40 percent in the 1980s and early 1990s to 20 to 25 percent in the late 1990s (1996-2000). At the same time, light media users, those whose combined consumption totals no more than 4 days a week, increased as a percentage of the population from 15 to 20 percent across the 1980s and early 1990s to 30 to 35 percent in the second half of the 1990s.

[FIGURE 7 OMITTED]

Other ANES data, which can be traced back to the 1950s, suggest a long-term decline in news consumption. Figure 3, presented above, plots the percentage of the public who claim to use newspapers and television for news about the presidential election campaigns. The utility of this measure is its series length. The deficit of this measure is that it does not seem to discriminate very well between heavy and light media users. As Figure 3 shows, and as reported above, television viewing increased in the 1950s, peaking at 80 to 90 percent by the 1960s, where it settled until the early 1970s. A small decline is noticeable in the mid-1970s, but a break in the series in 1988 makes it hard to specify the trend's shape in the mid-1980s. The mid-1990s' drop is easily seen in these data. The newspaper trend suggests a decline from the late 1950s onward, with of course some upward and downward spikes along the way.

General Social Survey (GSS) data, which use a more discriminating question by asking people to classify how many days a week they use the newspaper, are plotted in Figure 8. The classification scheme here and resulting figure require some explanation. The GSS has asked people since 1972 how many days a week they read a daily newspaper: every day, a few times a week, once a week, less than once a week, and never. I scored these categories 5, 4, 3, 2, and 1, respectively. The figure plots the averages of these individual scores, which should not be read as literal averages but category averages. The trend line shows a decline in which the average category was between every day and several times a week of newspaper reading in the 1970s. A steady state appears in the 1980s, when people seem to read the newspaper on average several times a week. In the 1990s, their newspaper-reading habits deteriorate even more to between once a week and several times a week. Like the ANES data, the GSS data indicate a steep step-like decline in the mid- to late 1990s.

[FIGURE 8 OMITTED]

Overall, these data indicate a shrinking audience for news for both television and newspapers. Moreover, not only is there a long-term deterioration in news consumption that begins for newspapers in the 1960s and television in the 1970s, but a steeper decline set in sometime in the mid-1990s. The 1990s decline appears not to be just a short-term trough, but a step-like drop.

The combination of the decline in news content with the shrinking news audience points to the conclusion that the public now is less well informed than was the case in the 1950s and 1960s. Whether the increased amount of negative news can influence public opinion about the president, considering the small audience for news, is another question. Whether negative news about public figures and government can have much impact on public opinion when public confidence in that institution has also declined raises still another question.

Declining Trust in News Media

A fourth trend of significance is the declining public confidence in the news media. Considerable debate exists over whether declining confidence in the news media is just a function of overall confidence declines in all major institutions (Bennett, Rhine, and Flickinger 1998) or whether some aspect of the decline is particular to the news media itself (Cook and Gronke 2001).

The GSS has been asking people about their confidence in the people running television and press off and on since 1972. Although the press question clearly points to the news media, the television question is more ambiguous. Many people may think of television entertainment executives as well as news executives. But as the data in Figure 9 indicate, both series display a downward trend of similar magnitude from the 1970s through the 1990s. Linear trend lines have been added to the figure to highlight the downward trend and make the series, which are often broken due to missing data, more easily interpretable. Patterson (2000) argues that the content and style of modern journalism has alienated many viewers and readers (also see Cappella and Jamieson 1997). One casual path linking news styles, confidence, and consumption is that the modern news style erodes confidence and trust in the news, which in turn leads people to abandon the news.

[FIGURE 9 OMITTED]

Summary

In the above pages, I have documented major trends in the style of news and the audience for news. Since the 1970s, the heyday of the golden age of television and news, the amount of the news hole devoted to traditional news about government, public policy, and the presidency has plummeted. At the same time, presidential news has turned increasingly negative. Now the news is more likely to be negative than positive toward the presidency. While these news content trends have been taking place, public habits and attitudes toward the news have also been changing. The public consumes less news than it did two decades ago and it has less confidence and trust in the news media than it once did. What are the implications of these changes for the press, the presidency, and public opinion?

