Skip to main content
Menu
What Virginia, Mississippi and Kentucky Can Tell Us About 2020

The 2019 elections are in the books, and now the country can finally start paying attention to 2020. In many ways, Tuesday’s elections were a dress rehearsal for those a year from now: The parties’ performance relative to partisanship can tell us which way the political winds might be blowing, and obvious trends from 2019 — like the widening urban-rural divide — provide clues as to where the battles of 2020 will be fought.

As I wrote in April and again in September, special elections in 2019 have not shown the same clear, overwhelming pattern of Democratic dominance as they did in 2017 and 2018. But the sample size was relatively small — until Tuesday, when Virginia held elections in all 40 of its state Senate seats and all 100 of its House of Delegates seats. Although they’re not full-fledged special elections, you can analyze these races in much the same way, looking for how each party performed relative to the base partisanship of the districts. And in Virginia, Democrats not only did well in absolute terms (they flipped two Senate seats and six House seats), but they also outperformed the FiveThirtyEight partisan leans1 of their districts in most of those races. (Elections also took place on Tuesday for the Mississippi state Senate, Mississippi state House and New Jersey General Assembly. However, we did not include these races in our analysis because 2016 and 2012 presidential results by legislative district are unavailable in Mississippi and because New Jersey Assembly districts elect two legislators each, with all candidates in the district running in the same race.2 This irregular arrangement makes Assembly races too different from typical elections to compare.)

Specifically, in the Virginia state Senate, the average winning margin was 4 points more Democratic than the average partisan lean.

How Democrats did in the Virginia state Senate

The 2019 election results relative to each district’s partisan lean

2019 Vote Share
District Dem. Rep. Margin Partisan Lean Dem. Margin Swing
SD-32 74% 26% D+47 D+30 +17
SD-06 60 40 D+19 D+8 +11
SD-13 54 46 D+9 R+1 +9
SD-03 38 62 R+23 R+32 +8
SD-08 48 52 R+4 R+12 +8
SD-39 66 34 D+32 D+24 +7
SD-33 65 35 D+30 D+23 +7
SD-11 45 55 R+10 R+17 +6
SD-14 40 60 R+20 R+26 +5
SD-12 49 51 R+2 R+7 +5
SD-17 48 52 R+3 R+8 +5
SD-10 54 46 D+9 D+5 +4
SD-07 49 50 R+1 R+5 +4
SD-28 42 57 R+15 R+18 +3
SD-26 35 65 R+30 R+31 +2
SD-04 37 63 R+26 R+26 EVEN
SD-27 36 64 R+29 R+29 EVEN
SD-19 28 71 R+43 R+40 -3
SD-24 28 71 R+43 R+38 -5
SD-15 32 68 R+37 R+30 -6
SD-22 37 63 R+26 R+17 -8
Average +4

Includes only races where both a Democrat and a Republican were on the ballot. Results are unofficial and as of noon Eastern on Nov. 7.

FiveThirtyEight’s partisan lean metric is the average difference between how a district votes and how the country votes overall, with 2016 presidential election results weighted at 50 percent, 2012 presidential election results weighted at 25 percent and results from elections for the state legislature weighted at 25 percent. Partisan leans do not incorporate the results of this year’s elections.

Sources: Virginia Department of Elections, Daily Kos Elections

In the House of Delegates, the average winning margin was 3 points more Democratic.

