PRESIDENTIAL Edwards / BUILDING STUDIES COALITIONS QUARTERLY / March 2000

Building Coalitions GEORGE C. EDWARDS III Texas A&M University

The necessity of forming coalitions is inevitable in a large, diverse nation in which political power is fragmented both vertically and horizontally. Presidents invest substantial time and effort building supporting coalitions for themselves and their polices among the public and within Congress. The author explores the president’s ability to build coalitions, examines the relationship between the institutional presidency and obstacles to coalition building, and inquires whether the presidency as an institution is adequate to the task of building the coalitions necessary for governing. He concludes that (1) the president has great difficulty building coalitions for governing among the public and within Congress, (2) the institutional structure of the presidency (and Congress) is at the core of the difficulty of building coalitions in Congress but does not inhibit obtaining public support, and (3) the changes required to notably improve presidential coalition building are so fundamental that they are unlikely to occur.

Building coalitions is at the core of governing in America. The necessity of forming coalitions is inevitable in a large, diverse nation in which political power is fragmented both vertically and horizontally. Because the president in most instances requires the approval of the legislature to make public policy, Congress is the proximate site of coalition building. Because advances in technology allow the president to reach the public directly and because the White House views public support as crucial to its success, presidents invest substantial time and effort in building support for themselves and their polices among the public. In this article, I explore the president’s ability to build coalitions among the public and within Congress, examine the relationship between the institutional presidency and obstacles to coalition building, and inquire whether the presidency as an institution is adequate to the task of building the coalitions necessary for governing.

Coalition Building among the Public Leading the public is at the center of the modern presidency. As parties weaken and bargaining resources diminish, presidents see themselves increasingly dependent on pubGeorge C. Edwards III is Distinguished Professor of Political Science and the director of the Center for Presidential Studies at Texas A&M University. He also holds the Jordan Professorship in Liberal Arts and has authored dozens of articles and written or edited sixteen books on the presidency and American politics. He is editor of Presidential Studies Quarterly. Presidential Studies Quarterly 30, no. 1 (March) © 2000 Center for the Study of the Presidency

47

48  PRESIDENTIAL STUDIES QUARTERLY / March 2000

lic support to accomplish their goals; and they devote substantial time, energy, and resources to obtaining this support. Presidents “go public” more than ever, depending on a steadily expanding White House public relations infrastructure to take their messages to the American people. In 1995, for example, the White House spent $18 million advertising on behalf of the president—a year before the presidential election (Woodward 1996, 344). Thus, presidents are involved in a permanent campaign to build supportive coalitions. As Bill Clinton reflected on the results of the 1994 elections, he concluded that the principal cause of the Democrats’ stunning defeat was his failure to communicate his achievements. “I got caught up in the parliamentary aspect of the presidency and missed the leadership, bully pulpit function which is so critical” (Woodward 1996, 22). The president’s remark reflects four fundamental and widely shared premises about presidential leadership: 1. Members of Congress are responsive to public opinion. 2. Public support is crucial to the president’s success. 3. The president must not only earn public support with his performance in office but also must actively take his case to the people. Moreover, he must not only do it at reelection time but all the time. 4. Through the permanent campaign, the White House can persuade or even mobilize the public. Leading the public—changing opinions and mobilizing citizens into action—is perhaps the ultimate resource of the democratic political leader. It is difficult for others who hold power to deny the legitimate demands of a president with popular support. Commentators on the presidency often assume that the White House can persuade or even mobilize the public if the president is a skilled enough communicator. One of the crowning ironies of the contemporary presidency is that at the same time that presidents increasingly attempt to govern by campaigning—“going public,” public support is elusive, perhaps more than ever before. In the century since Theodore Roosevelt declared the White House a “bully pulpit,” presidents often have found the public unresponsive to issues at the top of the White House’s agenda and unreceptive to requests to think about, much less act on, political matters. When asked about his “biggest disappointment as president,” George Bush replied, “I just wasn’t a good enough communicator.” To evaluate the president’s ability to build coalitions among the public, we must know how successful presidents are in employing the bully pulpit to lead public opinion. There is much that we have to learn about this topic, but we can obtain a reasonable sense of the challenges facing presidents by reviewing what we know about presidential leadership of public opinion and then briefly examining the experience of two of the most able communicators in the past half century: Ronald Reagan and Bill Clinton.

Edwards / BUILDING COALITIONS  49

What We Know As a society (and uniquely among developed democracies), we devote extraordinary attention to what the president says and how he says it. Political commentators in both the press and the academy routinely evaluate presidents in terms of their public leadership, their ability to articulate a vision, rouse a crowd, or even stick to a speech. Yet, we know very little about the impact of the permanent campaign. Despite the prominence of going public in scholarly commentary on the presidency, very few studies focus directly on the impact of presidential leadership of opinion, and no full-length studies do so. A few experimental studies provide suggestive findings. Sigelman (1980) ascertained public opinion on six potential responses to the 1979-80 hostage crisis in Iran. He then asked those who opposed each option whether they would change their view “if President Carter considered this action necessary” (see also Conover and Sigelman 1982). In each case, a substantial percentage of respondents changed their opinions in deference to the supposed opinion of the president. In another experiment during the Reagan presidency, Thomas and Sigelman (1985) posed policy proposals to sample subjects. When informed that the president was the source of the proposals, enthusiastic supporters of Reagan evaluated them in favorable terms; but when the source was withheld, Reagan supporters evaluated these same proposals unfavorably.1 Not all results are as positive, however. In another study, Sigelman and Sigelman (1981) asked sample groups whether they supported two proposals, a domestic policy proposal dealing with welfare and a proposal dealing with foreign aid. One of the groups was told that President Carter supported the proposals, while the president was not mentioned to the other group. The authors found that attaching the president’s name to either proposal not only failed to increase support for it but also actually had a negative effect because those who disapproved of Carter reacted very strongly against proposals they thought were his. Glaros and Miroff (1983) evaluated the reactions of some persons watching Ronald Reagan address the nation. Their conclusion was that the principal impact of the speeches was to reinforce the audience’s predispositions. They found little evidence of persuasion taking place. A few studies have examined aggregate responses to the president’s communications. Kernell’s (1997) work is the most prominent on the president going public but provides only a few case studies of actual opinion leadership. Since his focus is on describing and analyzing the strategy of going public, he does not provide a systematic study of the response of public opinion to the president. Ragsdale (1984) found a short-term increase of about 3 percentage points in presidential approval following a televised presidential 1. See also a poll of Utah residents that found that although two-thirds of them opposed deploying MX missiles in Utah and Nevada, an equal number said they would either “definitely” or “probably” support President Reagan if he decided to go ahead and base the missiles in those states. “Most Utah Residents Say ‘No’ to MX Missile Deployment,” Bryan-College Station Eagle, September 15, 1981, 5A.

50  PRESIDENTIAL STUDIES QUARTERLY / March 2000

address; and Page and Shapiro (1985) found that presidents could influence public opinion on issues but only when they themselves have high approval ratings.2 Cohen (1997) has done the most extensive work on aggregate opinion. He finds that presidents can influence the public’s agenda through symbolic speech in State of the Union messages, at least in the short run. He also finds, however, that presidents are only able to affect the public’s agenda over time on foreign policy and that substantive policy rhetoric has no impact on the public’s policy agenda. His work on agenda setting has received support from Hill (1998). In general, Cohen finds the president to have only a very modest impact on public opinion. In his examination of public opinion regarding the Gulf War, Zaller (1994) argues that those in the public most susceptible to presidential influence are those attentive to public affairs but who lack strong views. The president’s greatest chance of influencing public opinion is in a crisis (which attracts the public’s attention) in which elites articulate a unified message. At other times, most people are too inattentive or too committed to views to be strongly influenced by elite efforts at persuasion. Despite the limitations of their abilities to exercise direct opinion leadership over the public, presidents are aided by the willingness of Americans to follow their lead, especially on foreign policy. Foreign policy is more distant from the lives of most Americans than domestic policy and is therefore seen as more complex and based on more specialized knowledge. Thus, people tend to defer more to the president on foreign issues than on domestic problems, which they can relate more easily and directly to their own experience. Studies have shown public opinion to have undergone changes in line with presidents’ policies on the liberation of Kuwait, the invasion of Grenada, the testing of nuclear weapons, relations with the People’s Republic of China, isolationism, and both the escalation and the de-escalation of the Vietnam War (CBS News/New York Times Poll 1983, 2; Erikson, Luttbeg, and Tedin 1980, 144; Mueller 1970, 69-74; Page and Shapiro 1992, 182; Rossi 1965; Sussman 1983, 1, 18. However, compare Page and Shapiro 1992, 242, 250). The public generally does not have crystallized opinions on issues and is therefore often easy to sway in the short run. This volatility, however, also means that any opinion change is subject to slippage. As issues fade into the background or positions on issues are confronted with the realities of daily life, opinions that were altered in response to presidential leadership may quickly be forgotten. This is especially likely to occur in the area in which the president’s influence on public opinion seems to be greatest: foreign policy. A substantial and rapidly increasing literature focuses on presidential rhetoric, including works by Campbell and Jamieson (1990), Hart (1987), Medhurst (1993), Smith and Smith (1994), Stuckey (1990), Tulis (1987), Windt (1990), and Zarefsky (1990). Underlying most of the work on political rhetoric is the premise that rhetoric matters. The authors of these fine works concentrate on analyzing what the president said. In the process, they make numerous inferences regarding the impact of the president’s rhetoric on public opinion. However, scholars of presidential rhetoric virtually never provide evidence for their inferences about the president’s impact (Edwards 1996). 2. Cohen (1997), however, found little impact of approval on public leadership.

