INTERNATIONAL CRISES AND PUBLIC SUPPORT FOR PRESIDENT CLINTON
UNITED NATIONS' ARMS R|CHARD A. INSPECTION IN IRAQ brody
Abstract
Since the publication of John Mueller's Wars, Presidents and Public Opinion (1973) scholars have been alert to the possibility that international crises, irrespective of the wisdom of the policies from which they arise, can increase public support for the president of the United States. Mueller based his explanation, of the so-called "rally" phenomenon, on a patriotic public reaction to threats to a nation's safety. In Assessing the President: The Media, Elite Opinion and Public Support (1991 ) Brody and Shapiro challenge Mueller's patriotism hypothesis. Their data suggest that rallies appear if (and only if) opposition elites do not openly and actively criticise the policies, which gave rise to the crisis. Three ingredients are necessary and sufficient to produce a rally: (1) A presidential monopoly of the interpretation of the meaning of the events and the actions required to end the crisis; (2) A silent or supportive political elite; (3) News media that abjure independent policy criticism which, instead, depend entirely upon political elites to provide policy oppo-sition. The second and third ingredients produce the president's interpretive monopoly. Faced with but one interpretation of the events and the actions, to which they give rise, the public has no basis for withholding its support. Foreign affairs are inherently ambiguous to most of the American public and without good reason to question the wisdom of foreign policy, the benefit of the doubt goes to the president. This paper will examine the usefulness of the Brody-Sha-piro hypotheses in helping us understand the media, elite and public responses to the crisis that followed Iraq's attempt to limit the United Nations' inspection of its arms production facilities in February 1998.
Richard A. Brody is Emeritus Professor, Department of Political Science, Stanford University, email: brody@leland.stanford.edu.
KD KD
Ln Ln
no
ć? o o
fN
O >
A 3
a
0
LH LH
Since the publication of John Mueller's Wars, Presidents and Public Opinion (1973) scholars have been alert to the possibility that international crises, irrespective of the wisdom of the policies from which they arise, can increase public support for the president of the United States. Mueller bases his explanation, of the so-called "rally" phenomenon, on a patriotic public reaction to threats to a nation's safety. In Assessing the President: The Media, Elite Opinion and Public Support (1991, Chapter 3) Brody and Shapiro challenge Mueller's patriotism hypothesis. Their data suggest that rallies appear if (and only if) opposition elites do not openly and actively criticise the policies, which give rise to the crisis.
Three ingredients are necessary and sufficient to produce a rally: (1) A presidential monopoly of the interpretation of the meaning of the events and the actions required to end the crisis. (2) A silent or supportive political elite. And (3) news media which neither offer their own policy criticism nor seek out and report "unofficial" policy criticism. This is to say, if media build their coverage of international crises on the opinions of political elites and do not cover the opinions of other potential sources (Bennett 1990; Mermin 1999), rallies in public support will cycle with the absence or presence of elite criticism.1
In Mueller's formulation "rally" circumstances activate latent patriotism which manifests itself as an unwarranted and unexpected increase in aggregate public support. The increase is "unwarranted" because international crises, even those that eventuate in policy success, generally originate from policy failure. The increase is "unexpected" because in Mueller's formulation support for the president declines inexorably. Rallies need no intermediation; the public simply responds instinctively to the perception of threat to the United States. Brody and Shapiro's formulation treats a rally as a normal aggregate psycho-political response to an unusual situation. What is "unusual" is the active or passive acceptance, by the president's political opponents, of his assessment of the situation and his proposals for dealing with the crisis. Under this formulation the public responds to opinion leadership.
In principle, active or passive acceptance by the opposition of a president's assessment of a situation and his proposals for dealing with it could happen in either domestic or foreign policy. However, in the context of American politics, the president's domestic policy pronouncements are rarely accepted without criticism. The absence of acquiescence to the president in domestic politics is a natural consequence of the constitutionally designed separation of powers. Members of Congress, even those who share the president's partisanship, frequently have the incentive and opportunity to take (Mayhew 1974) and publicise (Cook 1989) domestic policy positions in opposition to the president's. Under "divided" government it will be rare for the president's domestic policy proposals to escape high profile criticism. Acquiesce in foreign affairs, even under divided government, is much less rare.2
A rally manifests itself as a departure from the trend in public support for the president. It may arrest a decline or add to an increase. In either case, we would expect the departure to be temporary but not necessarily short-lived. This raises the question of whether we can say anything systematic about the magnitude and duration of the "departure."
