Mississippi's anachronistic primary system made it possible for a mystery man to win the Democratic nomination for governor.
The announcement of the winner in last week’s Democratic gubernatorial primary sounded more like an answer in Jeopardy! — it came in the form of a question.
Who is Robert Gray?
Not a single one of the 147,043 people who voted for Mr. Gray knew who he was when they went to the polls. That’s no exaggeration: he made no effort to campaign, and he said he forgot to vote for himself. Even his mother thought she was voting for a different Robert Gray.
His opponents Vicki Slater and Valerie Short — legitimate candidates who actively sought the nomination — didn’t know either. They never crossed paths with Gray on the campaign trail. He didn’t spend a single dollar on his candidacy. He didn’t even have a Facebook page.
Mississippi elects each of its executive and legislative officials in the odd year prior to the presidential election. That includes eight statewide offices, three regional public service and transportation commissioners, 52 senators, and 122 members of the House.
A number of local elections are planned according to the state election calendar. They include — but are not limited to — sheriff, district attorney, chancery clerk, county supervisor, coroner, and in certain places, school district superintendent. Every one of these local races allows for party affiliation.
Party primaries occur on the first Tuesday in August, and the general election is held on the first Tuesday in November.
Mississippi has no official party registration, so voters have the option of choosing either party’s primary. But choosing a party primary means voting on the entire slate — you can’t vote in the GOP primary for governor and then the Democratic primary for county supervisor.
In the days since the election, the questions have shifted from who to how. And while no explanation is definitive, Gray almost certainly benefited from an uncommon confluence of factors.
The two active gubernatorial candidates had never run for elected office before, so neither possessed high name recognition or a pre-existing constituency. Slater and Short raised more money than Gray’s $0, but neither came close to what successful Democratic candidates have spent in past primaries. In the end, his opponents’ relative anonymity allowed Gray to capitalize, inadvertently, on his one major advantage: he was running against two women in the only state that has never elected a female governor or member of Congress. Without anything else to distinguish among them, more than half of primary voters opted for the man.
(In my opinion, too much credit has been given to Gray’s name appearing first on the ballot: Tim Johnson, the Democratic establishment’s preferred candidate for lieutenant governor, was listed second on the ballot. He prevailed easily over his male opponent despite spending less money than Slater.)
No matter whether you buy that explanation, I believe that the unique circumstances of Gray’s anomaly are less important than the conditions that made such a result possible in the first place. In other words, our question should change from, “Why did 51 percent of Democratic voters mark Robert Gray on August 4?” to, “Why did a majority of Democratic primary voters go to the polls unprepared to vote for the state’s top office?”
As I’ll explain down the page, the conditions that enabled Gray’s victory are deeply embedded in Mississippi’s anachronistic election process, a system designed for one-party white supremacist rule that has not kept pace with the expansion of voting rights or partisan realignment. As a result, the 21st century Democratic primary electorate is distorted through 19th century laws and traditions. Problems occur when these three vestiges of the old regime work in concert:
- Candidates for every state and local office can declare a partisan affiliation, but voters can not register with a party. Any voter can participate in either the Democratic or Republican primary.
- Many local officials and their challengers still run as Democrats without GOP opposition. Even in many majority-Republican counties, the Democratic primary operates as the de facto general election for certain local offices.
- Local races (of which there are many) drive turnout even when state offices are contested. The Democratic primary attracts more voters than the Republican primary despite GOP dominance in statewide elections. A large share of the Democratic primary voters do not support the Democratic nominees in the general election.
The roots of this system begin — as so many things do in Mississippi — in the era following the Civil War.
The White Primary
During Reconstruction, the 15th Amendment was added to the Constitution in order to prevent states from denying the right to vote based on “race, color, or previous condition of servitude.”
When white ex-Confederates regained control of Mississippi’s government in the late 19th century, they refused to accept the new suffrage guarantee. Instead, they exploited a loophole: the Constitution says states cannot prevent black voters from participating in elections — it didn’t say anything about political parties. They argued that primary elections were votes among the “private” party membership. Therefore political parties could legally exclude African American voters from their primaries.
In 1902, Mississippi adopted a law officially sanctioning a whites-only primary system. Nearly every state and local office — from governor down to county coroner — allowed candidates to declare a partisan affiliation. They all ran as Democrats. With no opposition, the Democratic primary became the de facto general election. This voided the political power of the few black voters who were able to register in spite of poll taxes, literacy tests, and physical threats.
Mississippi’s whites-only primary came to an end after the integrated Mississippi Freedom Democratic Party secured the national party’s support through direct action at the 1964 Democratic National Convention. The Voting Rights Act, passed the next year, finally enforced the 15th Amendment in the South. By 1968, Mississippi had the highest share of black voters in the country. Almost all became Democrats.
Soon thereafter, whites began voting en masse for the Republican Party in statewide elections. Mississippi last supported a Democratic presidential nominee in 1976, and no Democratic candidate for governor has won a majority of the popular vote since 1987. (Ronnie Musgrove, in 1999, fell short of 50 percent and was ultimately elected by the House of Representatives.) Today, the GOP controls seven of eight statewide offices, both U.S. Senate seats, three of four Congressional seats, and both chambers of the Legislature.
However, the partisan realignment has not filtered all the way down the ballot. Many local officials still run as Democrats out of tradition if not ideology. In a large number of races, all of the challengers run as Democrats, too, so if you want to have a say in local affairs, you must vote in the Democratic primary — even if you plan to vote for the Republican slate in November.