Implications for Presidents, the News Media, and the Public

Based on existing understandings of the impact of news on public opinion (e.g., Brody 1991), one would expect the increase in negative news to have a dampening effect on public support for the president. Not only does the case of Bill Clinton during 1998 belie this expectation, but the positive correlation between negative news and presidential approval since the late 1970s also confounds traditional understandings of the connections between news and public support for the president.

Several factors about the new news media may blunt its impact on public opinion. First, it is harder for the news media to reach and consequently affect public thinking because the news audience has shrunk. Fewer people are attentive to the news than was once the case. Furthermore, there is less news content for them to be attentive to. While it is true that people may encounter the news indirectly, through conversations with friends and family, such indirect, or two-stage flows, of information may weaken the impact of the news, as content is filtered in the course of being communicated this way. Moreover, some people's social networks may be nearly apolitical and lacking much news content, due to the degree of shrinkage in the news audience. Such a social network context will further blunt the impact of news on public opinion.

The rise of public discontent with the news media may blunt the impact of news reporting on the public even more than the shrinking news audience. Source credibility is an important ingredient in opening a person to communication effects. As source credibility declines, communication effects should also diminish. If the increasing levels of distrust and lack of confidence in the media measure the credibility of the press, then we can conjecture that the news has less impact on the public because the source that the information came from has lost some of its credibility.

Closely related, the fact that the news now is regularly negative may also undermine the ability of news content to affect public support for the president. In the golden age of broadcast television, negative news about the president was potent because it was rare. Its rarity signaled to the public that something was truly amiss in the White House. When all news about the president is negative, the signaling value of news to the public declines. The public can no longer tell whether there are problems with the president and the administration that it needs to consider or whether the negative news is just the same old story that the news media always seem to be reporting. When the signal from the news media is so noisy, the public discounts it heavily.

The combination of a smaller news audience, the loss of credibility of the news media, and the noisiness of the news signal to the public has undermined the ability of the media to affect public attitudes about the president compared to the impact of the news during the golden age. If my argument makes sense, there are even more profound implications. In a mass mediated age, the public relies heavily on the news media to act as its eyes and ears about government. The information that the public receives from the news media is indispensable in the public's ability to hold its leaders accountable. As such, the news media play a vital role in democratic processes. When the public pays little attention to the news, when it views the news as incredible, and the news as noise rather than information, then this linkage that binds the governed and the governors together is weakened.

This new system also has important consequences for the president and his ability to lead the public. One may read from the above comments that the president may be somewhat immunized from bad news. This is one way to read the Clinton example. But this new system is a double-edged sword for the president too. It also limits his ability to lead the public (DiClerico 1993). Rather than trying to build widespread support across a national mass public, in this new environment the president must build support a different way.

A smaller news audience means that the president can reach fewer people through the news media than he once could. Moreover, his ability to go public directly is undermined, as the audience for presidential addresses has declined (Baum and Kernell 1999; Welch 2000). As Edwards (2003) argues, one of the barriers to presidential leadership is gaining public attention. This barrier has always been in place, given the relatively low levels of public awareness and interest that polls have noted across the last 50 years. But the news system describe here has raised the height of that barrier. First, presidents have increasing difficulty even getting the news media to pay attention to them, as the declines in news content indicate. And even when presidents make the front page and nightly news broadcast, stories are more likely to be about scandals and less likely to be about policy making than was the case several decades ago. Moreover, when the president wants to address the nation in prime time, the networks have been increasingly likely to deny his request, and sometimes when access is granted, only one of the networks will broadcast the presidential address (Foote 1980). The lack of news attention by average citizens adds to the barrier between the president and the public.

In place of building public support through appealing to the broad mass public, the president engages in a more selective approach, targeting specific groups. We see this in the increase of presidential speaking, but not to the nation as a whole. Usually presidents target friends, ginning up their enthusiasm for the president. Presidential opponents counter with appeals to opposition groups.