Burbach (1995) proposes and tests four models to explain rallies in presidential support in the period from 1953 to 1991: The first, "null," model would deny any
regular and routine increase in public support for the president in response to foreign policy crises. Burbach (1995, 3) finds that, for the fifty cases he examines, the average 3.8% increase in public approval of the "way" the president is "handling his job" argues against the null model. There appears to be a rally phenomenon to explain.
The second model is "conflict-cohesion" (Burbach 1995, 5). This model, a variant of Mueller's "patriotism" hypothesis, construes rallies as the collective result of instinctual, individual responses to foreign threat. Applied to presidential support this hypothesis implies that members of the public, theretofore holding a negative opinion of presidential performance, treat the president as the embodiment of the nation and revise their opinion in the face of foreign threats. This model encounters a crucial difficulty: viz., there are a large number of events, that meet the criteria for rallies in public support in response to international crises, that are not associated with a rise in public approval of the president. This points to a "patriotic" response that is conditional sometimes it appears sometimes not. Among other things this means that our explanatory task shifts to specifying the conditions under which rallies do or do not appear.
The third model is "media signalling" (Burbach 1995, 6). It starts with the observation that the public is basically unfamiliar with foreign policy. A lack of familiarity and the uncertainty that accompanies rapidly developing foreign crises increase the public's reliance on elite opinion leadership. Media coverage faithfully3 reports elite consensus or dissent to the public and the public responds in accord with the state of elite opinion.
Brody and Shapiro (in Brody 1991, chapter 3) add a partisan component to this account: In situations fraught with uncertainty the public is likely to use its partisanship to resolve ambiguity. If opinion-leading elites are split along partisan lines and partisans in the public take their cues from the leaders of their party, we would not expect a rally. However, if the president's partisan opponents actively or tacitly support his response to a crisis, opposition partisans in the public will be likely to resolve ambiguity with information provided by the president and increase the likelihood of expressing approval of his actions.
Burbach's fourth model (1995, 7) treats the public as "rational" in the sense of being able to compare outcomes and preferences and to respond positively when its preferences are met and negatively when they are not. In the aggregate Americans are most likely to approve of the use of force when it is used to protect Americans abroad and least likely when force is used to intervene in the domestic affairs of another country or in support of an international humanitarian effort. The responses to other purposes for which force is used are arrayed in between (Burbach 1995, 22-23) successful application of force is likely to boost approval. For the set of events Burbach examines, the fourth model performs best.4
As interesting as is the Burbach (1995) study it still leaves open the question of how we can account for the magnitude and duration of a rally departure. Other things equal, the magnitude of a rally departure should relate to the volume of media coverage focused on the president's account of the causes of the crisis and the actions needed to end it. The duration of a rally should be a function of the length of time opinion-leading elites are supportive — actively or tacitly — and accepting of the president's explanation and actions.
The literature suggests that we need measures of media attention and elite consensus in order to account for the magnitude of the rally displacement:
News Attention: News attention is evinced by the time devoted to stories about the crisis by network television news and major newspapers on a given day. It is simply the fraction of the total time the three, twenty-two minute network evening news shows spend on the crisis. Newspaper lineage will serve as a measure of print media attention. Conditioned on elite consensus, the more news attention to a crisis, the greater the rally displacement.
Elite Consensus: Elite consensus represents the balance of negative or positive commentary on a crisis on a given day offered by congressional leaders in the president's party and in the opposition party. Our expectation is that the more positive and neutral consensus among political elites, the greater will be the magnitude of the rally.
Rally Displacement: Rally displacement should be a function of the product of news attention and elite consensus. Both the magnitude and duration of a rally also have a partisan component: It should be easier for the president to convert his fellow partisans from disapproval to approval than it is to convert opposing partisans. However, approval is subject to a ceiling effect and a president's fellow partisans are more likely to reach the ceiling than are his partisan opponents or political independents. Taken together these considerations suggest that, ceteris paribus, the magnitude of a rally is conditional on the pre-crisis level of support among the president's fellow partisans. The lower the pre-crisis support among the president's fellow partisans — the further support among his partisans is from the ceiling — the greater the magnitude of the rally.
