Document Type


Publication Date



When black Americans and white Americans want the president to do different things, who wins? When low-income earners prefer different government action than do middle and high-income earners, whose preferences are reflected in presidential behavior? Recent studies show that congressional behavior often most closely follows the preferences of the white and the wealthy, but we know relatively little about presidential behavior. Since the president and Congress make policy together, it is important to understand the extent of political equality in presidential behavior. We examine the degree to which presidents have provided equal representation to these groups over the past four decades. We compare the preferences of these groups for federal spending in various budget domains to presidents’ subsequent budget proposals in those domains from 1974 to 2010. Over this period, presidents’ proposals aligned more with the preferences of whites and high-income earners. However, Republican presidents are driving this overall pattern. Democratic presidents represent racial and income groups equally, but Republicans’ proposals are much more consistent with the spending preferences of whites and high-income earners. This pattern of representation reflects the composition of the president's party coalition and the spending preferences of groups within the party coalition. Who gets what they want from government? That is, whose preferences for government policy are best reflected in the policies government creates? Sidney Verba (2003, 663) argued “the equal consideration of the preferences and interests of all citizens” is “one of the bedrock principles in a democracy.” However, several recent studies of U.S. politics find that the wealthy and whites are more likely than the poor and racial/ethnic minorities to see their preferences reflected in government behavior and policy (e.g., Bartels 2008; Ellis 2012; Flavin 2012; Gilens 2012; Gilens and Page 2014; Griffin and Newman 2008; Jacobs and Page 2005;). Additional works qualify, critique, and complicate these studies (e.g., Bhatti and Erikson 2011; Soroka and Wlezien 2008;), finding that the degree of inequality in political outcomes varies across political contexts (e.g., Brunner, Ross, and Washington 2013; Ellis 2013;; Rigby and Wright 2011, 2013;), though few have argued that the American political system reflects the preferences of various income and racial groups equally. The majority of this literature examines the content of public policy or congressional behavior. We know much less about presidential representation of income and racial/ethnic groups. Although presidency scholars have made major strides in understanding when and how much presidential behavior mirrors public preferences (e.g., Canes-Wrone and Shotts 2004; Cohen 1997; Druckman and Jacobs 2011; Erikson, MacKuen, and Stimson 2002, Rottinghaus 2006;), according to Druckman and Jacobs (2009), we have just begun to appreciate which groups’ preferences presidents represent best. In this study, we seek a more complete understanding of whose preferences are best represented in presidential behavior. In doing so, we build on and contribute to a growing literature that examines inequality in political representation more broadly (e.g., Enns and Wlezien 2011). It is not a foregone conclusion that the patterns of inequality seen elsewhere in the American political system would also characterize the presidency. The president serves as a national leader, rather a representative of a smaller, sometimes more homogeneous, constituency like members of Congress (Baker 2008), which can generate different incentives to represent specific groups. Moreover, presidential and congressional representation may differ given institutional differences in method of election, term length, term limits, and citizens’ different expectations of these elected officials. For example, legislators who are retiring, thus free from electoral pressures to represent their constituents’ preferences, behave differently than legislators running for reelection (e.g., Rothenberg and Sanders 2007). Since second-term presidents spend half their tenure in office without the possibility of reelection, unlike the majority of members of Congress, presidential behavior may differ from congressional behavior. If so, minority representation may vary significantly across the branches of government. In the particular policy arena we study here, presidents’ proposals for federal government spending, the president and members of Congress may often have different incentives for representation. Presumably, the public holds the president more accountable than members of Congress for the composition of the budget simply because the president proposes an entire budget. Members of Congress can request additional spending on areas of particular ideological or economic interest to their constituents, but members do not propose entire budgets, meaning they can often make requests without the hard choices the president must make: a dollar increase in one program means a dollar decrease in another or the president must bear the political cost of a bloated budget. Moreover, it is important to examine equality of representation in the context of the presidency because the president plays the strongest and most direct role in representing citizens’ preferences of any single actor in the American political system. A Member of Congress may represent her constituents well or poorly, but in the end, she is but one of 435 or 100 members in a single chamber of a bicameral institution that comprises one of three branches of government. A constituent may be especially well represented by her member of the House, but that member has limited influence on the outputs of the House, much less the ultimate output of the policymaking process involving the House, Senate, and president. Thus, a connection between public preferences and policy outputs is important (Gilens 2012), but does not tell us much about the behavior of any specific individuals. In contrast, the president is a single actor who can often take direct action (Howell 2003). Of course, the president often relies heavily on others in the administration, but within the executive branch, the president's opinion is decisive, unlike individual lawmakers’. Thus, within the American system, the president has the most power to expand, shrink, or even reverse the patterns of unequal congressional representation. Consequently, the presidency should be of great significance to representation scholars. We examine the degree to which presidents have provided equal representation to racial groups (blacks and whites—unfortunately our data source did not identify Latinos for most of our period of study) and to income groups (low-, middle-, and high-income earners) over the past four decades. We observe whether these groups prefer government spending to increase, decrease, or remain about the same. We then compare group preferences to presidents’ subsequent budget proposals to see whether presidential behavior matches group preferences. Doing so enables us to see, for example, how often presidents propose a spending increase in a domain when a group prefers more spending in that policy area. Analyzing data from 1974 to 2010, we find that presidents’ proposals match whites’ and high-income earners’ preferences significantly more often than the preferences of African Americans and low-income earners. In particular, Republican presidents’ proposals are more often congruent with the spending preferences of whites and the wealthy. Democrats, on the other hand, tend to match the groups’ preferences equally. This pattern of presidential representation reflects the composition of current party coalitions. That is, presidents act most consistently with the preferences of the largest groups in their party coalitions, leading to different patterns of representation for Democrats and Republicans.

Publication Title

Congress & the Presidency