News & Events
Colorado Supreme Court Opinion May Spell the End of Site-Specific Zoning Ballot Measures
In its most significant land use decision in years, the Colorado Supreme Court yesterday ruled that an amendment to a “PUD agreement” in Telluride was administrative, not legislative in character, and accordingly, could not be the subject of a citizen-led initiative. Straightforward as that decision seems at first, its implications extend well beyond a picturesque box canyon—taken to its logical conclusion, Kavanaugh v. Telluride Locs. Coal. Petitioners’ Comm., 2026 CO 47, could put an end to project opponents’ ability to overturn site-specific land use approvals at the ballot box.
Kavanaugh involved a developer’s attempt to authorize an affordable housing development on a property originally set aside as open space in a planned unit development (“PUD”). Evidently unable to secure the required sign-off for an amendment from every other landowner in the PUD, the developer turned to Telluride’s initiative process, which would have put the PUD amendment to a town-wide vote (an unusual approach to be sure, but one presumably calculated to succeed in a town starved for affordable housing). The town clerk, though, rejected the ballot measure on the grounds that the proposed amendment to the PUD agreement, which set the uses and development standards within the PUD, did not concern a “legislative” matter eligible for an initiative. The developer challenged in court.
The Legal Framework and the Court’s Decision
By way of legal background, the fundamental question was whether the proposed PUD amendment was “legislative” or “administrative.” Across myriad opinions, Colorado courts have held that only “legislative” items are eligible for initiative and referendum. “Administrative” items are not. So, for example, in Vagneur v. City of Aspen, 295 P.3d 493 (Colo. 2013), the Colorado Supreme Court blocked Aspenites’ attempt to set the path of State Highway 82 into town by ballot initiative on the grounds that routing highways was an administrative function not a policy-setting one for voters.
Employing that framework here, the court of appeals initially agreed with the developer, explaining that the Colorado Supreme Court’s 1981 decision in Margolis v. District Court, 638 P.2d 297 (Colo. 1981) dictated the outcome. In Margolis, the court concluded that a site-specific rezoning is legislative in character and thus may be subject to the initiative process. The court of appeals thus reasoned that if a site-specific rezoning was legislative under Margolis, so too was the developer’s proposed PUD amendment since both were forms of zoning.
The supreme court reversed, distinguishing between an “original zoning decision” (legislative) and a proposal to “amend one specific PUD” (administrative). See 2026 CO 47, ¶31. The court reasoned that Telluride’s adoption of its PUD enabling ordinance was the legislative act, and that site-specific review of a PUD application to “enforce preexisting legislative goals” was administrative in nature—it also concluded that amendment of a PUD “agreement” was effectively an “amendment of existing contractual obligations.” 2026 CO 47, ¶33. Further, the court reasoned that the initiative process is “not equipped to address the complex assessments required by the Town’s [PUD] enabling ordinance, such as how to best mitigate geologic hazard areas . . .” and that the developer therefore could not “use the initiative power to evade the PUD amendment process.” 2026 CO 47, ¶36.
What Happens Next?
On its face, Kavanaugh decided the initiative process was not available to a developer who wanted to amend their PUD agreement by citizen initiative. The concept of a PUD agreement is itself a product of Telluride’s idiosyncratic PUD process which, depending on the development’s complexity, may require both a “PUD plan” and a “PUD agreement.” Most Colorado municipalities and counties, however, require a PUD plan without any provision for a supplemental agreement. Notwithstanding its origins in an unusual code provision, the reasoning behind the decision potentially holds far broader implications.
First, the decision is likely to apply to initiatives and referenda. Courts have treated the legislative/administrative test as applying to both. Thus, if a land use matter is not eligible for initiative, it stands to reason that it also cannot be challenged by referendum. For example, if Telluride’s Town Council had approved the developer’s PUD amendment and voters then sought to bring a referendum to overturn the approval, the supreme court presumably would have applied the same reasoning and reached the same conclusion: ineligible for the ballot.
Second, while the court apparently relied on the idea that Telluride’s PUD “agreements” functioned at least in part like a contract (and governmental contracts are certainly administrative matters), almost all of its other reasoning suggests that it would apply to amendments to PUD plans in their various flavors: official development plans, final development plans, etc.
Third, site-specific rezonings are now arguably out of bounds for initiative and referenda too. Kavanaugh confines Margolis to the idea that only “original” zoning is legislative—against Margolis’s express language to the contrary—then explains that a single local government’s site-specific review is a “complex and multi-layered administrative process” carrying out established policy, i.e., the PUD enabling ordinance. Barring unusual circumstances like setting original zoning after an annexation, that suggests that any site-specific land use decision that is reviewed “based on case-specific considerations that require specialized knowledge” is administrative and ineligible for initiative or referendum. If this were the case, citizens would no longer be able to challenge site-specific rezonings via referendum, as it’s not hard to extend the logic to “case specific considerations” set forth in the zoning code itself in the form of approval criteria.
Given that initiatives and referenda concerning development approvals are used almost exclusively in opposition, that would represent a significant shift in favor of development. Still, courts could also conclude that vanilla site-specific rezonings are not sufficiently “complex and multi-layered” or could review the question case-by-case. Site-specific rezonings are still quasi-judicial in character and hence, remain subject to challenge through judicial review under Rule 106(a)(4) of the Colorado Rules of Civil Procedure.
Overall, Kavanaugh’s reasoning could substantially curtail citizens’ ballot-box involvement in site-specific land use measures. Time will tell, and we will continue to monitor how courts apply this decision to the scenarios above.