Assuming that the president's fellow partisans are positive or passive, the duration of a rally will be a function of the passive or active acquiescence of the partisan opposition. All else equal, the more pervasive the supportive consensus is, among the opposition elite, the longer will the rally last.
To see these processes at work we will examine news attention, elite consensus and rally displacement during the crisis provoked by Saddam Hussein's shutting down of the United Nations' inspection of arms production facilities in Iraq. We will begin with a recapitulation of the state of public opinion before, during and after the playing out of the arms inspection crisis.
Public Support for President Clinton, 1997-2000
Public approval of President Clinton's handling of the job of president, after his reelection in 1996, is traced in Figure 1.5
Three statistically distinct sub-periods are evident in this time series. In 1997 the president's approval rating averaged 58.2%, in 1998 it averaged 63.9%, and for 1999-2000 the average is 60.4%. The 1998 average is 5.7 percentage points larger than that for 1997 (pt = 11.21) and 3.5 percentage points larger than 1999-2000 (pt = 4.05). The 2.2 percentage point difference between the averages for 1997 and 1999-2000 is also statistically significant (pt = 4.22). Although the differences are small they are unlikely to merely reflect the vagaries of sample surveys. President Clinton had less support in 1997 than in 1998 and was evaluated more positively in 1998 than in 1999 and the first half of 2000.
Figure 1: Evaluation of Clinton, January 1997 to July 2000
It is noteworthy that the thirty-six poll periods in 1998 reflect public opinion at the time that a series of related domestic news stories drew the attention of the American press and public: News of the president's involvement with Monica Lewinsky broke in mid-January 1998 and the scandal continued to attract coverage throughout the year. The "impeachment story" occupied a great deal of press attention during the second half of 1998 and the early winter 1999. The investigations of the president by Special Prosecutor Kenneth Starr and by the Congress were among the top news stories of the year. All of this domestic scandal news needs to be kept in mind as we carry out our examination of the effect of the Iraqi arms inspection crisis on public support for President Clinton.
It is surprising, with news of his personal misconduct swirling around President Clinton, that 1998 should have been the year in which the American people expressed the highest level of satisfaction with the president's "handling of his job." What role, if any, did the crisis over the inspection of Iraqi arms production facilities play in public evaluations of President Clinton? Did the Iraqi situation produce a "rally" which propped up support for the president among the American people?
These questions will be addressed via a brief review of news of the crisis in Iraq. Press attention to the story will be noted and the support or lack of support of presidential actions by opposition elites — Republican leaders, in this case — will be summarised, as well.
News Coverage of the Iraqi Crisis
Our review of coverage of the Iraqi arms inspection crisis begins on February 2, 1998 with a, relatively buried, report — page 10 of the A section of the New York Times7 — of foreign diplomats going to Baghdad to attempt to persuade Saddam Hussein to restart inspection, by the United Nations, of Iraq's arms production facilities. This story rather neatly encapsulates the elements of the crisis: By shutting down the United Nations' arms inspection program, Iraq was in violation of UN resolutions and abrogating its obligations under the conditions imposed upon it at the termination of Gulf War in 1991. Furthermore, Iraq's actions contributed to the suspicion that it was engaged in the clandestine production of so-called "weap-
ons of mass destruction." These actions challenged the argument that international sanctions, per se, were effective in holding in check Iraq's potential military threat to its immediate neighbours and Israel.
The shutting down of the arms inspection program brought into question many of the assumptions which underlay the means selected by the U.S. to carry out its policy of neutralising Iraq as a regional military threat. As such, it represented the kind of foreign policy failure that, if the conditions are right, is the occasion for a public opinion rally.
On February 3, 1998 the story became page-one news. The New York Times reported — page one of the A section — that Saudi Arabia had given tentative approval to the United States for air strikes against Iraq if diplomacy failed to get the inspections restarted. On February 4 the Times reports — page one of section A — that Defence Secretary Cohen had begun the process of seeking support from opposition leaders in Congress. The story mentions contacts with Trent Lott and John McCain in the Senate and Dick Armey in the House. White House contacts with United Nations' Secretary General Kofi Annan are also reported. These news stories reflect the president's desire to "rally" domestic and foreign leadership support for a policy shift from sanctions to military force.