Neither the president nor his opponents have much incentive to moderate their rhetoric or policy proposals in such a system. The moderate middle of the public is effectively left out of the picture. Instead, presidents and political elites in general try to mobilize already committed and loyal constituencies, composed of people whose political beliefs tend to veer far from the middle. This system of leadership and opposition further polarizes already polarized politics.

Thus, while the new system may immunize presidents to some degree from negativity in the news, this new system also limits opportunities for presidential leadership. Great events, like 9/11, that galvanize the public must exist for presidents to lead the nation. Rather than being the nation's leader, with a one national constituency, presidents in this new system act more as the leader of many constituencies.

Democracy suffers in such a system. Presidents are to a degree uncoupled from the mass public, but are tethered to special interests. The news media seem unable to act as a true watchdog for the reasons expressed above. The linkages once meant to bind governed and governors have come undone. And of course, the public feels increasingly dispossessed. The irony in all of this is that presidents may actually enjoy solid popularity levels. We may characterize this new system as popularity without responsibility or accountability.

A Coda: How Did Clinton Survive 1998?

I opened this paper with the puzzle of how Bill Clinton's popularity could have risen while being pummeled with so much bad news emanating from a hostile press. This is only puzzling if we look at politics in the late 1990s as structurally similar to politics 25 years ago. It is not the same structurally.

Bill Clinton survived 1998--a scandal with a young White House intern and a congressional impeachment--as well as uniformly bad press, because (1) much of the public was not paying much attention to what was going on, and (2) much of the public discounted what the press had to say about the president because the public had gotten used to the press knocking presidents. Bill Clinton's polls rose in part because of a good economy and in part because of a counter-reaction of the public to the extreme negativity of the press (Popkin 1998).

But it is also likely that Bill Clinton's public relations strategy, built on a system fully expressed by Ronald Reagan, had a role to play in explaining Clinton's rising polls. If the public refused to pay much attention to what the press said about the president, the public may have paid some attention to what the president had to say about himself. Clinton's public relations strategy was to emphasize his leadership and the effectiveness of his policies.

Unwittingly, the press may have enabled Clinton's strategy to reach the public by showing film of the president acting presidential and touting his accomplishments. In news stories, journalists would routinely comment on the "presidential strategy" in cynical terms. But if my ideas are somewhat correct, the public discounted, if not ignored what the journalists had to say, instead absorbing some of the president's message. We do not see a groundswell of support behind the president; the audience is too small for that and many people's political attitudes are too hardened to change. But we do see a modest increase in presidential support across 1998. The president reached a modest number of people through his public relations campaign, driving up his polls. Ironically, despite access to a large audience, he might not have been able to do this in the golden age of broadcast television.

TABLE 1
Tone of News about the President from Three News Organizations,
1953-1978

 Time New York Times CBS News *

Date Positive Negative Positive Negative Positive Negative

1953-78 60 28.2 48.7 27.2 38.5 38.6
1953-65 58.9 10.9 50 11 na na
1966-74 28.7 31.6 28.5 35.2 23.5 38.2
1974-78 ** 43.3 19.4 37.5 23.4 44.8 23.6

* Only from 1968 to 1978.

** Begins when Ford assumes office.

Source: Grossman and Kumar (1981, 256, 265).

TABLE 2
Percentage of News Stories on Government, Military, Domestic,
and/or Foreign Affairs, 1977-1997

Media Organization 1977 1987 1997

ABC 71.6 60.3 44.9
CBS 67.8 54.4 40.3
NBC 62.6 60.1 38.8
Los Angeles Times * 74.8 77.6 62.0
New York Times * 63.8 63.8 69.3
Time Magazine ** 48.1 51.9 19.2
Newsweek ** 46.1 42.3 25.0

Source: Project for Excellence in Journalism. Changing definitions of
the news, March 6, 1998. Available from http://www.journalism.org/
resources/research/reports/definitions/subjects.asp. The coding period
for all years was March of the year.

* Front-page stories.

