Split Ticket

On a night of surprisingly strong Democratic performances in the 2023 elections, Brandon Presley’s 3.2% loss in Mississippi’s gubernatorial election stands out. Facing an incumbent governor in a Trump +17 state in the Deep South, most had already written off the Presley campaign. Then, a catastrophic October Democratic underperformance in demographically-similar Louisiana made Presley’s chances look even slimmer.

Presley’s surprise overperformance was powered by three factors: strong Black turnout, exceptional Black support, and relatively strong white support. Tapping into the white electorate was one of the biggest reasons that Mississippi was so much closer than Louisiana. Furthermore, while Black turnout was abysmally low in the latter state, it was remarkably robust in the former.

In fact, the counties with the most Black electorates had the highest turnout rates in Mississippi relative to 2020 (a feat in and of itself, considering the comparatively low Black turnout in 2022). The month before in Louisiana, the reverse was true: relative change in turnout was negatively correlated with a county’s Black percentage.

Base turnout means little without maintaining high levels of support, however, and Presley’s margins with every group of voters may have set modern Democratic records for performances in the Magnolia State. We estimate that Presley won the support of 96% of Black voters and 22% of white voters, both of which represent an improvement on President Joe Biden’s 94% and 17% vote shares, respectively. He also got excellent turnout — our best estimate is that the 2023 Mississippi electorate probably would have voted for Trump by about 11 points in 2020, meaning it was significantly friendlier to Democrats than the actual 2020 electorate (which was Trump +17).

But while it may be somewhat unsurprising to see Presley outdo Biden’s anemic performance in the state, his improvement on 2019 Democratic gubernatorial nominee and then-Attorney General of Mississippi Jim Hood is significantly more impressive, especially considering that Hood himself only lost by 5.1%. Notably, Presley managed to mostly match Hood’s margins with whites while improving on his already-high margins among Blacks. In majority-Black counties, Presley got 73% of the vote. Given the simultaneous increase in relative turnout, this is a significant improvement on Hood’s 71% and Biden’s 68%, and it is also his largest overperformance relative to Hood’s numbers for any demographic grouping of counties (see below).

In Louisiana, however, the story was the exact opposite, as Democrats continued to decline relative to Biden’s 2020 margins. This cannot be attributed solely to lower Black turnout; in fact, it is highly likely that a significant and abnormally high fraction of Black voters actually voted for the Republicans.

This begs the question: why did Mississippi buck the trends seen in Louisiana, and what drove Presley’s overperformances?

Firstly, as we noted in our post-election recap, Democrats made an effort to invest in Mississippi, whereas Republicans outspent them 26:1 in Louisiana. Presley also had an exceptional voter outreach program among Black voters, with numerous campaign events at Black colleges and churches, ensuring he could capitalize on Tate Reeves’ weaknesses with them.

Secondly, while the reaction to the Supreme Court’s ruling in Dobbs v. Jackson has helped Democrats electorally, it has actually helped Republicans in a few highly religious areas; Mississippi and Louisiana are just two of them. Presley’s strong pro-life stance would have been campaign kryptonite in most regions, but in Mississippi, it probably helped him diffuse the issue of abortion to an extent unachievable for a  pro-choice Democrat.

The Impact of Felony Disenfranchisement

Not completely unrelated to the final gubernatorial margin are the extremely stringent felony disenfranchisement laws found in Article XII of the Mississippi Constitution of 1890. One such provision is Section 241, which disenfranchises felons for life even after the completion of sentences. The related Section 253 limits felons’ recourse to restoration of their voting rights. Both provisions were included in the Reconstruction-era constitution to disenfranchise Black Mississippians. The effects of Mississippi’s racist motivations are still visible today — between 55 and 60% of Mississippi’s disenfranchised residents are Black even though the state’s total Black population is just 38%.

In August of 2023, the traditionally conservative Fifth Circuit Court of Appeals struck down Mississippi’s lifetime felony voting ban (Section 241) as an unconstitutional violation of the Eighth Amendment’s prohibition on the imposition of “cruel and unusual punishments”. But a month later, the collective Fifth Circuit vacated the panel opinion pending rehearing en banc. This maintained the preexisting disenfranchisement laws for the November elections. 

We have seen suggestions regarding whether these laws may have changed the outcome of the election. While these questions are unanswerable at best, we can still test whether this is a plausible theory.

Reeves won by 26,619 votes. According to a recent study from The Sentencing Project, a racial justice-focused organization, 239,209 people are banned from voting for life due to felony convictions. That works out to roughly 11% of the state’s voting-age population, and the study estimates that 54% of those disenfranchised felons are Black. Other analyses place the number even higher (around 60–61%), but for the purposes of our analysis, we will begin with the 54% figure.

Not all of these newly enfranchised voters would vote, of course; according to pollster John Couvillon, overall turnout in Mississippi was roughly 40% of registered voters, and ex-felons tend to have a much lower turnout rate than the rest of the population. Extrapolations of turnout rates from Jeff Manza and Christopher Uggen’s 2006 work would give a turnout estimate of 24% among newly eligible former felons*, which would mean that we would see 31,320 new Black voters and 25,116 non-Black voters (a population that is almost entirely white).

If we projected some margin estimates onto these new electorates, we could, in theory, estimate their effect on the outcome. Split Ticket’s analysis suggests Presley won Black voters by 94 and lost white voters by 54. It is tough to be confident in the leans of ex-felons, as they tend to be a fairly unique group of voters, but using these splits for them would narrow Reeves’ margin to 11,330 votes and 49.96% of the overall vote, which is just under the 50% required to avoid a runoff. Using these assumptions, it is quite plausible that a maximally enfranchised electorate would have denied Tate Reeves an immediate victory and sent the race to a similarly-competitive runoff election.

But we hesitate to make any sweeping conclusions here, because we think those turnout estimates are on the higher side. In practice, restored felons tend to turn out at significantly lower rates, as Traci Burch found in 2011, due to a variety of factors such as fines, fees, and an unawareness that they could even vote again. Dialing the turnout down to a more reasonable 13% (more in line with findings from other studies, like the 2015 study from Meredith and Morse) would place Reeves at 50.4%, which is just above the runoff threshold. He would have beaten Presley by just over 2% in this scenario, suggesting that the current impact of felon disenfranchisement was no more than 1%, on margin, in Mississippi.

It is possible that a full restoration of rights, a heavy campaign regarding these restorations, and a removal of other barriers to voting (such as not conditioning re-registration on the payment of fines and fees) would have resulted in the higher turnout figure established above. However, an effort to reform the re-enfranchisement process would almost certainly not immediately result in the beneficial increases in turnout and political involvement this article describes.

Given the above numbers, we thus believe it to be fairly unlikely that the August 2023 decision restoring the rights of ex-felons would have been enough to tip the Mississippi gubernatorial election to a runoff even if it were allowed to stand, especially given the short timeframe between the order and the actual election. However, if these laws and the associated obstacles had never been on the books in the first place, and if these voters were actually aware of their right to vote, it is much more plausible that Reeves could have been pushed to a runoff.

It goes without saying that the racist disenfranchisement policies established decades ago continue to affect the modern political process. Regardless of the specific makeup of the disenfranchised or their propensity to vote, the lost opportunities to participate in the political process are apparent, especially in close races, and their ability to make their voice heard is silenced.

*The 2006 book from Manza and Uggen studies the question of ex-felon turnout and suggests that roughly a third of ex-felons would have cast ballots in the 2000 election if they were restored the right to vote. The 2000 election had 54% voter turnout; crudely normalized, this would imply that roughly 24% of ex-felons in Mississippi would have cast a ballot in 2023.