Edwards / BUILDING COALITIONS  51

Ronald Reagan Ronald Reagan’s standing with the American people is legendary. In contrast to his two immediate predecessors, the public viewed him as a strong leader. He overwhelmingly won a second term in office. This seeming love affair with the public generated commentary in both academia and the media about the ability of “The Great Communicator” to move public opinion to support him and his policies. Perhaps the most notable recent example of a president mobilizing public opinion to pressure Congress is Reagan’s effort to obtain passage of his bill to cut taxes in 1981. Shortly before the crucial vote in the House, the president made a televised plea for support of his tax cut proposals and asked the public to let their representatives in Congress know how they felt. Evidently this worked, for thousands of phone calls, letters, and telegrams poured into congressional offices. How much of this represented the efforts of the White House and its corporate allies rather than individual expressions of opinion will probably never be known. But on the morning of the vote, Speaker Tip O’Neill declared, “We are experiencing a telephone blitz like this nation has never seen. It’s had a devastating effect” (quoted in Congressional Quarterly 1981. See also Kernell 1997, 150-51). With this kind of response, the president easily carried the day. What is especially notable about this incident is its rarity. If we examine closely Reagan’s rise to power, we find that the basic themes Reagan espoused in 1980 were ones he had been articulating for many years: government was too big, the nation’s defenses were too weak, pride in country was an end in itself, and public morals had slipped too far. In 1976, conditions were not yet ripe for his message. It took the Carter years—with their gas lines, raging inflation, high interest rates, Soviet aggression in Afghanistan, and hostages in Iran—to create the opportunity for his electoral victory. By 1980, the country was ready to listen. Commentators such as Reagan advisers Anderson (1990, xviii-xix, 7) and Niskanen (1988, 22) and journalist Johnson (1991, 49) agree that Reagan gave the conservative movement focus and leadership, but did not give it life. He was the agent around which already existing conservative thought coalesced and rallied. More systematic data support the view that Reagan had a receptive audience. Stimson (1991) concluded that “movements uniformly precede the popular eras.” The conservative winds of the 1980s were “fully in place before the election of Ronald Reagan.” He was the beneficiary of a conservative mood, but he did not create it (pp. 64, 126-27). Similarly, Page and Shapiro (1992, 127, 136) found that the right turn on social welfare policy took place before Reagan took office and ended shortly thereafter. Davis (1992) also found that prodefense and antiwelfare conservative trends had occurred by the late-1970s, before Reagan’s nomination. Mayer (1992, 123) produced similar findings, while Smith (1990) found that liberalism had plateaued by the mid-1970s. Another aspect of Reagan’s coming to power is of direct interest to us. Although he was the preferred candidate of the American people in 1980 and 1984, Reagan also was the least popular candidate to win the presidency in the 1952 to 1988 period. His supporters displayed an unusual degree of doubt about him, and those who opposed him disliked him with unprecedented intensity (Wattenberg, 1991, chap. 4).

52  PRESIDENTIAL STUDIES QUARTERLY / March 2000

Thus, Reagan arrived at the White House on the crest of a preexisting tide of conservatism that he helped to articulate but not to create. What happened after he took office? Was he able to use the bully pulpit to move the public to support his policies if it were not already inclined to do so? Reagan knew better. In his memoirs, he reflects on his efforts to ignite concern among the American people regarding the threat of communism in Central America and mobilize them behind his program of support for the Contras. For eight years the press called me the “Great Communicator.” Well, one of my greatest frustrations during those eight years was my inability to communicate to the American people and to Congress the seriousness of the threat we faced in Central America. (Reagan 1990, 471) Time and again, I would speak on television, to a joint session of Congress, or to other audiences about the problems in Central America, and I would hope that the outcome would be an outpouring of support from Americans who would apply the same kind of heat on Congress that helped pass the economic recovery package. But the polls usually found that large numbers of Americans cared little or not at all about what happened in Central America—in fact, a surprisingly large proportion didn’t even know where Nicaragua and El Salvador were located—and, among those who did care, too few cared enough about a Communist penetration of the Americas to apply the kind of pressure I needed on Congress. (Reagan 1990, 479)

Numerous national surveys of public opinion have found that support for regulatory programs and spending on health care, welfare, urban problems, education, environmental protection, and aid to minorities increased rather than decreased during Reagan’s tenure (Mayer 1992, chaps. 5-6; Page and Shapiro 1992, 133, 136, 159; Schneider 1985; Lipset 1986, 228-29. See also CBS News/New York Times Poll 1987, Tables 16, 20; Davis 1992; Federal budget deficit 1987, 25, 27; Opinion outlook 1987a; Supporting a greater federal role 1987). On the other hand, support for increased defense expenditures was decidedly lower at the end of his administration than when he took office (Lipset 1986, 229; Mayer 1992, 51, 62, 133; Bartels 1994. See also CBS News/New York Times Poll 1987, Table 15; Defense spending 1987; Opinion outlook 1987b). (This may have been the result of the military buildup that did occur, but the point remains that Reagan wanted to continue to increase defense spending, but the public was unresponsive to his wishes.) In each case, the public was moving in the opposite direction to that of the president. In the realm of foreign policy, the president, as we have seen, was frustrated in his goal of obtaining public support for aid to the contras in Nicaragua (Page and Shapiro 1992, 276; Reagan 1990, 471, 479; Sobel 1993. See also CBS News/New York Times Poll 1986, Table 5; 1987, Table 17; Americans on contra aid 1988). However, his problem was even broader. Whether the issue was military spending, arms control, military aid and arms sales, or cooperation with the Soviet Union, by the early 1980s public opinion had turned to the left—and ahead of Reagan (Page and Shapiro 1992, 271-81; Reilly 1987, chaps. 5-6; Mayer 1992, chaps. 4 and 6). Finally, Americans did not move their general ideological preferences to the right (see, for example, Fleishman 1986; Wattenberg 1990, 169-71; 1991, 95-101). Indeed,

Edwards / BUILDING COALITIONS  53

rather than conservative support swelling once Reagan was in the White House, there was a movement away from conservative views almost as soon as he took office (Stimson 1991, 64, 127). According to Mayer (1992), “Whatever Ronald Reagan’s skills as a communicator, one ability he clearly did not possess was the capacity to induce lasting changes in American policy preferences” (p. 127). Thus, Ronald Reagan was less a public relations phenomenon than the conventional wisdom indicates. He had the good fortune to take office on the crest of a compatible wave of public opinion, and he effectively exploited the opportunity the voters had handed him. Yet, when it came time to change public opinion or mobilize it on his behalf, he typically met with failure. As press secretary Marlin Fitzwater put it, “Reagan would go out on the stump, draw huge throngs and convert no one at all” (Apple 1990; see also Welch 1997). Bill Clinton President Clinton is an articulate and energetic leader whose core governing strategy is based on obtaining public support (Edwards 1995). Yet, he has not been able to move the public on behalf of his programs. For purposes of illustration, let us focus on his first two years in office. When the president’s first major economic proposal, the fiscal stimulus plan, was introduced, it ran into strong Republican opposition. During the April 1993 congressional recess, Clinton stepped up his rhetoric on his bill, counting on a groundswell of public opinion to pressure moderate Republicans into ending the filibuster on the bill. (Republicans, meanwhile, kept up a steady flow of sound bites linking the president’s package with wasteful spending and Clinton’s proposed tax increase.) The groundswell never materialized, and the Republicans found little support for any new spending in their home states. Instead, they found their constituents railing against new taxes and spending. The bill never came to a vote in the Senate (Healey 1993, 1002-3). The president’s next major legislative battle was over the budget. On August 3, 1993, he spoke on national television on behalf of his budget proposal, and Senate Republican leader Robert Dole spoke against the plan. A CNN overnight poll following the president’s speech found that support for his budget plan dropped (Woodward 1994, 285).3 Several million calls were made to Congress in response to Clinton and Dole, and the callers overwhelmingly opposed the president’s plan (Switchboards swamped 1993, A18). The White House had more success on the North American Free Trade Agreement (NAFTA) when Vice President Al Gore debated Ross Perot on the Larry King Live television show. Among those who watched the debate, support for NAFTA increased and opposition decreased. Even here, however, the White House’s ability to move public opinion was limited: Gallup polls taken before the House vote showed that only 38 percent of the public favored the trade agreement. When the crucial rule regarding debate on the 1994 crime bill was voted down in the House, the president immediately went public. Speaking to police officers with flags in 3. A CBS News\New York Times poll with before-and-after samples on August 2 and 3 found that support for the president’s budget remained unchanged even in the immediate aftermath of the speech but that opposition weakened.

54  PRESIDENTIAL STUDIES QUARTERLY / March 2000

the background, he blamed special interests (the National Rifle Association) and Republicans for a “procedural trick,” but his appeal did not catch fire. Meanwhile, Republicans were talking about pork barrel spending, tapping public resentment. Clinton’s public push yielded only the votes of three members of the Black Caucus. So he had to go to moderate Republicans and cut private deals. President Clinton’s greatest disappointment was his inability to sustain the support of the public for health care reform, despite substantial efforts. Once again, the White House hoped for a groundswell of public support, but it never occurred (Waldman and Cohn, 1994, 32). Indeed, by mid-August 1994, only 39 percent of the public favored the Democratic health care reform proposals, while 48 percent opposed them (Gallup Poll 1994). Despite energetic efforts, then, the White House was not able to produce groundswells of support for the economic stimulus plan, nor for the budget deal, the crime bill, or for health care reform, and it only did a little better on NAFTA. The president’s own approval levels averaged less than 50 percent for each of his first two years in office. In 1994, an association with Clinton decreased votes for Democratic candidates for Congress, and the election was widely seen as a repudiation of the president (Jacobson 1994; Brady, Cogan, and Rivers 1995). In addition, the president’s public relations efforts did not go unanswered. In 1993 alone, President Clinton’s opponents ran television commercials on NAFTA, the balanced budget amendment, tort reform, and health care reform. In sum, contemporary presidents typically find the public unresponsive to their pleas for support. Given the critical role that public support plays in the governing strategies of presidents, their frustrations in obtaining public support pose obstacles to their ability to advance their agendas.

