The Evolution of Arizona’s Legislative Map, 2012–2018

When Kyrsten Sinema and Kathy Hoffman won their historic statewide victories in 2018, they also carried a bare majority of the Arizona’s state legislative districts. If Democrats can translate that success to the legislative level going forward, they’d put the capitol in play. They already nearly did so in in 2018: at the state senate level, Democrats carried thirteen of Sinema and Hoffman’s sixteen districts, and lost in the remaining three districts by a total of less than 5,000 votes (winning a cumulative 49% of the vote between the three).

This map and much of the data used here is from Drew Savicki. Other data is drawn from the Daily Kos Elections database.

It’s no surprise then, that Democrats are aiming to win legislative majorities in 2020, a significant shift from where things stood when the current legislative map was drawn. Back in 2012, Barack Obama carried only twelve legislative districts.

This piece traces the evolution of the legislative map from 2012, when Democrats’ best hope was for a string of luck that might tie a legislative chamber, to today, where majorities in both chambers are legitimately in play.

An early digression: Kyrsten Sinema and Kathy Hoffman’s historic victories relied on remarkably similar geography. They were within two points of each other in 29 legislative districts, and within one point of each other in 20. This is unusual, but not totally unique. In 2014, Terry Goddard and Felicia Rotellini experienced a similar degree of similarity in their coalitions. The two were within two points of each other in 28 legislative districts, within one point in 17.

The Pre-Trump Balance of Power

Few, if any, careful observers thought the Republicans’ legislative majorities were in serious danger before Donald J. Trump came along. There were few true swing districts — not unusual for a legislative map in today’s polarized, geographically sorted political environment — and Republicans had enough safe districts to virtually guarantee they would hold power for the foreseeable future.

For present purposes, I’m defining as a swing district those districts that either: 1) voted for different parties in the statewide races from 2012–14 for which I have legislative district data, or 2) gave a Democrat between 48–52% of the two-party vote in at least two of those statewide races. Only three districts met those criteria in the 2012–14 period: LDs 8, 18, and 28 (the “crazy eights”).

That left 15 solid Republican districts and 12 solid Democratic ones, meaning Dems would need to sweep the swing districts simply to tie a chamber. They never did so, instead holding one swing Senate seat (LD8) and only one swing House seat (in LD28) during this period. Making matters worse, Democrats could not count on their own solid districts: Republicans captured a House seat in “safe” blue territory in each cycle (LD9 in 2012, LD2 in 2014).

The Trump Era and Contested Partisan Control

Over the latest two cycles, the battleground map has shifted only marginally, but enough to change the fundamental calculus of partisan control of the legislature, resulting in knife’s edge majorities in both chambers.

Using the same criteria, there are now four swing districts for the 2016–18 period, with an even split of 13 solid districts for each party. LDs 8 and 18, districts in many ways emblematic of the salient cultural divides of the Trump Era, have moved solidly into their respective partisan camps. LD28 stays in the swing category thanks to Ducey’s comfortable win there, although voters there now seem to have a definite preference for Democrats on average. Joining LD28 are three districts that carried a distinctive red tinge during the pre-Trump period: LDs 6, 17, and 20.

To tie a chamber, Democrats needed to hold their own solid district and split the four swing districts. Any stronger performance would mean a majority.

Democrats continued the trend of holding all of their solid Senate seats, but in 2016 again they lost solid House seats. This time they lost two, in LDs 10 and 18, because they did not try to win the second seat in LD18, which had only become solidly Democratic that cycle. Half of their four-seat gain in 2018 came from consolidating these solid Democratic seats.

The other half of the gains came from the swing House seats. In 2016, Democrats captured only one of eight (in LD28), but in 2018 they added seats in LD17 and LD28, and nearly captured seats in LDs 6 and 20 as well. They have still failed to win a single swing Senate seat in any cycle this decade.

Tracing the Swing Districts Through the Decade

Lastly, here’s a summary table showing how the six current and/or former swing districts have evolved over the course of the decade. No two have followed quite the same path.

• LD18 started out as the most competitive at the statewide level, and jumped further towards the Democrats in both 2014 and 2016. That pattern was repeated at the state senate level, but delayed by a cycle. The two levels of the ballot only came into alignment in 2018 (that delay will be a trend across other districts).
• LD8 was the second most competitive statewide in 2012, and Sen. Barbara McGuire’s over-performances kept it in the Democratic column in 2012 and 2014. In 2016, however, the district lurched too far into the red for McGuire to hold on. In 2018, without McGuire on the ballot, the state senate results came into alignment with the statewide ones.
• LD28 was essentially as competitive statewide as LD8 back in 2012. While it lurched to the left in 2014, Sen. Adam Driggs was able to keep it solidly in the Republican column in 2014. Once again, it took an extra cycle for the legislative results to begin to reflect the statewide ones. But even still, Sen. Kate Brophy McGee has over-performed in both 2016 and 2018, barely holding on as the district has plateaued at the statewide level. Will we look back on the Brophy McGee as the suburban Republican parallel to McGuire? That may depend on LD28 taking one more statewide jump to the left, as LD8 did (but to the right) in 2016.
• LD20 didn’t look too promising for Democrats until after the 2018 cycle. While it was never deep red, it had largely flatlined at both the statewide and state senate levels from 2012–2016. Yet, in 2018 the district took a leap towards competitiveness at both levels.
• LD6 is the one competitive district where Democrats have consistently done better at the legislative level than in statewide races, especially since Sen. Sylvia Allen’s first election in 2014. Each state senate race since has been very close (as was the state house race in 2018) even while statewide Democrats failed to win here. Thus far, statewide Democrats have done better in midterm cycles, perhaps reflecting the difficulties facing the national Democratic brand in the rural parts of the district. Whether 2020 breaks this pattern, and whether Democratic legislative candidates continue to over-perform the statewide ticket, could well determine control of one of both legislative chambers in 2020.
• Finally, LD17 was the reddest of these districts in 2012 and 2014 at the statewide level, and Democratic state senate candidates were brought down by the top-ticket gravity. Like LDs 18 and 28, this district swung hard to Democrats in 2016 and held at about that same level in 2018. Also like those other districts, it took until the second cycle of competitiveness for the legislative races to match the district’s new competitive bent.