The president's desire notwithstanding, the effort to gain assent from Republican (i.e., the opposition party) congressional leadership was quickly frustrated. On February 5 the Times reports — page one, section A — that the White House and the Republican leadership in Congress were "at odds" on Iraq policy. Republican leaders in both the Senate (Majority Leader Trent Lott) and the House (Speaker Newt Gingrich) called for the use of American military power to remove Saddam Hussein and not simply to get the arms inspection program restarted. On February 6 more criticism of President Clinton's Iraq policy by Republican leaders was forthcoming — page one, Section A. The Times reported that Republicans were saying that Clinton's policy aims were unclear and that while Britain and the U.S. might agree on the use of air power other nations did not approve of the policy shift. Representative Floyd Spence, the Republican Chair of House Committee on National Security is quoted as declaring that the administration had not sorted its options nor defined the ultimate goal of attacking Iraq.
News reports on February 7, 1998 detailed the Clinton administration's warning of impending air strikes on Iraq. On February 8, the Times reported the administration's success in getting our European allies to back the policy shift. It also reported an increase in the size of our air strike force in the Gulf.
The president's diplomatic success on February 8 was offset on February 9 by reports of Saudi Arabia's signalling to the administration that U.S. war planes would not be welcome at Saudi airbases (New York Times, A1). On February 10 the press reported an increase of American ground forces in the Gulf (New York Times, A8). To ensure that our forces would be welcome, Secretary Cohen travelled to the region to meet with and gain the assent of our allies. Republican spokesmen announced that their policy for Iraq would be to support Saddam Hussein's Iraqi opponents in order to foment insurgency.
On February 11th a front-page Times news analysis opined that President Clinton was beating the drums for war but "very slowly" (New York Times, A1). On February 12 the Times reported that U.S. commanders were preparing the attack (A6). On February 13 more analysis of the situation was offered on page one.
On February 14 the Times detailed the White House's case against Iraq and its efforts to prepare the American public for war (page A1). In a sense, the United Nations supported the administration's effort by making public its experts' opinion that Iraq was hiding weapons of mass destruction. On February 15, 1998, as the chances of military action increased, the New York Times reported on President Clinton's detailed specification of the limited goals for air strikes on Iraqi targets (p. A6). This is interpreted, by the Times, as part of the president's strategy of "going public" to gain popular backing for his policy (Kernell 1986). The opposition, however, was not satisfied with the president's goal specification, Ted Stevens from Alaska, the Republican chair of the Senate Committee on Appropriations, expressed frustration at the lack of clarity of the administration's Iraq policy goals (New York Times, p. A6).
On Monday, February 16, the Times reported that plans for the attack on Iraq were moving ahead (page A1). It reports that the considerations that will affect the exact date of the air strike would include weather, Muslim holidays, and parents weekend at Stanford University.8 The next day the Times reported that the U.S. was seeking to limit the role of Secretary General Kofi Annan in negotiations with the Iraqis (February 17, 1998, p. A1). The Republican congressional leadership urged President Clinton to make the ouster of Saddam Hussein a top priority and Senators McCain and Lott and Speaker Gingrich produced a "check list" of policy goals for the president.
On February 18 President Clinton responded to the Republican challenge and described the goals for the air strikes (New York Times, page A1). The statement received a positive "review" from New York Times correspondent Philip Shenon whose analysis concluded that modest goals should produce easy victories (p. A1).
On February 19 President Clinton continued his campaign for the support of the American public. At a public forum, at Ohio State University, Clinton presented the arguments for air strikes as the way to force a restart of the United Nations' inspection program in Iraq. The audience expressed scepticism about the policy's viability (New York Times, p. A1). On February 20 the Times reports that the president is of the opinion that the country would united behind an attack (p. A1). But the Times also reports that the states bordering Iraq were fearful that they would suffer "collateral damage" from an American attack.
On Saturday, February 21, the Times published a page one story that detailed the administration's plan for four days of around-the-clock bombing of Iraq. The story carried no comment — positive or negative — from Republican leaders.