The Causes of Difficulties in Leading the Public We have seen that presidents frequently face difficulties in leading the public. The question for us is whether these difficulties are the result of the nature of the presidency as an institution or whether their roots lie deeper in the nature of democratic leadership itself. There are two ways to proceed to answer this question. First, we can compare the experiences of presidents in attempting to lead the public with those of leaders in parliamentary democracies. Second, we can examine the obstacles to leading the public to determine whether they are related to the institutional presidency. President or Prime Minister? Data regarding public leadership in other countries are not abundant. Fortunately, we can examine the impact of Margaret Thatcher, another strong conservative leader holding office at about the same time as Ronald Reagan, on public opinion in Britain. We find that the prime minister’s experience was similar to Reagan’s. In a series of studies, Crewe has analyzed the support for Thatcherite values, policy beliefs, and leadership style using opinion polls by MORI, Gallup, and the British Election Surveys (1970 to 1983). He

Edwards / BUILDING COALITIONS  55

concluded that with the exception of privatization, there was no evidence that Prime Minister Thatcher converted the electorate on the central values of strong government, discipline, and free enterprise during her first term (Crewe, 1989; see also Crewe 1988; Crewe and Searing 1988a). In addition, there was no increase in the Conservative vote, partisanship, or party members in the 1980s (Crewe and Searing, 1988b). Rentoul (1989) and Curtice (1986a) report similar findings (see also Curtice 1986b). Although the results of the Reagan-Thatcher comparison by no means provide a definitive answer to the question of the source of problems presidents face in leading the public, they support the argument that the challenges of leading the public are more closely related to the nature of the audience than to the institutional nature of the messenger. Obstacles to Leading the Public To determine whether the presidency as an institution is the source of the difficulties presidents face in leading the public, we must identify the primary obstacles the president must overcome to succeed. We have much to learn about public leadership, but there appear to be four principal obstacles: the public’s awareness, understanding, and acceptance of the president’s message and their retention of it. Awareness

Although the president has unequalled opportunities to reach the public, we should not assume that all, or even most, of the public is attentive to the president’s messages. This is especially true in an age of low interest in politics and one in which cable television provides easy access to such a large array of alternatives to news programs. Indeed, we know that the size of the audience for televised presidential speeches has declined over time (see, for example, Baum and Kernell 1999). Even President Bush’s triumphal speech to a joint session of Congress following the success of the Gulf War attracted only a minority of those watching television at the time (Kernell 1997, 131-32). Television is a medium in which visual interest, action, and conflict are most effective in attracting audiences; and presidential speeches rarely contain these characteristics. The public’s general lack of interest in politics constrains the president’s leadership of public opinion in the long run as well as on a given day. Although they have unparalleled access to the American people, presidents cannot make too much use of it. If they do, their speeches will become commonplace and lose their drama and interest. That is why presidents do not make appeals to the public, particularly on television, very often—four or five times a year on average (Brace and Hinckley 1993, 387). Understanding

For the president to influence public opinion, members of the public must not only receive his message, they must also understand it. First, people must understand the position the president would like them to adopt, whether it is support for himself or his policies. We cannot assume that the public understands the messages sent by the White House, however. Those who are not aware of a message cannot be said to understand it.

56  PRESIDENTIAL STUDIES QUARTERLY / March 2000

An inattentive public is unlikely to be conversant with the president’s pleas. Even those who pay attention to the president’s message may miss the president’s point because they lack the background to understand what the president is attempting to communicate or because they selectively perceive what the president is saying. Acceptance

Presidential messages are not processed by open minds. Individuals have orientations that are the result of their background, experiences, and interests. The orientations of interest to us are people’s political values and attitudes, commonly referred to as political predispositions. The significance of political predispositions for presidential public leadership arises from two factors. First, they are relatively stable and typically not subject to short-term change. Although the cumulative effects of short-term changes in opinion may ultimately alter predispositions, the president or other elites are unlikely to have a significant immediate impact on people’s predispositions. Thus, the second important factor about individuals’ predispositions of interest to us is that predispositions mediate the impact of the political messages they encounter. If a message is consistent with a person’s predispositions, the argument should reinforce those predispositions. If an argument is inconsistent with a person’s predispositions and the person recognizes this inconsistency, then the person is likely to resist accepting the argument. As a result, predispositions may exert a significant influence on the public’s acceptance or resistance of the president’s efforts to lead them. Those most attentive to the president are also those most likely to hold strong opinions, which are in turn the most difficult views to change. Those least likely to have views, and thus those most susceptible to persuasion, are also those least likely to be exposed to the president’s messages (Zaller 1992). Retention

An important but generally overlooked question regarding opinion leadership is the public’s retention of its views. Rarely is the president interested only in affecting attitudes for only a few days. If, for example, the White House is going to provide an electoral incentive for members of Congress to support the president’s policies, the public must remember the issue on election day. Presidential efforts at persuading the public, then, usually involve more than one speech. Most observers argue that to be effective, the president must focus the public’s attention on his policies for a sustained period of time (e.g., Edwards 1995, 1999). Sustaining such a focus is very difficult to do, however, as there are many competing demands on the president to act and speak. The Reagan White House was successful in maintaining a focus on its top-priority economic policies in 1981. It molded its communication strategy around its legislative priorities and focused the administration’s agenda and statements on economic policy to ensure that the president’s message was not diffused by discussing a wide range of topics (Hertsgaard 1988, 107-8; Speakes 1988, 301). Yet, after 1981 President Reagan had to deal with a wide range of noneconomic policies. The Clinton administration immediately blurred its focus by raising a wide range of issues soon after taking office and has never been able to focus public attention (Edwards 1999).

Edwards / BUILDING COALITIONS  57 Interaction between Understanding and Acceptance

A vital but little understood aspect of presidential persuasion is the interaction between understanding and acceptance. Although members of the president’s audience will interpret his message through the filter of their own experience, values, and knowledge of politics, individuals typically have at least two, and often more, relevant values for evaluating the president’s positions. For predispositions to affect the acceptance of a message, a person must understand that the message has implications for their predispositions. If he is to succeed in leading public opinion, the president cannot leave to chance the identification of which values are most relevant to the issues he raises. Instead, the White House must help audiences understand why his position is consistent with their values.4 The president must use metaphors and other techniques to try to frame issues in terms of some values to make them more salient to his audience. Much of politics revolves around trying to structure choices for the public. Other Obstacles

Other variables also may affect the success of presidents in building coalitions among the public, including the delivery of speeches. Among post–World War II presidents, only Kennedy, Reagan, and Clinton have mastered the art of speaking to the camera. The substance of the president’s message may be important, but it is not clear what approach works best. Many of the most effective speeches seem to be those whose goals are general support and image building rather than specific support. They focus on simple themes rather than complex details. Calvin Coolidge used this method successfully in his radio speeches, as did Franklin D. Roosevelt in his famous “fireside chats.” One scholar counted only four times that Roosevelt used a fireside chat to discuss legislation under consideration in Congress (Cornwell 1965, 263). The limitation of such an approach, of course, is that general support cannot always be translated into public backing for specific policies. On a broader scale, Skowronek (1993) argues that the key to understanding presidential leadership is the success of presidents in imposing an authoritative definition on their respective historical situations. Their success is dependent on the context in which they function: (1) whether they are affiliated with or opposed to the old regime and the ideology and interest embodied in preexisting institutional arrangements and (2) whether governing commitments embodied in previously established institutional arrangements are resilient or vulnerable, constraining or providing opportunities for leaders. In their efforts to lead, presidents undermine the status quo ante, creating a cycle of generation and degeneration of governing regimes. I cannot do justice to this sweeping and rich analysis here, but we can note that many presidents have failed in their attempts to control the political definition of their actions. Part of the reason for this failure is the historical context in which presidents find themselves, a context that is beyond their control. Reagan’s conservative regime cast a long 4. There is little chance that the president can succeed in persuading people to change their values.

58  PRESIDENTIAL STUDIES QUARTERLY / March 2000

shadow over the Bush presidency, and the apparent repudiation of liberalism in the 1980s erected significant constraints on Clinton’s ability to create a new regime based on a commitment to a substantially larger role for the state.

The Institutional Presidency and Obstacles to Public Leadership There are many obstacles to presidential leadership of the public, but they share the characteristic of not being directly related to the institution of the presidency. A prime minister or chancellor faces the same challenges as the president. All face problems in obtaining the public’s attention and its understanding, acceptance, and retention of their messages. Chief executives also face common problems of effectively communicating their messages. The fact that the president, unlike some chief executives, is also head of state seems to add little to the president’s persuasive resources. We cannot conclude, however, that the institution of the presidency is unrelated to coalition building among the public. Although the nature of the presidency as an institution does not inhibit presidents from obtaining support, the institution of the presidency denies the president the power to act alone. The necessity to build coalitions forces the president into more dependence on public support than chief executives in most other systems. Presidential dependence on popular support is not a new phenomenon. Henry Jones Ford (1898), one of the leading political scientists of the nineteenth century, argued that the strength of the presidency was based on the fact that the convention system of nomination and the alteration in the role of the Electoral College had turned the presidency into a representative institution. The popular mandate recorded by the presidential election made the president an instrument of popular control. We now know, of course, that the president’s dependence on popular support is no guarantee of obtaining the public’s backing. Thus, the White House must continuously court the public to govern. The way presidents attempt to govern has profound consequences for public policy. When political leaders take their cases directly to the public, they have to accommodate the limited attention spans of the public and the availability of space on television. As a result, choices are typically reduced to stark black-and-white terms. Such positions are difficult to compromise, which hardens negotiating positions as both sides posture as much to mobilize an intense minority of supporters as to convince the other side. Moreover, unlike bargaining, which provides benefits for both sides, the strategy of going public is designed to intimidate opponents by increasing the political costs of opposition. In the end, polarization, gridlock, and public cynicism, which characterize American politics today, are the likely results. Whether successful or not, the president’s efforts to build coalitions among the public are designed to assist building coalitions in Congress. In most cases, Congress holds the ultimate power to make public policy in the United States. Thus, coalition building in Congress is one of the president’s principal responsibilities.

