Why has economic inequality remained and expanded in a political system that features competitive battles for elected office? After all, competitive elections have long been expected to motivate politicians seeking office or re-election to follow the policy preferences of the median voter or, at a minimum, to rectify sagging economic circumstances in anticipation that voters will punish them for a disappointing record in office.^{Footnote 1} A large and longstanding body of research has confirmed these expectations by reporting “policy responsiveness”—the tendency of government policy to match the policy preferences of citizens.^{Footnote 2}

Nonetheless, growing bodies of research are demonstrating that economic inequality has expanded to levels in the United States that are without parallel among western democratic countries. The catalogue of causes include government policy that has either failed to act adequately in the face of accelerating market pressures for inequality or facilitated the concentration of income and wealth through tax policy, inadequate assistance for vulnerable populations, weaker protections for labor organizing, and growing deference to finance (Alvaredo, Atkinson, Piketty, and Saez Reference Alvaredo, Atkinson, Piketty and Saez2013; Piketty Reference Piketty2015; Jacobs and Skocpol Reference Jacobs and Skocpol2005; Jacobs and King Reference Jacobs and King2016).

But why has wealth and income become more concentrated even though competitive elections continue to offer the institutional means for prospective and retrospective choices by voters? A stream of recent studies offer a response: continued and perhaps increasing declines in “policy responsiveness” and, more precisely, what we refer to as “segmented responsiveness”—the disproportionate tailoring of policy to the wishes and wants of whites and the affluent as compared to middle and lower-income Americans and people of color.^{Footnote 3}

Two explanations stand out for the puzzle of waning policy responsiveness and deteriorating economic circumstances within liberal democracy.

First, the political rights to participate in elections, contact officeholders, contribute to campaigns, and other forms of engagement are not exercised equally. “Unequal voice” and the relatively muted participation of the less well-off enable the affluent to dominate what policy makers hear and the intensity with which those messages register in the halls of government and the ears of elected officials (Schlozman, Verba and Brady Reference Schlozman, Verba and Brady2012; Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995). The implication is that elected officials would rectify inequality if political participation by people of color and middle and lower income citizens was similar to that of whites and the affluent.

But how decisive is political participation in dictating policy? Daniel Butler’s Representing the Advantaged rejects the participation account based on his findings from clever field experiments that investigated the responses of officeholders to requests from putative constituents. “[E]ven if all voters participate at equal rates,” he reports, “we will still observe bias in representation…. [because] differences in participation alone cannot explain bias in representation” (pp. 1–3).

The second explanation, which Butler and our research advances, is that inequality and the breakdown of policy responsiveness results from “who governs”—the people in office and the incentives and institutions that condition their behavior. Before we elaborate this account, we want to clarify that participation and elite accounts are not mutually exclusive: Unequal voice shapes the motivations of elected officials and, in turn, the disregard of elites for less-well established citizens impacts their motivations and resources to participate (Jacobs and Shapiro Reference Jacobs and Shapiro2000; Campbell Reference Campbell2003; Mettler and Soss Reference Mettler and Soss2004). Stronger participation would bring new and louder voices into the halls of government, but the “inequality elite” account insists that more equal voice would not dislodge policy makers and elite governance that systematically cater to the affluent and the best organized.

Elite politics has erected several barriers to using liberal democratic institutions to reverse historic inequality. For Butler, governing elites bring “information, opinions, and attitudes… into office [that] lead to significant bias in representation.” The bias is baked into their personal preferences and experiences that inform what positions they adopt, how they evaluate citizen demands, and their decisions about whether or not to respond.

In our book, we shift from the personal mindsets of officials to the incentive structure and institutional capacities of officeholders, drawing on quantitative and qualitative analyses of presidential archives. Our study of the Reagan White House reveals that he systematically tailored his policies to the affluent and to party activists who had become increasingly influential in the nomination process. The incentives of the party nomination process reward those who uncompromisingly follow the policy goals that they and their supporters favor (Jacobs and Shapiro Reference Jacobs and Shapiro2000;Wood Reference Wood2009). What about the risks of not following the median voter? Reagan and other presidents used the White House’s institutional capacity to privately track and attempt to manipulate public opinion to create leeway: They focused on salient issues (expanding their freedom to act on less visible issues) and built appealing personal images.

Both our account and Butler’s indicate that the link between competitive elections and responsive policy making has been disrupted. Politicians stick to their personal agendas; they are guided by incentive structures that promise certain punishment for compromising policy goal while posing only uncertain consequences for wandering from the views of median voters; and they are confident in their institutional capacity to manipulate voters and shirk accountability. Political representation remains but has a different form: It has been displaced from competition among politicians over aligning their policies with those preferred by the median voter and, instead, involves segmented representation and a reconstitution of representation from policy to non-policy grounds.

Bias is, of course, not new to American politics (Schattschneider Reference Schattschneider1960; Mills Reference Mills1956). What is new today is the magnitude of the disparities and the consequences for reconcentrating wealth and income. Today’s extraordinary bias results from the confluence of intensifying global and domestic sources of market-generated inequality and breakdowns in the liberal democratic links between competitive elections and responsive policy making.