However, on Sunday (February 22), with air strikes imminent, the United Nations re-entered the situation. Kofi Annan expressed the hope that military action could be avoided. On the 23rd Annan announced that an agreement with Iraq had been reached which would allow the resumption of inspection of arms facilities including the so-called "presidential sites" (New York Times, p. A1). The White House expressed caution about the agreement but, on February 24, President Clinton accepted the agreement while warning Iraq that it must comply (New York Times, p. A1). On February 25 Secretary of State Albright laid out the terms of the agreement with Saddam Hussein (New York Times, p. A1).
On February 26, 1998, Trent Lott urged the rejection of the pact with Iraq and accused President Clinton of capitulating to Saddam Hussein. At this point the story disappears from the news. After three tense weeks, arms inspection crisis came to an end.
Confrontation between the United States and Iraq over United Nations inspection of arms production facilities flared up again briefly in November 1998. It was a six-day version of the February crisis replete with threats and an increase in the U.S. bomber force in the Gulf. In this case, a resolution acceptable to the U.S. came about more quickly when Saddam Hussein found himself without any support in the Arab world.
Public Support for President Clinton during the February 1998 Crisis
Three of the national polling organisations — Washington Post/ABC Television News, New York Time s/ CBS Television News, and Gallup/CNN/USA Today — have a time series that brackets the crisis in February 1998. None of the three contain any public opinion evidence that would suggest a rally, in public support for President Clinton, in response to the crisis in Iraq.
The Washington Post/ABC Television News poll measured the public's evaluation of President Clinton's "job performance" in its surveys on January 28-31, 1998 — before Iraq's defiance of the United Nations inspection requirement — on February 17-18, 1998 — in the middle of the crisis, and on March 16, 1998 — three weeks after the situation was resolved. President Clinton's level of approval in these three surveys is 67 percent, 67 percent and 63 percent respectively.
The New York Times/ CBS Television News poll had five surveys immediately before, during, and just after the Iraqi crisis. The field dates for these surveys are: January 26, February 1, February 8, February 19-21, and March 1-2, 1998. The percentages "approving" of the President's job performance in the five surveys are 57 %, 72 %, 66 %, 68 %, and 64 % respectively. The fifteen percentage point rise in support for President Clinton between the January 26 and February 1 polls is unlikely to be associated with the situation in Iraq since it preceded press attention to the crisis. The jump in support is likely to be a reflection of the public's reaction to President Clinton's State of the Union Address on January 28, 1998 (Zaller 1998; Fischle 2000).
The Gallup/CNN/USA Today poll was in the field four times in the period of the crisis: Surveys were taken January 30-February 1, February 13-15, February 20-22, and March 6-9, 1998. The levels of "approval" observed in these polls are 69 %, 66 %, 66 %, and 6 3%, respectively. The peak here is reached well before news of the crisis in Iraq reached the American people.
On this evidence there was no rally in response to the foreign policy crisis stemming from Iraq's refusal to cooperage with the United Nations' inspection of its arms production facilities. Support for President Clinton was strong throughout February 1998 but there was no upward displacement of support, which would indicate the presence of a rally.
Why no Rally?
We can consider the actions of the news media, the actions of the political elite, and the structure of public opinion in seeking to explain why support for the president did not increase in response to this rally opportunity.
Can the absence of a rally be attributed to a lack of media attention? It seems doubtful; there was coverage every day between February 2 and February 26, 1998. Coverage of the crisis was usually quite extensive, detailed, and prominently displayed — for example, most of the time the crisis was page-one news in the New
York Times. Since there was ample media attention, we will have to consider other factors, which may have prevented a rally in public support.
Opinion leadership is a possible source of the lack of displacement upward of support for President Clinton. As was hypothesised earlier, a rally is a function of the passive or active acquiescence to the policy of the partisan opposition. All else equal, the more pervasive the supportive consensus is, among the opposition elite, the longer will the rally last. Rallies in public support cycle with the absence or presence of elite criticism.
Neither support nor silence was characteristic of the opposition leadership's approach to the arms inspection crisis. Two themes dominated Republican commentary on the administration's policy of using air strikes to re-establish United Nations inspection of Iraq's arms production facilities as a means of curtailing military threat to its neighbours in the region and to Israel: The policy was said to be "unclear" and "short-sighted." "Unclear" because the administration had not specified what the air strikes were meant to achieve. "Short-sighted" because removing Saddam Hussein from the presidency of Iraq was not a declared aim of the policy. These themes were stated and repeated by high-ranking Republicans — that is, by Republican legislators who comprised the leadership of the majority party in both houses of Congress.