Edwards / BUILDING COALITIONS  59

Coalition Building in Congress Perhaps the central feature of the U.S. constitutional system is the horizontal fragmentation of power encapsulated by the phrases “separation of powers” and “checks and balances.” Because the power to pass legislation is much more likely to be shared than separated, it is the checks and balances that are most relevant to building coalitions. The White House must build coalitions in Congress because the president generally cannot act without Congress’s consent. Under the U.S. constitutional system, Congress must pass legislation and can override vetoes. The Senate must ratify treaties and confirm presidential nominations to the cabinet, the federal courts, regulatory commissions, and other high offices. Agenda Setting Attaining agenda status is a necessary prelude to the passage of a bill, and thus obtaining agenda space for his most important proposals is at the core of every president’s legislative strategy. Agenda setting is the focus of another article in this issue, so I will treat it only briefly. The burdens of leadership are considerably less at the agenda stage than at the floor stage, at which the president must try to influence decisions regarding the political and substantive merits of a policy. At the agenda stage, in contrast, the president only has to convince members that his proposals are important enough to warrant attention. The White House employs three principal sources of influence on behalf of agenda setting—service, incentives, and persuasion—and generally succeeds. Over the 1953 to 1996 period, the president obtained agenda status for 97.6 percent of his significant legislative initiatives (Edwards and Barrett 2000). Thus, the agenda-setting stage of the legislative process rarely poses an insurmountable barrier to the president (see, however, Edwards and Wood 1999). Between the Agenda and the Floor Once on the agenda, there are several important steps in the legislative process through which presidential proposals must pass, including subcommittee and committee consideration, the rules for floor debate, and voting on the floor. We have systematic data on the president’s success only at the floor stage. We do know, however, that the president’s proposals are frequently altered substantially in committee and that the opposition party, when in the majority, often uses its rule-making powers to disadvantage the president’s proposals on the floor. Support on the Floor There is no universally agreed on measure of the president’s success on floor votes in Congress. However, we can examine a series of indicators, each providing a somewhat

60  PRESIDENTIAL STUDIES QUARTERLY / March 2000

different view of presidential success, to determine whether the president has a problem obtaining support in the legislature. Winning Votes

One measure of presidential success is the percentage of votes on which the White House has taken a stand and the president’s position prevailed. Presidents typically win most of the votes on which they take a stand, although there is substantial variance over time. For example, Bill Clinton won on 86 percent of the votes on which he took a stand in 1993-94, but his winning percentage fell to only 36 percent in 1995 (see, for example, Congressional Quarterly 1999). Calculating the percentage of votes won does not differentiate among issues; instead, it weighs each one equally and may provide a distorted view of the president’s record. Thus, despite Bill Clinton’s high victory rate in 1993-94, he failed to secure passage of many of his high-priority proposals such as the fiscal stimulus bill and health care reform. To obtain a more accurate picture of the president’s success in building coalitions, we must examine other measures of the executive-legislative relations. Passage of Significant Legislation

A useful indicator of the president’s ability to build coalitions is the White House’s success in obtaining passage of its potentially most significant legislative proposals. Table 1 shows that there were 287 presidential initiatives of potentially significant legislation over the 1953 to 1996 period. Of these, 41 percent became law. In short, presidents fail most of the time. These data do not tell us whether other presidential initiatives were delayed or diluted by opposition in Congress, but they do indicate that presidents, unlike prime ministers, frequently fail to obtain passage of their initiatives. Congress is quite capable of taking its own initiatives. In the 1953 to 1996 period, 66 percent of the potentially significant legislation that made it onto the congressional agenda were congressional initiatives. When the White House supports these initiatives, as it did nearly half the time over the 1953 to 1996 period, the president’s position prevailed 93 percent of the time. We should be cautious in interpreting this figure as evidence of strength in coalition building, however. The opposition party (when in the majority) generates most congressional initiatives. If the president supports such an initiative, there is unlikely to be any substantial opposition to overcome. In addition, the congressional leadership takes responsibility for building coalitions on congressional initiatives. Opposition to Significant Legislation

Congressional initiatives often come from the opposition party, especially during periods of divided government. Unsurprisingly, the president often seeks to defeat these initiatives. Over the 1953 to 1996 period, the White House opposed slightly more than half of the 545 congressional initiatives of potentially significant legislation on the congressional agenda. The president succeeded in killing all but 8 of them.

Edwards / BUILDING COALITIONS  61 TABLE 1 Presidential Initiatives that Became Law Years 1953-54 1955-56 1957-58 1959-60 1961-62 1963-64 1965-66 1967-68 1969-70 1971-72 1973-74 1975-76 1977-78 1979-80 1981-82 1983-84 1985-86 1987-88 1989-90 1991-92 1993-94 1995-96 Total Unified Divided

Number of Initiatives 9 6 11 4 24 22 30 19 16 18 17 19 19 17 12 10 5 2 8 4 15 0 287 155 132

Number that Became Law 5 2 7 0 13 9 17 13 4 4 5 2 7 9 3 0 1 1 3 3 9 0 117 82 35

Percentage that Became Law 56 33 64 0 54 41 57 68 25 22 29 11 37 53 25 0 20 50 38 75 60 0 41 53 27

The threshold for a winning coalition is lower when the president’s goal is to kill legislation. All the White House requires is one vote more than 33 percent in one chamber of Congress to sustain a veto. And the veto plays a crucial role in the president’s success in opposing potentially significant congressional initiatives. A substantial portion of the potentially significant legislation that failed to pass failed as the result of vetoes (Edwards, Barrett, and Peake 1997). In addition, a number of other bills also may have failed to pass in one or both houses of Congress because the threat of a presidential veto discouraged the investment of time, energy, and political capital necessary to obtain passage. Normally, we view the veto as a negative weapon, one to wield in opposition to winning coalitions rather than on behalf of building them. Yet, we know that Bill Clinton used the veto during the 1995-96 government shutdown to create the perception of the culpability of the Republican Congress. By structuring the choice before the public, the president was able to leverage public support to advance his policy goals. Republican fear of a recurrence of this predicament helped the president obtain more than he otherwise would have from Congress at the end of the 1998 session. Ironically, the veto has proven to be Clinton’s most successful tool in his efforts to woo the public.

62  PRESIDENTIAL STUDIES QUARTERLY / March 2000 TABLE 2 Partisan Support for Presidents, 1953 to 1996 (in percentages)

Members of Congress Democrats Republicans Difference

House

Senate

President Democratic Republican

President Democratic Republican

70 28 42

35 67 32

66 35 31

33 70 37

Source: Edwards and Wayne (1999). Note: On roll calls on which the president has taken a stand and on which the winning side was supported by fewer than 80 percent of those voting.

Support from Groups

Most winning coalitions in Congress have one of the parties at their core. The president is the leader of his party and depends heavily on it to pass his initiatives and stop legislation to which he is opposed. How much support does the president typically receive from his party? Table 2 shows that on contested votes on which the president has taken a stand, the president obtains the support of approximately two-thirds of his copartisans, twice the level he receives from members of the opposition party. On one hand, the president can depend on the support of most members of his party most of the time. On the other hand, there is plenty of slippage in party support, and the opposition party opposes him most of the time. If the opposition party is in the majority, which it frequently is, the odds are against the president building a winning coalition.

Obstacles to Building Coalitions in Congress Presidents frequently fail in their efforts to build coalitions in Congress. There are many potential causes for such failures. For example, the president and some members of Congress will always disagree because of their personalities or past histories. Yet, these differences are not the source of systematic conflict between the branches. In this section, I examine the principal obstacles to the president’s success in coalition building. In the following section, I analyze the contribution that the nature of the presidency as an institution makes to the occurrence of these obstacles. Checks and Balances The system of checks and balances is designed to produce sound, moderate legislation through a process of negotiation and compromise that accommodates minority viewpoints. The open and deliberative nature of the process is to confer legitimacy on the legislation that it generates. At the same time, checks and balances complicate coalition building. Indeed, the necessity of congressional support forces the president to build

Edwards / BUILDING COALITIONS  63

coalitions in the first place. The bicameral structure of Congress further complicates the process by requiring the president to build not one but two coalitions from among quite different sets of representatives. In addition, the requirement that the Senate ratify treaties by a two-thirds vote is a structural provision that increases the burden of coalition building because it forces the president to achieve a supermajority to achieve ratification. Checks and balances alone do not explain the president’s challenges in forming supportive coalitions in Congress. Theoretically, the two branches could be in agreement. However, checks and balances provide the context within which other potentially divisive factors may become obstacles to coalition building. For example, the Senate’s rules, especially those regarding debate, protect minority interests and force advocates of change to build coalitions of at least 61 percent of the members. Party Opposition The most important resource the president can have in building coalitions is like-minded members of Congress (Edwards 1989; Bond and Fleisher 1990). Such members are most likely to be found among those in the president’s party. We have seen that the president receives twice as much support on the average from members of his party as from members of the opposition party. Presidents have had little success in exercising legislative skills and systematically changing the minds of many senators and representatives as legislation comes to the floor (Edwards 1989, chap. 9; Bond and Fleisher 1990, chap. 8). As a result, the president is largely dependent on the cards voters have dealt him in previous elections. There are a number of methods of electing chief executives and members of the legislature. The most common is a parliamentary system in which the chief executive is elected from a single legislative district during the general elections for parliament. Thus, the members of the legislature and the chief executive are elected at the same time. This simultaneous election encourages voters across the nation to support the leader’s party by voting for candidates of the leader’s party. Since the prime minister is selected by a majority of the legislature, the prime minister’s party or party coalition must have the support of a majority in the legislature. In the United States, voters cast their votes separately for executive and legislative officials, who have terms of different lengths. Thus, they may split their votes between candidates of different parties. Not supporting the president’s party is even easier in midterm elections, when the president is not on the ballot. In addition, one-third of the Senate is not elected in any election during a president’s four-year term. The result is often divided control of the executive and legislative branches, which has occurred nearly two-thirds of the time in the past half century and in all but one of the national elections since 1978. Divided government has important consequences for the president’s policies. Under unified government, the president succeeds in obtaining passage of 53 percent of his significant legislative proposals (Edwards and Barrett 2000). Under the less sanguine conditions of divided government, however, the success rate for presidential initiatives is cut nearly in half, falling to 27 percent because of the opposition of the majority party in

64  PRESIDENTIAL STUDIES QUARTERLY / March 2000

Congress. Divided government matters (Edwards, Barrett, and Peake 1997). Moreover, divided government exacerbates the tendency stemming from the Senate’s filibuster rules to delay or dilute bills that eventually pass. Intraparty Diversity The president faces a second party-related obstacle to building coalitions: the diversity of policy preferences within his own party. In a large and diverse country with a two-party system, it is not surprising that representatives of each of the two parties reflect a range of constituents’ policy positions. This diversity inevitably poses a challenge to intraparty cohesion. The system of primaries for selecting congressional candidates undermines at a minimum the ability of party leaders to control who runs under their party’s label and thus weakens their ability to discipline errant members for not supporting the president. Most members of Congress gain their party’s nomination by their own efforts, not the party’s. Because virtually anyone can vote in party primaries, party leaders do not have control over those who run under their parties’ labels. Moreover, even though national party organizations have been active in fund-raising in recent years, candidates remain largely responsible for providing the money and organization for their own election, precluding party control over another aspect of electoral politics. The relative independence of presidential and congressional elections is illustrated by the modest number of coattail victories in which presidential coattail votes provide the increment of the vote necessary to win a seat for a representative of the president’s party. Such victories may provide the president an extra increment of support out of a sense of gratitude for the votes winners of congressional races perceive were received due to presidential coattails or out of a sense of responsiveness to their constituents’ support for the president. However, the outcomes of very few congressional races are determined by presidential coattails (Edwards 1983, 83-93; Flemming 1995). For example, in 1988 George Bush won election while his party actually lost seats in both houses of Congress. Similarly, the Democrats lost ten seats in the House and gained none in the Senate when Bill Clinton won election in 1992, and they lost two seats in the Senate when he won reelection in 1996. (Clinton ran behind all but a handful of members of Congress in their states or districts in both elections.) This is nothing new: in 1792, George Washington easily won reelection, but the opposition Democrat-Republicans captured the House of Representatives. Most House seats are too safe for a party, and especially for an incumbent, to have the election outcome affected by the presidential election. Senate elections are more affected by the president’s standing with the public (Abramowitz and Segal 1992, 121, 233, 238; Atkeson and Partin 1995; Campbell and Sumners 1990), but the president’s party typically gains no seats at all in a presidential election year. Modern presidents have tried to increase the size of their party cohort in Congress and encourage party cohesion by taking an active role in midterm congressional elections. Typically, however, they are disappointed in the results of their efforts (see Cohen, Krassa, and Hamman 1991).