This effect has waned in recent years as local offices in suburban GOP strongholds such as DeSoto, Madison, and Rankin counties have gone red. But in many counties, the Democratic primary remains the only show in town.
Which brings us to what happened last Tuesday.
The Purple Primary
Elections are all about turnout, and in Mississippi, most turnout is local. Despite the GOP’s popular majority and hotly-contested races for state auditor and treasurer, the Republican primary still had 15,000 fewer voters than the Democratic primary.
Of the 562,794 voters who showed up, 288,686 (51 percent) chose the Democratic table. That’s actually a considerable decline in the Democratic advantage from previous primaries, but primary turnout has no predictive power for the general election. The Democrats’ gubernatorial nominees in 2007 and 2011 finished with 42 and 39 percent of the vote, respectively.
In both elections, the Democratic nominees received fewer total votes in November than the aggregate cast for Democratic candidates in the primaries.
The Democratic turnout advantage is only present when local races are on the ballot. In federal races, Democratic primary turnout drops precipitously.
Here is a graph of the party turnout during the past four major statewide primaries: the state/local races in 2011 and 2015 and the federal races in 2012 and 2014. I’ve included the vote total for the top statewide office in each of those elections.
Democratic turnout in the 2015 state/local primary was more than three times as large as in the 2014 and 2012 federal primaries. The 85,000 Democratic faithful who showed up to vote in those low-key House and Senate primaries offer a minimum approximation of how many loyal Democrats would show up to vote if only statewide offices were on the ballot. Under even the most conservative estimates, it’s likely that a majority of Democratic primary voters turned out because of the bottom of the ticket rather than the top.
On the other hand, you’ll notice that Republican turnout has been consistent across state/local and federal primaries. The Republicans have had competitive races in each of those years, but it’s clear that GOP diehards turn out for every primary, rain or shine, without much crossover voting to the Democratic primary. (This mainly serves to dispel the conspiracy theory that Gov. Bryant’s Republican loyalists plotted to vote for the weakest Democratic candidate.)
The relative turnout figures suggest that most of the elasticity in the Democratic primary comes from loosely-attached Republicans and independents who cross over to vote in local races. While the effect of crossover voting is declining on a statewide level, it is still prevalent in many counties: the Democratic primary received at least 75 percent of the voters in 47 of Mississippi’s 82 counties. Of those 47 counties, only 20 voted for the Democratic gubernatorial nominee in 2011.
Three-fifths of counties had more than a 25 percentage point gap between the Democratic share of the primary and general election vote. Those counties accounted for 68 percent of the total Democratic primary turnout in 2015 — but only 39 percent of the Democratic nominee’s votes in 2011.
The size of the dropoff between August and November can be staggering. Nearly 90 percent of Tippah County voters participated in the Democratic primary, but only 25 percent cast their ballot for the Democratic nominee in 2011. Ninety-two percent voted in Carroll County’s Democratic primary, but only 29 percent went blue in 2011. In total, 17 counties (21 percent) have at least a 50-point gap between Democratic share of the primary and general election vote.
Explore the map for yourself. The counties shaded in blue vote more Democratic in August than November. Red counties vote more Republican.
The Gray Primary
So what does all of this have to do with Robert Gray? Well, plenty.
It is logical to assume — a dangerous thing to do with politics, I’ll admit — that most voters who choose to cast a ballot for a mystery candidate have little or no familiarity with anyone who is running. Since Democratic base voters are more likely to follow the party’s candidates and receive contact from their campaigns, a majority of the loyalists probably voted for Slater or Short. Conversely, those who were least engaged with the Democratic campaigns — specifically, the independents or loose Republicans who crossed over to vote in their local races — likely voted in greater numbers for Gray. (Note: I’m merely talking about probabilities here; many loyal Democrats voted for Gray, and many crossover voters did not.)
If that assumption is correct, we’d expect Gray to average larger majorities in the counties with the widest voting disparities between Democratic primary and general election, and we’d expect him to fall short in areas with more proportional turnout.
That’s exactly what happened.
Gray won an outright majority in 44 of the 50 counties (88 percent) in which the Democratic share of the primary vote was at least 25 percentage points greater than the 2011 general election vote share. He received 55 percent of the total vote in these high-crossover counties. They accounted for 71 percent of his primary votes, but — as I mentioned in the previous section — only 39 percent of the Democratic general election votes.
Gray only carried a majority in 10 of the 32 counties (30 percent) where turnout was more proportional. He won 39 percent of the total vote in these low-crossover counties, which produced the remaining 61 percent of the Democrats’ 2011 general election votes.
A linear regression analysis confirms a statistically significant correlation between a county’s crossover voting and Gray’s level of support: Gray’s margin of victory over his opponents’ combined total grew by one percentage point for every five points that the 2015 Democratic primary turnout exceeded the 2011 general election vote. The rate of crossover voting accounted for 40 percent of the variance in Gray’s county-to-county performance.
In effect, the crossover voters diluted the preference of the voters who will support the party’s nominee in November.
This isn’t to say that loyal Democrats didn’t also succumb to uninformed voting. Surely many did. Gray still racked up nearly 40 percent of the vote in low-crossover counties — more than enough to get into a runoff against one of his opponents. The state Democratic Party organization (or lack thereof) bears responsibility for that. No established candidate was willing to run for the state’s top office, and little financial or messaging support was available for the candidates who jumped in.
But even in the Democrats’ atrophied condition, it would have been almost impossible for Gray or any unknown candidate to have won a majority of the vote without the systemic distortion and dilution of the electorate created by Mississippi’s antiquated primary laws and traditions.
Featured photo courtesy of Robert Gray for Governor Facebook page.