Under these circumstances the "no rally" response is consistent with the opinion leadership hypothesis: Citizens who looked to the president's opponents for guidance on how to judge his performance in the arms inspection crisis would have found, in the Republican leaders' statements, reasons for maintaining their negative judgement of President Clinton.
The "no rally" response is also consistent with the general state of opinion on presidential performance in 1998, the second year of his second term. Within the American public, opinion on the question of the president's "handling his job" had become increasingly polarised. By 1998, the vast majority of those who "approved" of President Clinton's job performance did so "strongly." The mirror image is also true: Three in four of those who "disapproved" of the way Clinton was handling the presidency claimed to "strongly" disapprove. Other things equal, we would expect those professing strongly held opinions to be more steadfast than citizens lukewarm in their support or opposition. Polarised opinion work against a rally being observed: Those citizens who are steadfast in their approval cannot contribute to an increase in public support; they already "approve." Those who are steadfast in their disapproval are unlikely to change their mind or to change their mind sufficiently to move from "strongly disapprove" to one or another "approve" response. The likelihood of change, for these citizens, is further reduced when opposition elites are, as they were in this case, questioning the motives for and wisdom of presidential actions.
Nineteen ninety-eight, saw a remarkable transformation of the level and structure of President Clinton's support. The overall level of disapproval declined to its stable hardcore. Those who were tentative or lukewarm in their opposition to the president came to comprise less than seven percent of the electorate. In the December 1998 polls, just under a third of adult Americans disapproved of President Clinton's handling of his job but those who declare that they "strongly" disapprove of the president, always a majority of disapprovers, were fully three-quarters of citizens evaluating him negatively.
On the other hand, growth in support for the president — the overall level was sixty-six percent in December 1998 — was accompanied by growth in the fraction of "approvers" who acknowledge a strongly held opinion. In January, 1998 just over half of those approving said they "strongly" approved; by December strong approvers comprised seventy percent of President Clinton's supporters.
During 1998, opinion on President Clinton's handling of the job as president came to be polarised in the American electorate to a degree not previously seen in his presidency. From 1993 to 1998, the fraction of the electorate expressing strong disapproval increased twenty-six percent, the fraction who were tentative or lukewarm in either their support or opposition decreased thirty-eight percent, and the ranks of those strongly approving of the president's handling of the job grew by seventy-four percent.
"Polarisation" indicates the extent to which the electorate is comprised of those holding strong opinions pro or con on an issue or individual. In its assessment of the president in 1993 less than half of the electorate (45.4 %) was polarised; by the time of the Iraq arms inspection crisis, in February 1998, sixty-five percent of the electorate was polarised. And, at the time that the impeachment battle moved into the Senate, three-fourths of the American public was "polarised" on the question of the quality of President Clinton's job performance. Under circumstances like those in 1998, if strong opinion holders are truly more steadfast in their opinion holding than those with less strong opinions, because of the low probability of "strong disapprovers" switching to "approval," we would expect the absence of rallies to be the norm.
The "opinion leadership" and "structure of opinion" explanations are complementary rather than rivals. They can both be true. Both may have been operating in the February 1998 arms inspection crisis. On can easily imagine than an American strongly disapproving of President Clinton's job performance before February 1998 would have the small probability of changing his/her opinion reduced even further by the criticisms of Iraq policy provided by Republican leaders in Congress. In this case, the individual's prior opinion and opinion leadership reinforcing the priors may have precluded a rally in support.9
Discussion and Conclusion
The study of a single case can be, at best, suggestive. This case suggests that the mass opinion response to news of an international crisis is conditioned by the actions of those, reporters and editors, who construct and present information to the public, by the actions of opinion leaders who instruct the media and public on the meaning of foreign policy events, and by the structure of public opinion at the time the crisis becomes a news story. "Rallies" depend upon a particular configuration of these conditions: An absence of news reporting alternative and conflicting interpretations of the events by partisan elites and a public opinion, sufficiently ductile, to respond to the information it receives. This configuration was absent in February 1998.