Edwards / BUILDING COALITIONS  65

Party leaders also have few ways to enforce party discipline among those who are elected. What sanctions might be applied, such as poor committee assignments, are rarely used because legislators are very hesitant to set precedents that could be used against them. The strong centralization of party leadership in the House under Newt Gingrich soon gave way to the more collegial leadership of Dennis Hastert. The finding in Table 2 that the typical member of the president’s party fails to support him about one-third of the time is not surprising. The independent tenures of the president and members of Congress also diminish cohesion in the president’s party. In a parliamentary system, the government falls if the prime minister loses the support of the legislature. Typically, such a loss of support leads to new elections for the entire legislature. Since facing the electorate under circumstances of party disunity is usually not in the interests of the prime minister’s party, its members have an incentive to support their leader (Epstein 1967). There is no such incentive in the United States, however. Members of Congress retain their jobs (at least in the short run) independently of the president’s legislative success. In recent years, congressional parties have become more ideologically homogeneous and, as a result, more cohesive (Rohde 1991, Aldrich and Rohde 2000). This change has not advantaged presidents, however. Presidents have a more difficult time obtaining votes from the opposition party as parties become more polarized. When their parties are in the minority in Congress (as they were for Reagan’s, Bush’s, and most of Clinton’s tenures), polarization makes it more difficult to prevail on votes. In addition, with the exception of Ronald Reagan, winning candidates in the past three decades have positioned themselves as more moderate than their congressional parties. At the same time that winning presidential candidates have moved to the center, their congressional cohorts have become more polarized. As the number of conservative Democrats and liberal Republicans has diminished, there has been less pressure to compromise within the party caucus. The inevitable tension between centrist presidents and polarized party caucuses has meant that party support for the president has not increased in conjunction with party homogeneity (Fleisher and Bond 2000). Public Support Earlier sections of this article focused on the central role that obtaining public support plays in the modern presidency. Presidents believe that they must obtain public support for themselves and their policies to build coalitions in Congress (see Edwards 1997). We also found, however, that presidents’ attempts to govern by “going public” frequently fail. Because we focused on leading the public in first sections of this article, I will address other aspects of public support here. Mandates

An electoral mandate—the perception that the voters strongly support the president’s character and policies—can be a powerful symbol in American politics. It accords added legitimacy and credibility to a newly elected president’s legislative proposals. More-

66  PRESIDENTIAL STUDIES QUARTERLY / March 2000

over, concerns for both representation and political survival encourage members of Congress to support new presidents if they feel the people have spoken. As a result, the perception of a mandate can alter the premises of congressional decision making and substantially increase the president’s success in building coalitions (Edwards 1989, chap. 8). The Constitution’s requirement that the entire nation participate in electing the president provides the White House the potential to claim a public mandate for the president’s policies. The fact that presidents are elected through the mechanism of the Electoral College adds the potential for creating the impression that a plurality or narrow majority victory is actually larger if the presidential vote is dispersed uniformly across states. However, there is no evidence that a larger electoral than popular vote leaves a lasting mark on perceptions of the president’s victory. The Electoral College also creates the potential (last realized in 1888) to elect a president who lacks even a plurality of the vote. Such an event surely undercuts a president’s ability to claim public support, much less a mandate to govern. Constituency Support

Presidents typically attempt to build public support broadly throughout the country. For public support of the president or his policies to serve as a source of influence for the president in building coalitions in Congress, members of Congress must be responsive to that broad opinion. However, the same electoral system that provides the president with the potential for a mandate also ensures that the White House and each member of the legislature will interpret public opinion differently. James Madison anticipated this when he wrote in The Federalist no. 46, The members of the federal legislature will likely attach themselves too much to local objects. Measures will too often be decided according to their probable effect, not on the national prosperity and happiness, but on the prejudices, interests, and pursuits of the governments and the people of the individual states. (Hamilton, Madison, and Jay 1938)

With these prescient words, Madison focused on the greatest obstacle to presidential coalition building in Congress: the different constituencies of the two branches. Only the president (and his vice presidential running mate) is elected by the entire nation. Each member of Congress is elected by only a fraction of the populace. Inevitably, the president must form a broader electoral coalition to win his office than does any member of Congress. Moreover, two-thirds of the senators are not elected at the same time as the president, and the remaining senators and all the House members seem to be substantially insulated from the causes of presidential victories, leading to a large number of districts supporting a president of one party and a member of the House from the other party (Ornstein, Mann, and Malbin 1998). In addition, the Senate overrepresents rural states because each state has two senators regardless of its population.5 Thus, the whole that the president represents is different from the sum of the parts that each legisla5. I use the term overrepresents because each person living in a rural state has more influence on the election of a senator by virtue of being one of a small number of voters. Thus, rural voters have more representatives per person than urban voters.

Edwards / BUILDING COALITIONS  67

tor represents. Each member of Congress accords special access to the interests that he or she represents. Some members of Congress make responsiveness to narrow interests a firm decision rule (Fenno 1979, chap. 3). Sometimes the decision rule may be less clear, but the results are the same. From military bases to rural post offices, from highway funds to school aid, and from foreign aid to foreign trade, members of Congress respond to the parochial needs of their constituencies (even Hillary Clinton challenged some of her husband’s policies toward New York in anticipation of becoming a Senate candidate). In the process, they often find themselves opposing the president’s efforts to limit expenditures, make expenditures more effective, or develop policies that advance broad, long-term interests of the nation. The concentrated benefits and dispersed costs provide significant incentives for supporting these projects (just as logrolling provides incentives for members to support each other’s projects) and little incentive to oppose them (see, for example, Ferejohn 1974). Members of Congress are aware that, given the chance, their colleagues will appeal to their constituents with policies that would work against the common good. Therefore, Congress has made some attempts to limit responsiveness to narrow segments of public opinion. For instance, Congress has removed itself from making specific tariff decisions to avoid making policies that historically have had disastrous consequences for the American economy. Congress delegates many other decisions, including military base closings and regulations, to officials in less visible parts of the government system, such as executive agencies and regulatory commissions. Similarly, members of some crucial House committees are chosen in part for their relative insulation from public opinion, and once bills reach the floor, the House leadership often employs closed rules to prevent parochial amendments. Of course, members of Congress are not, at least not usually, merely parochial and selfish representatives of special interests, with no concern for the general welfare. Moreover, presidential policies may be ill considered or be designed primarily to benefit the president’s electoral coalition. Whatever the policies’ inherent merits, the structure of American government exerts strong pressure on the two branches to represent different sets of interests. The president and Congress’s differing conceptions of constituency are an important obstacle to presidential coalition building in Congress. Public Accountability

If the president is to most effectively employ public support to build coalitions, the public must hold members of Congress accountable for their policy stances. The executive branch of government is a hierarchy, with the president at the pinnacle. It is difficult for the White House to evade responsibility for the entire branch, even if the president took no direct action on an issue. Moreover, when the president exercises power, it is clear who is acting and who should be held accountable. Congress, on the other hand, is highly decentralized, and each member is relatively obscure compared with the president. Moreover, Congress is not responsible for implementing policies. As a result, members of Congress can disclaim responsibility for policies or their consequences.

68  PRESIDENTIAL STUDIES QUARTERLY / March 2000

The public seems content with this arrangement and does not seem to hold members of Congress accountable for their collective decisions (Fenno 1979; Parker and Davidson 1979). Thus, members of Congress often run for reelection by campaigning against Congress (Fenno 1977, 914). Members also frequently articulate vague views on controversial issues and deal with controversial issues in Congress indirectly and procedurally, following the dictum that “no one ever got defeated for something he didn’t say.” The differences in the accountability of the president and members of Congress to the public present another obstacle to presidential coalition building. They provide the potential for representatives and senators to make irresponsible or self-serving decisions and then let the president take the blame. Those members of Congress who are more interested in the immediate distribution of the benefits of policies to their constituents than in the ultimate consequences of policies will be especially tempted to do so. The short-run results please the interested parties and easily can be attributed to congressional actions; long-term effects are more obscure, as are their causes. Thus, members of Congress gain points for reelection while evading accountability for their actions. Electoral Reprisals

The ultimate incentive for members of Congress to be responsive to the president’s public support is the fear of electoral reprisals if they fail to do so. The Constitution limits the president to a maximum of ten years in office and eight years if he does not complete the unfinished term of a predecessor. In contrast, there is no limit to the tenures of members of Congress. Political observers often claim that a president in his second term, a “lame duck,” is weakened in his ability to deal with Congress. The principal reason offered for this weakness is that the president will never again be on the ballot, and therefore members of Congress need not fear public retaliation for opposing his policies. The line of reasoning underlying this conclusion is not clear, and there is no theoretical or systematic empirical support for it (see, for example, Edwards 1991; Sullivan 1991; Sundquist 1992, 176). This feature of the institutional presidency seems to have little effect on coalition building. Time Perspectives The differences in the length of terms of presidents and members of Congress encourage them to adopt different time perspectives. Presidents fear that their mandates (most presidents feel they obtained one in their election) are short-lived, and they know their tenures will be short. Thus, they can waste no time in pushing for the adoption of their policies. (Presidents, of course, can procrastinate in proposing or reacting to others’ solutions to national problems, and they may focus on short-term political gains. The issue is one of institutional tendencies.) Congress has a different timetable. Its members tend to be careerists and therefore do not have the same compulsion to enact policies rapidly. This sluggish approach is aggravated by the decentralization of Congress, which ensures that a great deal of negotiating and compromising must take place on all but a few noncontroversial (and usually unimportant) issues. This process can, and often does, take years. President Nixon pro-

