Trepp | Local Ballot Measures Delivery Strong Voter Pushback on Data Center Projects

Data center developers faced significant opposition in the November 3 elections. Of the 38 local data center measures reviewed by Trepp, 37 went against developers, with voters approving 25 restrictions and rejecting 12 measures that would have supported development. Three additional measures remain pending.

Other development measures produced more mixed results. Of the 29 local measures with a clear outcome or direction, 11 (38%) went developers’ way, including six of 10 measures that would expand housing or multifamily development. The contrast highlights how voters are approaching different types of development, with some communities supporting new housing while others are strengthening restrictions or overturning projects previously approved by city councils.

Trepp reviewed local development measures across five states, examining how voters are shaping what can be built and where. While several California results remain pending as mail ballots continue to be counted, current returns point to a clear geographic divide in support for new development.

Table 1 breaks these votes down by type of measure and by whether the outcome went developers’ way. Measures that would directly add homes or multifamily space did best. The four of them that lost or are trailing are in coastal towns or communities at the edge of growth. Measures tightening the rules were developers’ worst category.

Table 1: November 3 Local Development Measures

Source: Trepp compilation of unofficial returns from county and city election offices and local reporting, reviewed November 5, 2026. Passed and failed mean a completed count or a call by The Associated Press (AP), the election office, or local press. A measure went developers’ way if it loosens rules and passed (or is likely to pass) or tightens rules and failed (or is likely to fail). Included are measures defined as likely to pass or fail with a current margin of at least 5 points and no local news outlet describing the race as undecided. Not shown: 3 measures too close to call and 3 mixed or procedural measures.

Infill Is Winning in Built-Up California

California voters were the most receptive to development. Votes loosening the rules passed or are leading in 8 of 10 California measures with a clear direction. Davis approved the 1,250-home Willowgrove project (73.2% yes), five months after Davis voters rejected a different project by 276 votes. San Bruno (64.2% yes) and Albany (63.3% yes) are on track to repeal 1970s rules that gave voters or neighbors a veto over zoning changes. A citizen initiative to cut Newport Beach’s zoned housing capacity from 8,174 to 2,900 units was defeated (32.6% yes). Affluent coastal towns were the exception: height increases in Coronado (21.8% yes) and Sausalito (39.5% yes) are trailing badly.

At the Edges, Voters Are Keeping the Veto

Outside California, measures loosening the rules passed or are leading in just 1 of 6 measures. Most of the losses were referendums on approvals a city council had already granted. Salida, Colo., voters overturned a 72-unit rezoning. Prescott Valley, Ariz., voters are on track to reject annexing a 3,500-home development, which would leave the site under county jurisdiction. Meanwhile, measures that add a public vote, a cost, or an open-space requirement to new growth passed or are on track to pass; across all states, 11 of 13 tightening measures passed or are likely to pass (Table 1). The data center fight spilled over in Grove City, Ohio, where voters approved a referendum requirement for any project over 50 acres, regardless of use but targeting data centers.

Housing Money Passes; Housing Rules Split

Few of these measures were pitched as ways to lower housing costs. Trepp classified each November 3 measure, statewide and local, by whether its ballot text, official analysis, or main campaign cited housing costs or income-restricted homes. Of 34 rule changes, 25 did not have an affordability aim (Table 2 omits the five too close to call).

When the question was money, voters said yes. Measures funding affordable housing or homebuyer aid passed or are likely to pass in 9 of 11 votes, including all four statewide bond and finance measures, which drew 59.3% to 62.2% yes.

Results were mixed and remained unsettled for rule changes pitched on affordability. Of six votes classified as affordability measures, one passed, two failed, and the remaining three were too close to call. The one that passed was Davis’s Measure W, which approves the 1,250-home Willowgrove neighborhood with at least 250 deed-restricted affordable units. The two statewide measures are undecided: Massachusetts Question 7, promoted to build smaller, lower-cost starter homes, is at 49.4% yes, and California’s Prop 45 is at 46.2% yes. Newport Beach shows the label can cut both ways: its defeated Measure H cited 2,900 affordable units in its ballot text while cutting the city’s total zoned capacity. Separately, three cities put measures on the ballot to meet state housing requirements. Votes in Monterey Park (85.2% yes) and Albany (63.3% yes) are leading, and Coronado’s (21.8% yes) is trailing.

Approval Risk Sits at the Local Vote

For developers and lenders, the lesson is that an entitlement is not always the last word. Salida voters this week and Lakewood, Colo., voters in an April special election undid, by referendum, rezonings previously approved by their city councils. Grove City voters have approved, and Castle Pines is on track to add, a public vote before large projects proceed. A 2022 industry survey put the cost of local opposition at 5.6% of multifamily development cost and 7.4 months of delay.

For existing collateral, the supply effect cuts both ways: where infill reform succeeds, new supply competes with existing buildings. Austin, Texas, added about 120,000 units from 2015 to 2024 and saw its median rent fall from $1,546 in December 2021 to $1,296 in January 2026. Where voters keep the veto, scarcity is preserved, but new projects carry entitlement risk long after a council says yes.

Data centers ran into a wall. Other development ran into geography: density is gaining ground in built-up cores, while the edge still wants the final say. Final counts in several states will settle the closest calls.

Data note: Trepp calculations, not published statistics. Election outcomes are a Trepp compilation of unofficial returns from county and city election offices, Ballotpedia, and local reporting, reviewed November 5, 2026; none is certified. The catalog covers local measures that change development rules (zoning, land use, growth and annexation limits, and voter-approval requirements) or approve specific projects, rezonings, and annexations; it is not a complete census, and California is overrepresented because its results are most accessible. Data center measures are excluded and are covered in Trepp’s data center ballot review; the counts cited here include the later returns reported in that post’s note. Affordability aim reflects each measure’s stated purpose, not its likely effect on prices.