Media has the responsibility of bringing to the public the "facts" of the situation. These facts include, if they exist, alternative interpretations of the meaning of the events as they unfold. Given the media's penchant for reporting politics as conflict it is likely that differing interpretations, offered by "legitimate" sources,
will be brought to public attention. How broad must be the media's search for "alternative" explanations and, in consequence, the judgement of how well the media meets their responsibility to inform the public, is a matter of debate (Mermin, 1999).10
The responsibilities of the opposition political elite to offer an alternative construction of events and of the public to keep an "open mind" are even harder to specify than the responsibilities of the media. The best we can hope for is an ability to shed some light on the circumstances under which members of an opposition elite will state its beliefs and the circumstances under which the public will attend to an implicit policy "debate" with an open mind. But even these modest hopes cannot be satisfied within the scope of this paper.
Notes:
1.	Bennett (1990) and Mermin (1999) label the tying of coverage to governmental-elite opinion "indexing."
2.	This could have been a product of the Cold War. The magnitude of the perceived threat from the Soviet Union and the bi-polar structure of the international system may have created pressures for bipartisanship in foreign policy before the breakup of the Soviet Bloc and the putative end of the Cold War beginning in 1989. It is an open question whether or not the "new world order" carries with it a reduced probability of elite acquiescence to the president in foreign affairs.
3.	Mermin criticises the press for narrowing the "spectrum" of foreign policy debate to the views of supportive and opposition elites. However, he leaves little question that these elites provide most (if not all) of the information available to the public (1999, 17-35).
4.	Burbach's focus on the use of force leads to the inclusion of events in which U.S. military involvement was sustained over fairly long periods, e.g., the Gulf of Tonkin, bombing in North Vietnam, troop deployments in Vietnam, the invasion of Cambodia, and the war in the Persian Gulf. In these situations success or failure is expected to be an important ingredient of the public's assessment of presidential performance. There is, however, no need to employ the notion of a "rally" to account for public support in these situations. The idea of a rally was developed to try to help us understand the public's positive response when foreign policy outcomes were manifestly negative, i.e., contrary to the United States' interests.
5.	Data for Figure 1 are drawn from eight national polls — viz., Washington Post/ABC Television News, New York Times/ CBS Television News, CNN/Time Magazine, Gallup/CNN/USA Today, Wall Street Journal/NBC Television News, Newsweek Magazine, and The Pew Research Center. Each month is divided into three ten-day periods. Multiple polls from a given polling organisation, within a ten-day period are averaged. The entries in Figure 1 are the average "approval" percentages across polling organization for a given period.
7.	The source of news reports will be identified in the main body of the text rather than in the reference and endnotes sections in order to accomplish three things simultaneously: 1. Give the reader a complete rendering of the location of the source material; 2. Provide information on "news attention" and "elite consensus" which is germane to the analysis; and 3. Eliminate redundancy.
8.	President Clinton's daughter was then a student at Stanford University. He and/or the First Lady were regular participants at parents events.
9.	Fischle (2000, 145) shows that a process like this conditioned citizen responses to news of the Monica Lewinsky affair.
10.	This is really a debate of how extensive or restricted is the set of "legitimate" sources. This is not a debate that we can settle within the compass of this paper.
References:
Bennett, W. Lance. 1990. Toward a Theory of Press—State Relations in the United States. Journal
of Communication 40, 103-125. Brody, Richard A. 1991. Assessing the President: The Media, Elite Opinion, and Public Support.
Stanford, CA: Stanford University Press. Burbach, David T. 1995. Foreign Policy Preferences, Presidential Approval, and the Use of Force.
Paper presented at the American Political Science Association Annual Meeting, Chicago, IL. Cook, Timothy H. 1989. Making Laws and Making News: Media Strategies in the U.S. House of
Representatives. Washington, DC: Brookings Institution. Fischle, Mark. 2000. Mass Response to the Lewinsky Scandal: Motivated Reasoning or Bayesian
Updating. Political Psychology 21, 138-159. Kernell, Samuel. 1986. Going Public. Washington, DC: Congressional Quarterly Press. Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press.
Mermin, Jonathan. 1999. Debating War and Peace: Media Coverage of U.S. Intervention in the
Post-Vietnam Era. Princeton, NJ: Princeton University Press. Mueller, John. 1973. Wars, Presidents and Public Opinion. New York: John Wiley. Zaller, John. 1998. Monica Lewinsky's Contribution to Political Science. PS: Political Science & Politics 31, 182-189.