How Democrats did in the Virginia House of Delegates

The 2019 election results relative to each district’s partisan lean

2019 Vote Share
District Dem. Rep. Margin Partisan Lean Dem. Margin Swing
HD-04 37% 63% R+26 R+55 +29
HD-97 27 56 R+29 R+45 +15
HD-68 55 45 D+11 R+1 +12
HD-96 46 53 R+6 R+18 +12
HD-28 52 48 D+4 R+6 +10
HD-21 55 45 D+9 EVEN +9
HD-72 53 47 D+7 R+3 +9
HD-12 54 46 D+7 R+2 +9
HD-91 55 45 D+10 D+1 +9
HD-27 50 50 R+1 R+9 +9
HD-25 40 58 R+18 R+27 +9
HD-93 56 44 D+11 D+3 +8
HD-34 58 42 D+17 D+10 +7
HD-87 62 38 D+24 D+17 +7
HD-44 71 29 D+41 D+34 +7
HD-39 68 32 D+37 D+30 +7
HD-42 60 40 D+19 D+12 +7
HD-23 36 64 R+28 R+35 +7
HD-51 55 45 D+9 D+2 +7
HD-02 61 39 D+22 D+15 +7
HD-10 52 48 D+5 R+2 +7
HD-20 41 58 R+17 R+24 +7
HD-73 52 48 D+5 R+2 +7
HD-88 44 56 R+12 R+18 +6
HD-26 46 54 R+8 R+14 +6
HD-18 39 60 R+21 R+27 +6
HD-85 52 48 D+3 R+2 +6
HD-84 49 51 R+2 R+8 +6
HD-33 43 57 R+14 R+19 +5
HD-94 58 40 D+17 D+12 +5
HD-81 48 52 R+4 R+9 +5
HD-55 40 60 R+20 R+25 +5
HD-40 52 48 D+5 D+1 +4
HD-13 56 44 D+12 D+9 +3
HD-65 35 65 R+30 R+33 +3
HD-59 37 63 R+27 R+29 +3
HD-82 41 59 R+19 R+21 +3
HD-56 39 61 R+22 R+25 +3
HD-52 73 27 D+46 D+44 +3
HD-24 33 66 R+33 R+36 +2
HD-29 36 64 R+29 R+31 +2
HD-76 56 44 D+13 D+11 +2
HD-64 40 60 R+20 R+21 +1
HD-31 53 47 D+5 D+4 +1
HD-06 23 77 R+54 R+54 EVEN
HD-83 50 50 EVEN EVEN EVEN
HD-08 34 66 R+33 R+32 -1
HD-50 53 46 D+7 D+8 -1
HD-07 33 67 R+34 R+32 -1
HD-54 42 58 R+16 R+14 -2
HD-62 45 55 R+10 R+9 -2
HD-15 26 74 R+49 R+46 -3
HD-22 31 69 R+38 R+34 -4
HD-58 37 62 R+25 R+21 -4
HD-100 48 52 R+4 EVEN -4
HD-99 38 62 R+25 R+21 -4
HD-66 47 52 R+5 EVEN -5
HD-98 31 69 R+38 R+32 -6
HD-75 51 49 D+2 D+8 -6
HD-14 39 61 R+23 R+10 -12
HD-61 33 67 R+34 R+21 -13
HD-60 34 66 R+33 R+15 -17
Average +3

Includes only races where both a Democrat and a Republican were on the ballot. Does not include HD-80, because an independent received more than 10 percent of the vote. Results are unofficial and as of noon Eastern on Nov. 7.

FiveThirtyEight’s partisan lean metric is the average difference between how a district votes and how the country votes overall, with 2016 presidential election results weighted at 50 percent, 2012 presidential election results weighted at 25 percent and results from elections for the state legislature weighted at 25 percent. Partisan leans do not incorporate the results of this year’s elections.

Sources: Virginia Department of Elections, Daily Kos Elections

Because partisan lean is meant to represent a district’s partisanship in a neutral political environment, that’s consistent with a somewhat Democratic-leaning national mood, like the one we see in polls of the generic congressional ballot — though not nearly as Democratic-leaning as 2018, when Democrats won the House popular vote by almost 9 points.

Any discussion about Virginia and partisan lean, though, should include the caveat that the state is well-educated (only five states and D.C. have a higher share of the population over age 25 with a bachelor’s degree or higher) and highly suburban (nine of its 11 congressional districts are classified as some level of suburban, according to CityLab) — traits that might make it more likely to rebel against the Republican Party in the age of Donald Trump. Indeed, most of the few seats where Republicans did better than partisanship would suggest were in more rural areas. Meanwhile, districts Democrats flipped were largely suburban.