Edwards / BUILDING COALITIONS  69

posed revenue sharing in 1969; it passed in 1972. President Truman proposed a national health plan in 1948; a limited version (Medicare) was passed in 1965. One consequence of Congress’s frequent sluggishness in handling legislation is that the president is not likely to get much of what he wants now until later, if at all. A second consequence is that presidential policies may be passed too late to become fully effective. At the same time that differences in tenures encourage the president and Congress to process legislation at different speeds, they also invite them to adopt different time perspectives on policy issues. A president, especially one in his second term, may choose to tackle long-term issues such as Social Security financing or tax reform. They are more worried about their legacies than about providing short-term benefits to voters that will serve electoral needs. On the other hand, most members of Congress, especially those in the House, are constantly facing election. As one representative put it during the Reagan presidency, “My neck’s on the chopping block—not Reagan’s. He can talk about longer-term solutions to interests rates and unemployment. I can’t. I need something to tell my people now” (Ornstein 1982, 102-3). Bases for Decision Another obstacle to presidential coalition building in Congress is the different internal structures of the executive and legislative branches. The executive branch is hierarchically organized, facilitating the president’s examining a broad range of viewpoints on an issue and then weighing and balancing various interests. This structure also helps the president to view the trade-offs among various policies. Since one person, the president, must support all the major policies emanating from the executive branch, he is virtually forced to take a comprehensive view of those policies. Members of Congress frequently do not take such a broad view. Unlike in a parliamentary system, members of Congress are prohibited from serving in the executive branch. In addition, each house of Congress is highly decentralized, with each member jealously guarding his or her independence and power. The structure of Congress ensures that a diversity of views will be heard and that many interests will have access to the legislative process. Decentralizing power and responsibility in Congress also allows for specialization in various policy areas. Congressional decentralization and specialization do not guarantee, however, that each member will hear all the views and see the proponents of each interest. Indeed, the decentralization of Congress almost guarantees that the information available to it as a whole is not a synthesis of the information available to each legislator. The Congress as a whole does not ask questions—individual members do. Thus, not all members receive the answers. Because of specialization, legislators make decisions about many of the policies with which Congress must deal in form only. In actuality, they tend to rely on the cues of party leaders, state party delegations, relevant committee leaders of their party, and other colleagues to decide how to vote (see, for example, Matthews and Stimson 1975; Kingdon 1989). The different structures of the executive and legislature would not necessarily lead to a divergence of viewpoints between the president and Congress if the people who

70  PRESIDENTIAL STUDIES QUARTERLY / March 2000

influenced the votes of each legislator represented the full range of views in the chamber, but they do not. Although the findings are not clear-cut, it appears that committees attract members from constituencies having special interests in subject areas of the committee or who have ideological interests of their own to promote (Shepsle 1978; Hall and Grofman 1990; Snyder 1992; Groseclose 1994; Londregan and Snyder 1994, Maltzman 1995; Adler and Lapinski 1997; Bianco 1997; Deering and Smith 1997; Peterson and Wrighton 1998. But see Krehbiel 1991, 1994). Thus, the committees’ members are frequently unrepresentative of each house. Moreover, the individuals or state party delegations who serve as cue givers are chosen because they represent constituencies or maintain ideologies that are similar to those of the member who is consulting them. They also do not represent a cross section of viewpoints. Besides not considering the full range of available views, members of Congress are not generally in a position to make trade-offs between policies. Because of its decentralization, Congress usually considers policies serially, that is, without reference to other policies. Without an integrating mechanism, members have few means by which to set and enforce priorities and to emphasize the policies with which the president is most concerned. This latter point is especially true when the opposition party controls Congress. In addition, Congress has little capability to examine two policies, such as education and health care, in relation to each other. Not knowing that giving up something on one policy will result in a greater return on another policy, members have little incentive to engage in trade-offs. The budget committees have a broader scope than other committees and are involved in making some trade-offs between policies and setting some priorities. But they deal only with direct expenditures (and then usually only with increases over past expenditures), not taxes (except for general revenue estimates), tax expenditures, treaties, regulation, or other important areas. Moreover, they only recommend general limits on spending, leaving it up to the more parochial subject-area committees to go into specifics. The House committee is also composed of temporary members whose permanent committee assignments undoubtedly limit their scope. Similarly, the decentralization of Congress limits its ability to deal comprehensively with major policy domains. Congress distributes its workload among committees, but committee jurisdictions do not usually cover entire policy areas. For example, no one congressional committee handles energy, welfare, economic stability, or national security (the last requiring a coordination of defense policy and diplomacy). Conflict with the president may occur, because the more centralized nature of the presidency encourages the White House to evaluate legislation in terms of its relationship to related issues in policy domains. Finally, the different internal structures of the president and Congress influence the amount and quality of the information available to them for decision making, further encouraging the two branches to see issues from different perspectives. Members of Congress rarely have available to them expertise of the quantity and quality that is available to the president. Members of Congress tend to hire generalists, even on committee staffs. Sometimes these individuals develop great expertise in a particular field, but more often they are only amateurs compared with their counterparts in the executive branch.

Edwards / BUILDING COALITIONS  71

Many are selected to serve legislators’ needs and desires that have little to do with policy analysis. Congress is especially at a disadvantage in national security policy, in which the president relies on classified information that is generally unavailable to Congress.

The Institutional Presidency and Obstacles to Building Congressional Coalitions Unlike the lack of relationship between the institutional presidency and coalition building in the public, the structure of the presidency (and of Congress) is at the core of the obstacles to coalition building in Congress. Checks and balances are at the heart of the constitutional system and force the president to build multiple coalitions on any bill and in some instances force him to build very broad coalitions. The rules for separate elections for separate terms create the potential for divided government, while the independence of the tenures of the president and Congress discourage intraparty unity. (The system for nominating and electing members of Congress also weakens party leadership.) The electoral system also invites the president and members of Congress to define their constituencies differently, while limitations on the president’s tenure encourage different time perspectives in the executive and legislature. The hierarchical nature of the executive, in contrast to the more decentralized legislature, highlights the president’s accountability while obscuring Congress’s and provides the president and Congress different bases for their decisions. These relationships are summarized in Table 3. The most important influences on congressional voting are party, ideology, and constituency (Edwards 1989; Bond and Fleisher 1990). These factors are largely beyond the president’s control, especially in the short run. With the exception of providing the president the potential to claim a broad mandate to govern, the institutional presidency does little to help the president turn these influences on Congress to his advantage. Although it does not contribute directly toward building coalitions, the ability to remain in office in the face of the loss of public and congressional support does provide the foundation for exercising influence in the future. Aside from the veto power, the president has few institutionalized legislative powers. He may call Congress into special sessions and adjourn it in the case of disputes between the two chambers. Both of these powers have fallen into disuse and give the president little leverage in an age of year-round congresses. The president also may give a State of the Union message and recommend legislation to Congress. As we have seen, the president’s role in setting Congress’s agenda is substantial, although it is difficult to see how the “right” to recommend legislation is at the core of it. Given the First Amendment, the right to recommend legislation is a truism. How could anyone limit it? The White House has a modest-sized institutional staff devoted to congressional relations. Although presidents have employed this staff in a variety of ways and although some operations work more effectively than others (Collier 1997; Holtzman 1970; Jones 1988; Wayne 1978), the performance of the legislative liaison office is not at the core of presidential leadership in Congress (Edwards 1980, 1989; Bond and Fleisher 1990).

72  PRESIDENTIAL STUDIES QUARTERLY / March 2000 TABLE 3 The Impact of the Structure of the Presidency (and of Congress) on Obstacles to Coalition Building in Congress • Checks and balances force the president to build multiple coalitions on any bill and supermajorities on treaties. • The rules for separate elections for separate terms create the potential for divided government. • The independence of the tenures of the president and Congress discourage intraparty unity. (The system for nominating and electing members of Congress also weakens party leadership.) • The electoral system invites the president and members of Congress to define their constituencies differently. • Limitations on the president’s tenure encourage different time perspectives in the executive and legislature. • The hierarchical nature of the executive, in contrast to the more decentralized legislature, highlights the president’s accountability while obscuring Congress’s. • The hierarchical nature of the executive, in contrast to the more decentralized legislature, provides the president and Congress different bases for their decisions.

Restructuring the Presidency The institutional presidency has limited effect on the president’s ability to build coalitions among the public but has a major impact on building coalitions in Congress. In most instances, this influence is negative, creating obstacles to coalition building. How we evaluate the adequacy of the institutional presidency to meet the demands of coalition building in the twenty-first century will be based on what we want government, and thus the president, to do. Those who see little need for federal action or who are skeptical of national policies are likely to be more satisfied with a presidency in which the White House is rarely able to rally the public and Congress to pass significant new legislation. Where, they might ask, is the evidence that there are now demands on government so different from those in previous eras that we require the president to be a more reliable coalition builder? Where is the evidence that we require new constitutional arrangements to solve the problems of governing? To evaluate the president as coalition builder, I have approached the problem of coalition building from the perspective of the president. Thus, I have focused on how well the president can make the system work in response to his demands rather than the contribution the president makes to the functioning of the system. Some, following Mayhew (1991), may conclude that the system works well enough and produces significant legislative change, even under divided government. (Such a conclusion, of course, begs the question of whether such changes meet our needs.) Jones (1994, chap. 8) prefers “balanced participation” and does not feel that we will be better off with simplifying the system to encourage presidency-centered leadership. Indeed, he warns against an excess of presidential power, such as he finds in the early Johnson and Reagan administrations when dominant presidents obtained sweeping changes that, he feels, were not necessarily

