THE COURT'S VOTING POWER PRINCIPLE
Baker v. Carr stands for the (at times contested) principle that the federal courts may intervene to assess whether citizens' votes carry equal weight in state voting procedures. (1) According to the Baker Court, before Baker--and the revolution that it spawned--citizens' votes did not have equal influence on the lawmaking process, contrary to what the Constitution requires because differently-sized populations could be represented by the same number of elected officials.
We contend that even after Baker and its associated cases, voters remain unequal. At least from the perspective of a member of Congress seeking reelection, some citizens (i.e., likely voters and "swing" voters) are more important than others. Consequently, members of Congress tend to act in closer concert with the preferences of constituents whose votes matter more. Since some groups (e.g., the young, racial/ethnic minorities) are less likely to vote or be swing voters, these groups tend to receive less of what they want from government compared to other groups.
The Court's approach in Baker presumes the relative homogeneity of citizens' policy preferences within electoral districts and conceives of the legislature as primarily a distributive body where all citizens desire a larger share of the federal pie. In this case, if differently sized (and disagreeing) districts are both represented by a single legislator, the citizens of the more populous district will have their votes diluted. In order for citizens to have equal influence over legislative outcomes, district populations must be equal. As the Court later articulated in Reynolds v. Sims--"the right of suffrage can be denied by a debasement or dilution of the weight of a citizen's vote just as effectively as by wholly prohibiting the free exercise of the franchise." (2)
Gray v. Sanders (3) then signaled that within a single electoral district (in Gray, a state), electoral rules that aggregate votes in such a manner that votes are not counted equally are also unconstitutional. At issue in Gray was whether the Democratic Party primary for the nomination of U.S. Senate and other statewide officers could employ a "county--unit" system wherein unit votes were allocated to counties by population, but the relationship between population and units in the allocation formula was logarithmic rather than linear. In striking down this election procedure, the Court extended Baker to require equally populated geographical units within electoral districts. That is, citizens should have equal say in the selection of their elected representatives.
Here, the Court contemplates that citizens in a district who belong to politically relevant demographic groups will not always agree with one another about the best policy direction and, in this case, all individuals should have equal say in the selection of their elected representative: "Once the geographical unit for which a representative is to be chosen is designated, all who participate in the election are to have an equal vote--whatever their race, whatever their sex, whatever their occupation, whatever their income, and wherever their home may be in that geographical unit." (4) To aggregate individual votes in such a way that a minority could defeat a majority, as was true in Georgia before Gray, meant that citizens' votes carried unequal weight. This, of course, the Court could not abide: "Once the class of voters is chosen and their qualifications specified, we see no constitutional way by which equality of voting power may be evaded." (5) Finally, invoking the Equal Protection Clause of the 14th Amendment, the Gray Court concluded: "The conception of political equality ... can mean only one thing--one person, one vote." (6)
This Article contends that, in these cases, the Court was concerned about whether citizens have equal say in the selection of officials not only because it may affect who wins an election. Additionally, and perhaps even more importantly, the Court was concerned about this issue because regardless of who wins any given race we have reason to expect that in any system where votes are unequally weighted, elected officials will serve the interests of those whose votes matter most. Stated another way, one important motivation behind the one person, one vote principle is to move our system of government toward equal representation.
Behind Baker is an empirical assumption that when citizens hold unequal voting weight, their elected representatives will be more attentive to those holding more electoral weight. A half-century later, we test this assumption. Below, we ask first whether citizens really have equal voting power. Finding that they do not, we examine whether Baker's assumption is correct, asking whether citizens with more voting power are better represented. We argue that despite Baker and Gray, voting power is not equal across the population. Some votes affect the outcome of an election more than others. Moreover, citizens' preferences are not reflected in the actions of their representatives in equal measure. Those with more voting power tend to enjoy greater correspondence between their policy preferences and their representatives' actions than is true for those with less voting power. We argue that the Baker court assumed correctly that unequal votes would lead to unequal representation of constituents' policy preferences in their representatives' behavior in office. Although the Baker decision and those following it may have moved the country closer to one of equal voting power, significant inequality in voting power remains.
THE MEANING AND IMPORTANCE OF VOTING POWER
Members of Congress (MCs) have electoral incentives to treat constituents unequally. To see this, think of an MC who has the capacity to choose a small number of her total constituents who would become slightly more favorable toward her. Who would she choose? Presumably, she would choose those constituents who are likely to vote. Improving a small group of citizens' favorability toward the MC would not help her re-election effort if the target constituents do not typically vote. Only those who vote can weigh in on the re-election or rejection of the incumbent] In addition, the MC would presumably choose constituents who sit on the fence between voting for or against her. A small bit of newfound favorability toward the MC will not affect the votes of constituents who long ago determined to vote for or against her. Thus, MCs, who have to allocate scarce resources like their advertising dollars, time, and roll-call votes, have incentives to allocate those resources to affect the votes of likely voters who are not yet committed to voting for or against the MC. This is not to dismiss an MC's need to attend to his or her "base" supporters; however, "swing" or "marginal" voters will receive disproportionate attention from a calculating politician. (8) As Larry Bartels put it, "[r]ational candidates seeking to maximize their electoral prospects must 'go hunting where the ducks are,' tailoring their appeals to those prospective voters who are both likely to turn out and susceptible to conversion." (9)
The concept of voting power reflects the contribution an individual will make to an incumbent's re-election effort. MCs can easily assess any given citizen's voting power from fairly basic and relatively easily ascertained political characteristics. The relevant political characteristics include demographic or attitudinal qualities a citizen possesses that provide a cue to elected officials about the citizen's likely political behavior. For instance, in the United States we might say that a citizen's race/ethnicity, income, gender, and age--as well as whether the individual is a partisan or an independent--are likely to convey information to an elected official about two aspects of the citizen's likely political behavior.
First, these qualities may convey information about the likelihood that the individual will participate in the next election. For instance, if an individual is a twenty-five year old Latino male, based on his age or ethnicity or gender (or a combination of the three) and the turnout propensity of these groups, the official can estimate the likelihood that the individual will participate in the next election. Individuals with high turnout propensities are more likely to influence an MC's re-election prospects because they are very likely to vote for either the incumbent or the challenger. These individuals, therefore, have more voting power than those unlikely to vote.
Second, these qualities provide cues about an individual's tendency to favor one of the parties. So, using the individual in the preceding paragraph once again as an example, the elected official will want to assess the tendency of the young, Latinos, and males to favor one of the parties. Individuals whose qualities suggest that they do not have a tendency to favor one of the parties possess more voting power. Citizens whose votes are up for grabs are especially attractive because their votes are both not already secured and not unattainable. Thought of another way, these votes can be won, but if they are not, they are very likely to go to one's opponent.
Voting power merits our attention because inequalities in voting power are likely to be translated into inequalities in political influence. To see this, we build a theoretical argument from a series of assumptions to two hypotheses:
ASSUMPTION 1: AN INSTITUTIONAL CONFIGURATION WITH SINGLE-MEMBER DISTRICTS
Article I, section 2 of the Constitution provides that, "The House of Representatives shall be composed of Members chosen every second year by the People of the several States ... Representatives ... shall be apportioned among the several States which may be included within this Union, according to their respective Numbers." (10) The Constitution did not, however, specify the manner in...