** Cover stories.

(1.) These data come from Thomas Patterson's (2000) random sample of 5,000 stories from Lexus-Nexus service from 1981 through 1998. His data cast a wide net beyond stories on the presidency; I only present presidential news stories here, about 20 percent of the entire sample. The number of presidential stories per year is often modest, which precludes making definitive statements about news coverage of the presidency. Yet because of the long span of time and the random selection of stories, we can gather a sense of the comparative tone of news reporting on the presidency across these nearly two decades.

(2.) We should not overly interpret the 1987 figure because of the modest number of stories coded that year (25).

(3.) In a perceptive analysis, Newman (2002) finds that the scandal actually hurt Clinton's ratings. Had the scandal not existed, Clinton's polls would have been about two to three points higher. Still, this is a meager effect of scandal on presidential polls in light of Ostrom and Simon's (1989) analysis of the impact of Watergate and Iran-Contra.

(4.) We need to take these numbers with caution. People seem to inflate their news exposure, as they over-report turnout. Given that surveys indicate that some people consider programs such as "Entertainment Tonight" and reality crime shows such as "Cops" to be news, people may overstate their news consumption when all that they have done is watch an entertainment program. Similarly, newspaper-reading attention to the campaign may be inflated as sports readers glance at the news headlines about the campaign. Still, despite these sources of measurement error, assuming that such problems are relatively constant across years, we can track some trends. But it is quite likely that these numbers understate those who are inattentive to either medium. Other data suggest that the percentage of the population that is inattentive to all forms of hard news has increased (Patterson 2000).

(5.) Centralization was occurring in the newspaper sector, as many papers failed, leaving only a handful of cities with more than one daily paper, and as corporate chains took over many other newspapers. Yet, as Hamilton (2003) argues, the newspaper sector was not immune to competitive pressures. Corporate offices were often highly sensitive to the economic implications of their news product and altered daily papers to reflect these new economic considerations.

(6.) For more details of their study, see http://www.journalism.org/resources/research/reports/ definitions/subjects.asp.

(7.) The first 100 days: How Bush versus Clinton faired in the press. Available from http://www. journalism.org/resources/research/reports/100days/default.asp.

(8.) A regression of negative news on a time counter (1949 = 1) is strongly statistically significant, and suggests that each year that the series progresses the president will receive about half a percentage point more negative news. The regression equation (standard errors in parentheses) is y = 10.10 (2.93) Constant +.54 (.11) Counter; [R.sup.2] = .35. A similar regression for positive news finds a near mirror result. Each additional year subtracts about .4 percentage points of positive news. The resulting equation is y = 50.41 (2.83) Constant -.39 (.11) Counter; [R.sup.2] = .23.

(9.) This judgment is made by comparing the percentage of positive news and negative news. For almost every year of the 1980s and 1990s, there was more negative than positive news.

(10.) This is calculated by summing the number of days in the past week that respondents claimed to have read the newspaper and watched the evening network news.

References

Balutis, Alan P. 1976. Congress, the president, and the press. Journalism Quarterly 59 (Autumn): 509-15.

--. 1977. The presidency and the press: The expanding public image. Presidential Studies Quarterly 7 (Fall): 244-51.

Baum, Matthew A. 2003. Soft news goes to war: Public opinion and American foreign policy in the new media age. Princeton: Princeton University Press.

Baum, Matthew A., and Samuel Kernell. 1999. Has cable ended the golden age of television? American Political Science Review 93 (March): 99-114.

Bennett, Stephen Earl, Staci L Rhine, and Richard S. Flickinger. 1998. Assessing Americans' opinions about the news media's fairness in 1996 and 1998. Political Communication 18 (April): 163-83.

Brody, Richard. 1991. Assessing the president: The media, elite opinion, and public support. Stanford: Stanford University Press.

Cappella, Joseph N., and Kathleen Hall Jamieson. 1997. Spiral of cynicism: The press and the public good. Oxford, England: Oxford University Press.

Cook, Timothy E., and Paul Gronke. 2001. Dimensions of institutional trust: Is public confidence in the media distinct from other institutions? Paper presented at the Annual Meeting of the Midwest Political Science Association in Chicago, IL.

