In Parts One and Two of this series, we introduced an important pending federal court challenge to Mississippi’s gubernatorial election regime, situated the dispute in constitutional context, and examined various procedural aspects of the case. Today we turn in more detail to the merits of the constitutional claims the plaintiffs assert, focusing in particular on plaintiffs’ contention that Mississippi’s method of electing governors violates the so-called one-person, one-vote principle the Supreme Court has held that states must respect under the Equal Protection Clause of the Fourteenth Amendment.
As we explained in the earlier installments of this series, Mississippi elects its governors using a system fundamentally different from that of any other state. To be elected by the voters of the state, a successful gubernatorial candidate must garner both: 1) a majority (rather than just a plurality) of statewide votes; and 2) a majority of the Mississippi House of Representatives districts. If no candidate satisfies both of these requirements, the state House of Representatives selects the governor from the top two popular-vote getters.
The complaint filed in federal district court earlier this year alleges that the Mississippi system, which was added to the Mississippi Constitution in 1890, violates the federal constitutional rights of the plaintiffs, who are individual African American registered voters in the state. In particular, the plaintiffs contend that Mississippi’s system was intended to and does indeed harm African American voters, who make up a large group (but still a minority) of Mississippi voters, in several ways. First, because African American voters (and white voters too) often engage in bloc voting, the requirement (the so-called “Popular Vote Rule”) that a successful candidate win a majority of the popular votes—rather than simply more than any other candidate—makes it hard for a minority group like African Americans to elect the candidate of their choice. Moreover, because the state’s African American population is not distributed evenly throughout the state but is concentrated in a relatively small number of House districts, the requirement (known as the “Electoral Vote Rule”) that a winning candidate prevail in a majority of House districts across the entire state hurts African Americans (and also Democrats).
But even putting aside the claim of invidious racial intent (which we addressed in Part One and which will ultimately depend upon the specific evidence of bad legislative motive plaintiffs are able to adduce), the plaintiffs assert that the Electoral Vote Rule also violates the one-person, one-vote principle the Supreme Court has recognized for elections for state and local offices, insofar as someone running for Mississippi governor could win a majority of state House districts but get far fewer votes than her opponent statewide. For these reasons, say the plaintiffs, Mississippi’s regime violates the Fourteenth Amendment. (The plaintiffs further assert that the Mississippi scheme violates the Fifteenth Amendment and the federal Voting Rights Act, but in this series we limit ourselves to the Fourteenth Amendment arguments.)
Nuances in the One-Person, One-Vote Line of Cases
The most well-known of the so-called malapportionment cases the Supreme Court has decided, such as the 1964 cases of Wesberry v. Sanders and Reynolds v. Sims, invalidated the use of legislative districts within a state (for purposes of electing congresspersons and state legislators, respectively) that are significantly different in population size. Prior to these cases in the 1960s, some legislative districts (typically in an urban region with a high percentage of people of color) would frequently have far more persons in them than other (often rural and whiter) districts, even though each district would elect one representative to the legislature.
But the Warren Court’s work in this realm was not limited to legislative elections. In particular, in Gray v. Sanders, the Court in 1963 (the year before Wesberry and Reynolds) invalidated a Georgia so-called “county unit” system for tabulating votes in primary elections for statewide office including the governorship. Under Georgia’s system, “[c]andidates for [statewide office] nominations who received the highest number of popular votes in a county were considered to have carried the county and to be entitled to” a certain number of county unit votes. A candidate for governor (or U.S. Senator) needed a “majority of the county unit vote [to be] nominated.” But the number of county unit votes allocated to each county was not proportional to that county’s population. For example, Fulton County (which includes Atlanta and a large African American population) had about 14% of the state’s people, but Fulton County’s six county unit votes made up only 1.5% of the total 410 county unit votes in the state. As the Court would observe in describing Gray in a later case, “[t]he result was that the number of votes of persons living in large counties was given no more weight in electing state offices than was given to a far fewer number of votes of persons residing in small counties. This discrimination against large[-] county voters was held to deny them the equal protection of the laws.”
A potentially significant distinction between Mississippi’s gubernatorial election regime in 2019, on the one hand, and the schemes invalidated in Wesberry, Reynolds, and even Gray, on the other, is that every state House district in Mississippi today (each of which selects a member to the lower house of the Mississippi legislature) contains (in order to comply with Reynolds) an equal (or nearly so) number of persons. In other words, there is no overt discrimination against more populous areas of the state under Mississippi’s modern regime the way there was in each of the cases described above from the 1960s.