Edwards / BUILDING COALITIONS  73

wise. Moreover, the separated system may help avoid stalemate by giving the other branch the option of taking the lead on policy change. Others will rightly raise the issue of the difficulty of anticipating the full range of consequences of changing to some form of a parliamentary system of government. The implications of such changes are too far reaching to be done justice in this article, but they are likely to be substantial (why else make them?). In addition, a constitutional convention may open a Pandora’s box of changes that result in constructing yet new obstacles to coalition building. On the other hand, those who believe that there is a pressing need for significant legislation at the national level are likely to be more concerned. Advocates of change might point out that both conservatives and liberals seek major changes in policy, and they typically want the president to take the lead in bringing about this change. Ronald Reagan led the fight for some of the most fundamental changes in public policy in the past three decades, including areas as central to the national government as taxes and national defense. If the president is pivotal to achieving the goals of those on both ends of the political spectrum, the ability of the White House to build coalitions is a prime concern as we evaluate our political system. If we want to encourage policy change and if the president is a key to generating that change, we must be concerned that the president cannot reliably build winning coalitions. The question is whether we can alter the context in which the president seeks to lead in a way that will improve the White House’s ability to build coalitions without sacrificing important values, including the representative, substantive, and other benefits of Congress’s intimate participation in policy making. Our analysis of the obstacles to coalition building provides a place to start to answer this question. At the core of the president’s challenges in building coalitions is the separation of powers and checks and balances. The electoral system of separate elections for separate terms and the independent tenures of the president and members of Congress follows from the premise of separate institutions. The greatest obstacles to presidential coalition building arise from this electoral system. It creates the potential for divided government, discourages intraparty unity, and invites the president and members of Congress to define their constituencies differently. Differences in the length of terms of presidents and members of Congress and limits on the president’s tenure encourage the White House and Congress to adopt different time perspectives for considering legislation. Identifying the sources of obstacles to coalition building is one thing. Altering the sources so as to alleviate the problem is something quite different. Reinventing the presidency as a more effective coalition builder requires nothing less than a fundamental revision of our constitutional system. Eliminating the electoral impediments to coalition building by unifying the timing, length, and limits on the terms of all elected national officials would increase the probability of successful presidential coalition building. However, such changes also would assault the central premises of separation of powers and checks and balances and the theory of representation underlying the structure of the two houses of Congress. The system for nominating and electing members of Congress also weakens party leadership. It would be possible to change rules about primaries, campaign financing, and

74  PRESIDENTIAL STUDIES QUARTERLY / March 2000

party discipline in Congress without undermining the structure or underlying theory of the Constitution, but here we are talking about changing the institution of the legislature rather than the presidency. In addition, the probability of turning to a centralized system of nominations and campaigns is low. The most recent attempts to institute strong centralized leadership in Congress are no more encouraging. Some of the obstacles to coalition building that we have traced to the institutional presidency are probably unavoidable. The contrast between the bases of decision of the hierarchical executive and the more decentralized legislature would not be easy to change. Although a parliamentary system would place some members of the legislature in executive positions, it does not automatically follow that cabinet ministers would adopt a broader view. Contemporary U.S. cabinet members are frequently criticized for adopting parochial viewpoints. The nature of the electoral and party systems seems to be more influential than executive experience in affecting the bases of decision. Where does this leave us as we evaluate the suitability of the institutional presidency for the twenty-first century? (1) The president has great difficulty building coalitions for governing among the public and within Congress, (2) the institutional structure of the presidency (and Congress) is at the core of the difficulty of building coalitions in Congress, and (3) it will be very difficult to change the fundamental structure of either institution. Institutions matter, but they are difficult to alter, especially when they are structured by the world’s oldest functioning constitutional system. Given the difficulty of making the fundamental changes that are necessary to substantially improve coalition building, we could agree with Jones (1994) that it is “more important to make the separated system work well than to change systems” (p. 297). Such a conclusion begs the question of how to make the system work better, however. The reality of our predicament is that no matter how skilled the White House may be, the president is not and is unlikely to be a consistently successful coalition builder. Only the presence of contextual conditions that encourage deference to the president, such as occurred in Franklin D. Roosevelt’s first term, is likely to provide the president the opportunity to dominate the policy-making process.

References Abramowitz, Alan I., and Jeffrey A. Segal. 1992. Senate elections. Ann Arbor: University of Michigan Press. Adler, E. Scott, and John S. Lapinski. 1997. Demand-side theory and congressional committee composition: A constituency characteristics approach. American Journal of Political Science 41 (July): 895-918. Aldrich, John H., and David W. Rohde. 2000. The consequences of party organization in the House: The role of the majority and minority parties in conditional party government. In Polarized politics: Congress and the president in a partisan era, edited by Jon R. Bond and Richard Fleisher. Washington, DC: CQ Press. Americans on contra aid: Broad opposition. 1988. New York Times, January 31, sec. 4, p. 1. Anderson, Martin. 1990. Revolution: The Reagan legacy. Stanford, CA: Hoover Institution. Apple, R. W. 1990. Bush sure-footed on trail of money. New York Times, September 29, 8. Atkeson, Lonna Rae, and Randall W. Partin. 1995. Economic and referendum voting: A comparison of gubernatorial and senatorial elections. American Political Science Review 89 (March): 99-107. Bartels, Larry M. 1994. The American public’s defense spending preferences in the post-cold war era. Public Opinion Quarterly 58 (Winter): 479-508.

Edwards / BUILDING COALITIONS  75 Baum, Matthew A., and Samuel Kernell. 1999. Has cable ended the golden age of presidential television? American Political Science Review 93 (March): 99-114. Bianco, William T. 1997. Reliable source or usual suspects? Cue-taking, information transmission, and legislative committees. Journal of Politics 59 (August): 913-24. Bond, Jon R., and Richard Fleisher. 1990. The president in the legislative arena. Chicago: University of Chicago Press. Brace, Paul, and Barbara Hinckley. 1993. Presidential activities from Truman through Reagan: Timing and impact. Journal of Politics 55 (May): 382-98. Brady, David W., John F. Cogan, and Douglas Rivers. 1995. How Republicans captured the House: An Assessment of the 1994 midterm elections. Stanford, CA: Hoover Institution. Campbell, James E., and Joe A. Sumners. 1990. Presidential coattails in Senate elections. American Political Science Review 84 (June): 513-24. Campbell, Karlyn Kohrs, and Kathleen Hall Jamieson. 1990. Deeds done in words: Presidential rhetoric and the genres of governance. Chicago: University of Chicago Press. CBS News/New York Times Poll. 1983. October 28. ———. 1986. December 1. ———. 1987. October 27. Cohen, Jeffrey E. 1997. Presidential responsiveness and public policy-making. Ann Arbor: University of Michigan Press. Cohen, Jeffrey E., Michael A. Krassa, and John A. Hamman. 1991. The impact of presidential campaigning on midterm U.S. Senate elections. American Political Science Review 85 (March): 165-80. Collier, Kenneth E. 1997. Between the branches. Pittsburgh: University of Pittsburgh Press. Congressional Quarterly. 1981. Tax cut passed by solid margin in House, Senate. Congressional Quarterly Weekly Report, August 1, 1374. ———. 1999. Clinton’s biggest prize was a frustrated GOP. Congressional Quarterly Weekly Report, January 9, 75-77. Conover, Pamela Johnston, and Lee Sigelman. 1982. Presidential influence and public opinion: The case of the Iranian hostage crisis. Social Science Quarterly 63 (June): 249-64. Cornwell, Elmer E., Jr. 1965. Presidential leadership of public opinion. Bloomington: Indiana University Press. Crewe, Ivor. 1988. Has the electorate become Thatcherite? In Thatcherism, edited by Robert Skidelsky. London: Chatto and Windus. ———. 1989. Values: The crusade that failed. In The Thatcher effect: A decade of change, edited by Dennis Kavanagh and Anthony Seldon. Oxford: Oxford University Press. Crewe, Ivor, and Donald Searing. 1988a. Ideological change in the British Conservative Party. American Political Science Review 82 (June): 361-84. ———. 1988b. Mrs. Thatcher’s crusade: Conservatism in Britain, 1972-1986. In The resurgence of conservatism in Anglo-American democracies, edited by B. Cooper, Allan Kornberg, and William Mishler. Durham, NC: Duke University Press. Curtice, John. 1986a. Interim report: Party politics. In British social attitudes: The 1987 report, edited by Roger Jowell, Sharon Witherspoon, and Lindsay Brook. Aldershot Hants, UK: SCPR\Gower. ———. 1986b. Political partisanship. In British social attitudes: The 1986 report, edited by Roger Jowell, Sharon Witherspoon, and Lindsay Brook. Aldershot Hants, UK: SCPR\Gower. Davis, James A. 1992. Changeable weather in a cooling climate. Public Opinion Quarterly 56 (Fall): 261-306. Deering, Christopher J., and Steven S. Smith. 1997. Committees in Congress. 3d ed. Washington, DC: CQ Press. Defense spending. 1987. Gallup Report, May, 2-3 Edwards, George C. III. 1980. Presidential influence in Congress. San Francisco: W. H. Freeman. ———. 1983. The public presidency. New York: St. Martin’s. ———. 1989. At the margins: Presidential leadership of Congress. New Haven, CT: Yale University Press. ———. 1991. Presidential influence in Congress: If we ask the wrong questions, we get the wrong answers. American Journal of Political Science 35 (August): 724-29. ———. 1995. Frustration and folly: Bill Clinton and the public presidency. In The Clinton presidency: First appraisals, edited by Colin Campbell and Bert A. Rockman. Chatham, NJ: Chatham House. ———. 1996. Presidential rhetoric: What difference does it make? In The future of the rhetorical presidency, edited by Martin J. Medhurst. College Station: Texas A&M University Press.