But that in and of itself is an important takeaway of the 2019 elections: Rural areas got redder, and urban and suburban ones got bluer — and not only in Virginia. Even for centrist Democrats like Mississippi gubernatorial candidate Jim Hood, the old, pre-Trump Democratic coalition has been replaced by one that increasingly relies on suburban voters to make up for losses among rural whites. Four years after being elected Mississippi attorney general with 55 percent of the vote, Hood won just 47 percent on Tuesday — yet he still managed to improve his vote share in DeSoto County (the Memphis suburbs) by 6 points, more than in any other county. And Madison County (the Jackson suburbs), the county with the highest per-capita income in the state, was the one county he lost in 2015 but carried in 2019. Meanwhile, his vote share dropped precipitously in the rural white counties of northeast Mississippi — from 54 percent to 21 percent in Itawamba County, for example.

In Kentucky, Democratic gubernatorial candidate Andy Beshear turned in a more impressive performance overall, turning a 9-point Democratic loss in the 2015 race into (probably) a half-point win. According to an exit poll conducted by a consortium of political scientists,3 he did so by winning 16 percent of Republican voters and defeating (admittedly unpopular) Republican Gov. Matt Bevin by 27 points among independents. But Beshear’s strongest overperformances relative to his 2015 counterpart, Jack Conway, came in urban and suburban areas. Beshear won Jefferson County (home of Louisville) with a whopping 67 percent of the vote, while Conway won Jefferson with just 58 percent. And perhaps most impressively, Beshear won a combined 47 percent in the three most populous counties in the Cincinnati suburbs: In 2015, Conway got only 37 percent across Boone, Campbell and Kenton counties. In the rest of the commonwealth — mostly rural areas — Beshear won 42 percent of the vote, not too much better than Conway’s 39 percent. So it’s fair to say that Beshear won where Conway lost because of his strength in metropolitan areas.

Elsewhere in the country, Democrats’ suburban victories seemed to confirm that there is still ground for the party to gain in suburban congressional districts, even after the inroads it made in 2018. In Hamilton County, Indiana, for instance, Democrats unexpectedly won three seats on city councils that have not had Democratic members in living memory. These Indianapolis suburbs are the anchor of Indiana’s 5th Congressional District, whose GOP representative is retiring in 2020. In the suburbs southwest of St. Louis, Democrats flipped Missouri House District 99 in a special election. The district had a FiveThirtyEight partisan lean of R+11, and the Democratic candidate won by 8 points, making for a dramatic 19-point margin swing. If the same thing happens in Missouri’s 2nd Congressional District, which covers Missouri House District 99, it would be enough to flip the seat from red to blue (the 2nd District’s FiveThirtyEight partisan lean is R+16). Finally, in the northern suburbs of Philadelphia, Democrats put Republican Rep. Brian Fitzpatrick on notice by winning every row office and a majority on the board of commissioners in Bucks County, which makes up most of Fitzpatrick’s Pennsylvania 1st Congressional District. (Democrats also seized control of county government in the other two major Philadelphia collar counties — Chester and Delaware — for the first time in more than 150 years.)

There’s one last lesson that the 2019 results suggest about 2020, but it’s one that we already knew: Turnout is likely to be through the roof. In Kentucky, we estimate based on preliminary data that 43 percent of the voting-eligible population cast a ballot for governor; not only is that much, much higher than the 30 percent of the voting-eligible population that we estimate turned out in 2015, but it’s also higher than the 42 percent who voted in the Senate race in the regular 2010 midterm election. In Virginia, we already know that more people voted than in any state-legislative-only election since at least 1976 — and The Washington Post estimates that there could be thousands of votes left to count. If we see a corresponding spike in turnout between 2016 (already a pretty high-turnout election by recent standards) and 2020, polling places could be overwhelmed with voters. Americans are telling pollsters that their levels of interest in the upcoming election are at unprecedented highs — and according to one recent poll, they are already more excited about voting than they were on the eve of the 2016 and 2012 elections!

Geoffrey Skelley contributed research.

Footnotes

  1. FiveThirtyEight’s partisan lean metric is the average difference between how a district votes and how the country votes overall, with 2016 presidential election results weighted at 50 percent, 2012 presidential election results weighted at 25 percent and results from elections for the state legislature weighted at 25 percent. The partisan leans in this article do not incorporate the results of this year’s elections.

  2. The top two vote-getters are elected.

  3. One of the survey directors, Dan Hopkins, is a FiveThirtyEight contributor.

Nathaniel Rakich is FiveThirtyEight’s elections analyst.

Comments