One thus might think that while Mississippi’s Electoral Vote Rule (requiring a candidate to win a majority of state House districts) is undesirable and unfair, it doesn’t necessarily violate the Fourteenth Amendment’s commands as understood by the Supreme Court. But Judge Jordan (the district court judge in the Mississippi case), in concluding that the plaintiffs had established a likelihood of success on their one-person, one-vote challenge, pointed to the fact that the Supreme Court in Gray, in footnote 12, asserted (albeit in dicta) that Georgia’s regime would be problematic “even if [county] unit votes were allocated strictly in proportion to population,” because of the winner-take-all nature of earning county unit votes. To explain what it saw as the problem, the Court provided an example: even assuming county unit votes were allocated in strict proportion to each county’s population, “if a candidate won 6,000 of 10,000 [popular] votes in a particular county, he would get the entire unit vote [for that county], the other 4,000 votes for a different candidate being worth nothing and being counted only for the purpose of being discarded.”
In a 1971 case several years later (Gordon v. Lance), the Court echoed this observation from Gray’s footnote 12, saying indeed that that Gray Court “held that the county-unit system would have been defective even if unit votes were allocated strictly in proportion to population” because of the problem of votes for a losing candidate being “discarded” (emphasis added). Although Gordon characterized Gray’s footnote 12 as a “h[o]ld[ing],” we should mention that this remark in Gordon was itself not essential to Gordon’s outcome.
One possible reason for skepticism about Gray’s footnote 12 is that in many American elections votes for losing candidates are “discarded” in this way. For example in any election for a legislator within a district (not to mention election for a statewide office like U.S. Senator), all the votes of candidates other than the leading vote getter are discarded in the sense that these votes do not determine the selection of the representative of that district, nor the partisan makeup of the legislature as a whole. (In parliamentary systems, by contrast, where each party earns a proportional share of the legislature’s total number of seats based on that party’s share of voter support in the entire state/nation, perhaps votes are not discarded in this way, but nothing in the Constitution requires that American elections within states follow the parliamentary model.)
Indeed, and to illustrate the complexity of the Supreme Court’s decisions in this realm, the Court upheld, in the 1966 case of Fortson v. Morris, three years after Gray but five years before Gordon, a system in which Georgia’s governor was elected by the Georgia General Assembly if no gubernatorial candidate received a majority of statewide votes, even though the party that controlled a majority of Assembly seats might have lacked majority or plurality voter support statewide because of the vote-discarding, winner-take-all nature of legislative district elections (something that is potentially exacerbated by partisan gerrymandering). As the Court majority observed in Fortson, “[t]here is no provision of the United States Constitution or any of its amendments which either expressly or impliedly dictates the method a State must use to select its Governor. . . . The method [of legislative selection of governor that Georgia] chose for this purpose was . . . well known and frequently utilized before and since the Revolutionary War. Georgia Governors were selected by the State Legislature [without any popular election being held for governor at all] until 1824.” And, said the Court, none of this was called into question by Gray.
So some observers might discern tension within the case law. Nevertheless, one might try to harmonize all the cases by saying that it is perfectly permissible for a state to make use of legislative districts (provided they are of equal population size) to elect persons (e.g., state legislators) who then select the governor; but a state may not use legislative districts to create non-human determinants (such as unit votes) that are then employed automatically to select governors without any intermediation of individual representatives.
Perhaps this distinction between the use of geographical districts to select human delegates who can then reflect and deliberate before choosing a governor, on the one hand, and the use of the same kind of geographical units to implement automatic formulae like unit-vote tabulation that in turn dictate who shall be governor without additional human involvement, on the other, makes sense. Or maybe it doesn’t, and maybe a majority of the current Supreme Court would disclaim the footnote dicta in Gray, and limit that case to instances in which county unit-vote allocation deviated from population proportionality. But Judge Jordan in the Mississippi case was at least arguably right that he should follow the Gray footnote’s direction until the Court revisits matters. And if the reasoning in footnote 12 is binding on lower court judges, then Mississippi’s Electoral Vote Rule could be vulnerable, insofar as it counts votes that are necessary to elect a governor in terms of geographic units rather than in terms of individual voters.