Cook, Timothy E., and Lyn Ragsdale. 1998. The president and the press: Negotiating newsworthiness in the White House. In The presidency and the political system. 5th ed. Edited by Michael Nelson, 323-57. Washington, DC: CQ Press.

Cornwell, Elmer E., Jr. 1959. Presidential news: The expanding public image. Journalism Quarterly 36 (Summer): 275-83.

Davis, Richard, and Diana Owen. 1998. New media and American politics. New York: Oxford University Press.

DiClerico, Robert E. 1993. The role of media in heightened expectations and diminished leadership capacity. In The presidency reconsidered, edited by R.W. Waterman, 115-43. Itasca, IL: F.E. Peacock.

Edwards, George C. III. 2003. On deaf ears: The limits of the bully pulpit. New Haven: Yale University Press.

Erikson, Robert S., Michael MacKuen, and James A. Stimson. 2002. The macropolity. New York: Cambridge University Press.

Foote, Joe S. 1990. Television access and presidential power: The networks, the presidency, and the "loyal opposition." New York: Praeger.

Gilbert, Robert E. 1981. Television and presidential power. Journal of Social, Political, and Economic Studies 6 (1): 75-93.

--. 1989. President versus Congress: The struggle for public attention. Congress and the President 16 (1): 91-102.

Groeling, Tim, and Samuel Kernell. 1998. Is network news coverage of the president biased? Journal of Politics 60 (4): 1063-87.

Grossman, Michael Baruch, and Martha Joynt Kumar. 1981. Portraying the president: The White House and the news media. Baltimore: Johns Hopkins University Press.

Hamilton, James T. 2003. All the news that's fit to sell: How the market transforms information into news. Princeton: Princeton University Press.

Hess, Stephen. 2000. Federalism and news: Media to government: Drop dead. Brookings Review 18 (Winter): 28-31.

Kernell, Samuel, and Gary C. Jacobson. 1987. Congress and the president as news in the nineteenth century. Journal of Politics 49 (November): 1016-35.

Kurtz, Howard. 1998. Spin cycle: Inside the Clinton propaganda machine. New York: Free Press.

Maltese, John Anthony. 1992. Spin control: The White House Office of Communications and the management of presidential news. Chapel Hill: University of North Carolina Press.

Newman, Brian. 2002. Bill Clinton's approval ratings: The more things change, the more they stay the same. Political Research Quarterly 55 (December): 781-804.

Ostrom, Charles W., Jr., and Dennis M. Simon. 1989. The man in the teflon suit? The environmental connection, political drama, and popular support in the Reagan presidency. Public Opinion Quarterly 53 (Autumn): 353-87.

Patterson, Thomas E. 2000. Doing well and doing good: How soft news and critical journalism are shrinking the news audience and weakening democracy--and what news outlets can do about it. Joan Shorenstein Center for Press, Politics, and Public Policy, John F. Kennedy School of Government, Harvard University.

Popkin, Samuel. 1998. When the people decline to be spun. New York Times, November 10, p. A29.

Ragsdale, Lyn. 1997. Disconnected politics: Public opinion and presidents. In Understanding public opinion, edited by Barbara Norrander and Clyde Wilcox, 229-51. Washington, DC: CQ Press.

Sabato, Larry J. 1991. Feeding frenzy: How attack journalism has transformed politics. New York: Free Press.

Welch, Reed L. 2000. Is anybody watching? The audience for televised presidential addresses. Congress and the Presidency 27 (Spring): 41-58.

Zaller, John R. 1992. The nature and origins of mass opinion. New York: Cambridge University Press.

Jeffrey E. Cohen is professor of political science at Fordham University. His most recent book, Presidential Responsiveness and Public Policy-Making, received the 1998 Richard E. Neustadt Award of the Presidency Research Group of the American Political Science Association. He currently serves as the editor of the "Polls" feature in Presidential Studies Quarterly.

AUTHOR'S NOTE: An earlier version of this article was presented at the Conference on Researching the Public Presidency, Bush Presidential Library Conference Center, Texas A&M University, February 27-28, 2004. I want to thank Tom Patterson and Lyn Ragsdale for allowing me to use data that they have collected. This article would not be possible without their data.