← Back to map

About This Map

Methodology and data sources for the RTA Special Election Results Map

What this map shows

This interactive map displays precinct-level results from the Pima County RTA Next special election held on March 10, 2026. The election included two propositions:

Proposition 418 (Plan) — Authorizes the RTA Next regional transportation plan, a $2.67 billion, 20-year program of roadway, transit, safety, and environmental projects across Pima County.

Proposition 419 (Tax Authorization) — Authorizes a half-cent sales tax to fund the plan.

Both propositions passed. The map offers four views: individual yes-vote-share maps for each proposition, a bivariate comparison showing where the two diverged, and a voter turnout view.

Data sources

Three primary datasets are used:

1. Election results — Official certified precinct-level results from Pima County Elections, containing registered voter totals, ballots cast, and yes/no/over/under vote counts for both propositions across 265 precincts. This dataset contains all 209,533 ballots. All yes and no percentages on the map use valid votes (yes + no) as the denominator, excluding over votes and under votes. Across the county, over and under votes account for about 1.0% of ballots on Prop 418 and 1.4% on Prop 419.

2. Geographic boundaries — Three shapefiles from Pima County GIS and City of Tucson GIS provide precinct boundaries (265 polygons), jurisdiction boundaries (City of Tucson, Oro Valley, Marana, Sahuarita, South Tucson, and unincorporated Pima County), and City of Tucson ward boundaries (Wards 1–6).

3. Registered voter file — A county spreadsheet listing registered voter counts broken out by precinct, ward, and jurisdiction. This is the key file that allows vote apportionment (explained below).

Why apportionment is needed

Election results are reported by precinct. But the map also shows results by ward and jurisdiction. The challenge is that precinct boundaries do not nest cleanly within ward or jurisdiction boundaries:

• 8 Tucson precincts are split across multiple wards (e.g., a precinct might be 70% in Ward 2 and 30% in Ward 3).

• 2 precincts span multiple incorporated jurisdictions.

Because the election results only report totals per precinct — not per ward or jurisdiction — we need a method to apportion each split precinct's votes to the wards or jurisdictions it overlaps.

Apportionment method: registered voter weighting

For each precinct that spans multiple wards or jurisdictions, we use registered voter counts as the apportionment weight. The county's registered voter file provides the exact number of registered voters in each precinct-ward and precinct-jurisdiction intersection.

Example — Precinct 153

Suppose Precinct 153 has 2,000 registered voters total, with 1,400 in Ward 2 and 600 in Ward 3. Its election results show 800 ballots cast and 480 yes votes on Prop 418.

The ward weights are:
• Ward 2: 1,400 / 2,000 = 70%
• Ward 3: 600 / 2,000 = 30%

We apportion the precinct's results proportionally:
• Ward 2 gets: 800 × 70% = 560 ballots, 480 × 70% = 336 yes votes
• Ward 3 gets: 800 × 30% = 240 ballots, 480 × 30% = 144 yes votes

Each ward's totals are the sum of these apportioned amounts across all precincts.

Why registered voters, not area?

An earlier approach apportioned votes based on the geographic overlap between precincts and wards (areal weighting). This assumes voters are evenly distributed across the land area of a precinct — a poor assumption when a precinct includes both dense urban neighborhoods and empty desert or parkland.

Registered voter counts directly measure where voters actually live within each precinct. This produces much more accurate ward and jurisdiction totals, particularly for wards like Ward 5, where areal weighting severely undercounted voters because much of the precinct land area in that region is sparsely populated.

Validation: When jurisdiction-level apportioned ballots are summed, they equal the county-wide total of 209,533 exactly. Ward-level totals sum to the City of Tucson total. No votes are lost or double-counted.

Precinct-level numbers (direct, not apportioned)

When you click on an individual precinct, the popup shows exact reported numbers from the election results file — registered voters, ballots cast, yes votes, no votes, and margin. Yes percentages are computed as yes / (yes + no), excluding over and under votes. These precinct-level numbers are not apportioned or estimated. The apportionment method only applies to the ward and jurisdiction breakdowns shown in the sidebar and in the precinct popup's "by ward" or "by jurisdiction" subsections.

Bivariate comparison view

The "418 vs 419" view uses a 3×3 bivariate choropleth. Each axis is divided into terciles (three equal groups of precincts):

VariableLowMidHigh
Prop 418 yes %< 56.3%56.3 – 64.7%> 64.7%
Prop 419 yes %< 54.2%54.2 – 62.0%> 62.0%

Each precinct is classified into one of nine cells based on where it falls on both axes. Here's how to read the legend:

Prop 419 yes % Low High Low Prop 418 yes % High A B C D
A — Low 418 / High 419
Voters favored the tax more than the plan
B — High 418 / High 419
Strong support for both propositions
C — Low 418 / Low 419
Weak support for both propositions
D — High 418 / Low 419
Voters favored the plan more than the tax

The most informative precincts are the off-diagonal ones — teal precincts where voters supported the plan but not the tax, and pink precincts where the reverse was true. Most precincts in this election fall along the diagonal (gray to purple), since the two propositions tracked closely, but the spatial pattern of divergence reveals where the tax was a harder sell than the plan itself.

Turnout view

Turnout is calculated as ballots cast divided by registered voters for each precinct. The amber-to-red color scale runs from 10% to 50%+. County-wide turnout was approximately 31.7%. Jurisdiction and ward turnout figures use the same registered-voter apportionment method to allocate registered voter counts to each geography, then divide the apportioned ballot totals by the apportioned registered voter totals.

Ward council members

WardCouncil member
Ward 1Lane Santa Cruz
Ward 2Paul Cunningham
Ward 3Kevin Dahl
Ward 4Nikki Lee
Ward 5Selina Barajas
Ward 6Miranda Schubert

Contact

Questions about this map or the underlying data can be directed to City of Tucson staff. For official certified results, contact Pima County Elections.