76  PRESIDENTIAL STUDIES QUARTERLY / March 2000 ———. 1997. Aligning tests with theory: Presidential approval as a source of influence in Congress. Congress and the Presidency 24 (Fall): 113-30. ———. 1999. Campaigning is not governing: Bill Clinton’s rhetorical presidency. In The Clinton legacy, edited by Colin Campbell and Bert Rockman. Chatham, NJ: Chatham House. Edwards, George C. III, and Andrew Barrett. 2000. Presidential agenda setting in Congress. In Polarized politics: Congress and the president in a partisan era, edited by Jon R. Bond and Richard Fleisher. Washington, DC: CQ Press. Edwards, George C. III, Andrew Barrett, and Jeffrey Peake. 1997. The legislative impact of divided government, American Journal of Political Science 41 (April 1997): 545-63. Edwards, George C. III, and Stephen J. Wayne. 1999. Presidential leadership. 5th ed. New York: St. Martin’s. Edwards, George C. III, and B. Dan Wood. 1999. Who influences whom? The President, Congress, and the media. American Political Science Review 93 (June): 327-44. Epstein, Leon D. 1967. Political parties in Western democracies. New York: Praeger. Erikson, Robert S., Norman R. Luttbeg, and Kent L. Tedin. 1980. American public opinion: Its origins, content, and impact. 2d ed. New York: John Wiley. Federal budget deficit. 1987. Gallup Report, August. Fenno, Richard F., Jr. 1973. Congressmen in committees. Boston: Little, Brown. ———. 1977. U.S. House members in their constituencies: An exploration. American Political Science Review 71 (September): 883-917. ———. 1979. If, as Ralph Nader says, Congress is “the broken branch,” how come we love our congressmen so much? In Congress in change: Evolution and reform, edited by Norman J. Ornstein. New York: Praeger. Ferejohn, John A. 1974. Pork barrel polities. Stanford, CA: Stanford University Press. Fleisher, Richard, and Jon R. Bond. 2000. Partisanship and the president’s quest for votes on the floor of Congress. In Polarized politics: Congress and the president in a partisan era, edited by Jon R. Bond and Richard Fleisher. Washington, DC: CQ Press. Fleishman, John A. 1986. Trends in self-identified ideology from 1972 to 1982: No support for the salience hypothesis. American Journal of Political Science 30 (August): 517-41. Flemming, Gregory N. 1995. Presidential coattails in open-seat elections. Legislative Studies Quarterly 20 (May): 197-212. Ford, Henry Jones. 1898. The rise and growth of American politics. New York: Macmillan. Gallup Poll. 1994. August 15-16. Glaros, Roberta, and Bruce Miroff. 1983. Watching Ronald Reagan: Viewers’ reaction to the president on television. Congress and the Presidency 10 (Spring): 25-46. Groseclose, Tim. 1994. Testing committee composition hypotheses for the U.S. Congress. Journal of Politics 56 (May): 440-58. Hall, Richard L., and Bernard Grofman. 1990. The committee assignment process and the conditional nature of committee bias. American Political Science Review 84 (December): 1149-66. Hamilton, Alexander, James Madison, and John Jay. 1938. The federalist. New York: Modern Library. Hart, Roderick P. 1987. The sound of leadership: Presidential communication in the modern age. Chicago: University of Chicago Press. Healey, Jon. 1993. Democrats look to salvage part of stimulus plan. Congressional Quarterly Weekly Report, April 24, 1001-4. Hertsgaard, Mark. 1988. On bended knee: The press and the Reagan presidency. New York: Farrar, Straus, and Giroux. Hill, Kim Quaile. 1998. The policy agendas of the president and the mass public: A research validation and extension. American Journal of Political Science 42 (October): 1328-34. Holtzman, Abraham. 1970. Legislative liaison. Chicago: Rand McNally. Jacobson, Gary C. 1994. The 1994 midterm: Why the models missed it. Extension of Remarks. APSA Legislative Studies Section. Johnson, Haynes. 1991. Sleepwalking through history. New York: Norton. Jones, Charles O. 1988. The trusteeship presidency. Baton Rouge: Louisiana State University Press. ———. 1994. The presidency in a separated system. Washington, DC: Brookings Institution. Kernell, Samuel. 1997. Going public. 3d ed. Washington, DC: Congressional Quarterly. Kingdon, John W. 1989. Congressmen’s voting decisions. 3rd ed. Ann Arbor: University of Michigan Press.

Edwards / BUILDING COALITIONS  77 Krehbiel, Keith. 1991. Are congressional committees composed of preference outliers? American Political Science Review 84 (March): 149-63. ———. 1994. Deference, extremism, and interest group ratings. Legislative Studies Quarterly 19 (February): 61-77. Lipset, Seymour Martin. 1986. Beyond 1984: The anomalies of American politics. PS 19: 228-29. Londregan, John, and James M. Snyder. 1994. Comparing committee and floor preferences. Legislative Studies Quarterly 19 (May): 233-66. Maltzman, Forrest. 1995. Meeting competing demands: Committee performance in the postreform House. American Journal of Political Science 39 (August): 653-82. Matthews, Donald R., and James A. Stimson. 1975. Yeas and nays: Normal decision-making in the U.S. House of Representatives. New York: John Wiley. Mayer, William G. 1992. The changing American mind. Ann Arbor: University of Michigan Press. Mayhew, David R. 1991. Divided we govern. New Haven, CT: Yale University Press. Medhurst, Martin J. 1993. Dwight D. Eisenhower: Strategic communicator. Westport, CT: Greenwood. Mueller, John E. 1970. War, presidents, and public opinion. New York: John Wiley. Niskanen, William A. 1988. Reaganomics. New York: Oxford University Press. Opinion outlook. 1987a. National Journal, April 18, 964. ———. 1987b. National Journal, June 13, 1550. Ornstein, Norman J. 1982. Assessing Reagan’s first year. In President and Congress: Assessing Reagan’s first year, edited by Norman J. Ornstein. Washington, DC: American Enterprise Institute. Ornstein, Norman J, Thomas E. Mann, and Michael J. Malbin. 1998. Vital statistics on Congress, 1997-1998. Washington, DC: CQ Press. Page, Benjamin I., and Robert Y. Shapiro. 1985. Presidential leadership through public opinion. In The presidency and public policy making, edited by George C. Edwards III, Steven A. Shull, and Norman C. Thomas, 22-36. Pittsburgh: University of Pittsburgh Press. ———. 1992. The rational public. Chicago: University of Chicago Press. Parker, Glenn R., and Roger H. Davidson. 1979. Why do Americans love their congressmen so much more than their Congress? Legislative Studies Quarterly 4 (February): 53-61. Peterson, Geoffrey D., and J. Mark Wrighton. 1998. The continuing puzzle of committee outliers: A methodological reassessment. Congress and the Presidency 25 (Spring): 67-78. Ragsdale, Lyn. 1984. The politics of presidential speechmaking, 1949-1980. American Political Science Review 78 (December): 971-84. Reagan, Ronald. 1990. An American life. New York: Simon and Schuster. Reilly, John E., ed. 1987. American public opinion and U.S. foreign policy 1987. Chicago: Chicago Council on Foreign Relations. Rentoul, John. 1989. Me and mine: The triumph of the new individualism. London: Unwin Hyman. Rohde, David W. 1991. Parties and reform in the postreform House. Chicago: University of Chicago Press. Rossi, Eugene J. 1965. Mass and attentive opinion on nuclear weapons test and fallout, 1954-1963. Public Opinion Quarterly 29 (Summer): 280-97. Schneider, William. 1985. The voters’ mood 1986: The six-year itch. National Journal, December 7, 2758. Shepsle, Kenneth A. 1978. The giant jigsaw puzzle: Democratic committee assignments in the modern House. Chicago: University of Chicago Press. Sigelman, Lee. 1980. Gauging the public response to presidential leadership. Presidential Studies Quarterly 10 (Summer): 427-33. Sigelman, Lee, and Carol K. Sigelman. 1981. Presidential leadership of public opinion: From “benevolent leader” to kiss of death? Experimental Study of Politics 7 (3): 1-22. Skowronek, Stephen. 1993. The politics presidents make. Cambridge, MA: Harvard University Press. Smith, Craig Allen, and Kathy B. Smith. 1994. The White House speaks: Presidential leadership as persuasion. Westport, CT: Praeger. Smith, Tom W. 1990. Liberal and conservative trends in the United States since World War II. Public Opinion Quarterly 54 (Winter): 479-507. Snyder, James M. 1992. Artificial extremism in interest group ratings. Legislative Studies Quarterly 17 (August): 319-45. Sobel, Richard, ed. 1993. Public opinion in U.S. foreign policy. Lanham, MD: Rowman and Littlefied.

78  PRESIDENTIAL STUDIES QUARTERLY / March 2000 Speakes, Larry. 1988. Speaking out. New York: Scribner’s. Stimson, James A. 1991. Public opinion in America: Moods, cycles, and swings. Boulder, CO: Westview. Stuckey, Mary E. 1990. Playing the game: The presidential rhetoric of Ronald Reagan. New York: Praeger. Sullivan, Terry. 1991. The bank account presidency: A new measure and evidence on the temporal path of presidential influence. American Journal of Political Science 35 (August): 686-723. Sundquist, James L. 1992. Constitutional reform and effective government. Washington, DC: Brookings Institution. Supporting a greater federal role. 1987. National Journal, April 18, 924. Sussman, Barry. 1983. Reagan’s talk gains support for policies. Washington Post, October 30, sec. A, pp. 1, 18. Switchboards swamped with calls over tax plan. 1993. New York Times, August 5, A18. Thomas, Dan, and Lee Sigelman. 1985. Presidential identification and policy leadership: Experimental evidence on the Reagan Case. In The presidency and public policy making, edited by George C. Edwards III, Steven A. Shull, and Norman C. Thomas, 37-49. Pittsburgh: University of Pittsburgh Press. Tulis, Jeffrey K. 1987. The rhetorical presidency. Princeton, NJ: Princeton University Press. Waldman, Steven, and Bob Cohn. 1994. Health care reform: The lost chance. Newsweek, September 19. Wattenberg, Martin P. 1990. From a partisan to a candidate-centered electorate. In The new American political system, edited by Anthony King, 169-71. Washington, DC: American Enterprise Institute. ———. 1991. The rise of candidate-centered politics: Presidential elections of the 1980s. Cambridge, MA: Harvard University Press. Wayne, Stephen J. 1978. The legislative presidency. New York: Harper and Row. Welch, Reed. 1997. Tuning in or tuning out: The influence of televised presidential addresses on public opinion. Unpublished doctoral dissertation, Texas A&M University. Windt, Theodore Otto, Jr. 1990. Presidents and protestors: Political rhetoric in the 1960s. Tuscaloosa: University of Alabama Press. Woodward, Bob. 1994. The agenda: Inside the Clinton White House. New York: Simon and Schuster. ———. 1996. The choice. New York: Simon and Schuster. Zaller, John R. 1992. The nature and origins of mass opinion. Cambridge, UK: Cambridge University Press. ———. 1994. Elite leadership of mass opinion: New evidence from the Gulf War. In Taken by storm: The media, public opinion, and U.S. foreign policy in the Gulf War, edited by W. Lance Bennett and David L. Paletz, 186-209. Chicago: University of Chicago Press. Zarefsky, David. 1990. Lincoln, Douglas, and slavery: In the crucible of public debate. Chicago: University of Chicago Press.