In retrospect, CodeNEXT failed not because staff produced a bad land development code, but because Council did not instruct it to produce a better one.
To be sure, Council did the right thing last summer when it killed the last iteration of CodeNEXT. CodeNEXT 3.0 was a bad code. It remained a bad code even after the Planning Commission’s heroic effort to improve it. The text was bad; the map was worse. The environmental and transportation codes would have made infill development harder, not easier. Design professionals panned the draft. The development community quietly opposed it. And the pro-supply, urbanist faction gave it tepid support at best. The Austin Neighborhoods Council and the like opposed the draft, too, but they dislike the idea of a new code altogether and will likely oppose any deviation from the status quo no matter what.
So the last iteration of CodeNEXT had lots of opposition but no true constituency.
Staff had the opportunity to produce a code with a constituency, but it floundered. Its handling of “missing middle” housing is a case in point. Imagine Austin directs that the land development code be liberalized to allow a variety of missing middle housing types. It calls for the “introduction and expansion into the market of housing types such as row houses, courtyard apartments, bungalow courts, small-lot single-family, garage apartments, and live/work units.” But after providing a modest amount of such housing in draft version 1.0, staff progressively wrung it from the succeeding drafts. Version 3.0’s most radical innovation was to allow accessory dwelling units with duplexes; this was literally the smallest possible liberalization of the status quo. Version 3.0 otherwise limited missing middle housing mostly to the few places where it is already permitted. Staff thereby dampened the support of those who want more missing middle housing. It gained no allies in return.
The prevailing opinion seems to be that there was a breakdown in the public process, as if a little more citizen input could have kept staff from producing a draft that made nearly everyone unhappy. But there was no problem with the public process. Everyone who wanted it got his say. There were ample more opportunities to participate at every stage of the process.
Public input began with Imagine Austin itself. The comprehensive plan—a broadly pro-housing document—was the culmination of uncountable hours of work by staff and citizens. Its development was a model of community engagement. And it was approved unanimously by Council.
The CodeNEXT process started almost immediately after Imagine Austin was adopted. Six years of public input ensued. I served on the Citizens Advisory Group from its inception in 2013 through late 2014. We held monthly meetings at which citizens were invited to speak; these meetings continued into 2018. But one didn’t have to attend CAG meetings to be heard. There were open houses. There were “listening sessions.” There were community character sessions, where residents were invited to identify types of housing they wanted for their neighborhoods. In the middle of the process, staff issued policy prescription papers, triggering new rounds of public input. There was outreach targeted to affordable housing and community organizations. The draft codes and maps were published online in a format that allowed line-by-line comments. The city published “white papers” by citizens on its CodeNEXT page. And, of course, there were numerous briefings to Planning Commission, the Zoning and Platting Commission, and City Council itself.
The public process did not produce consensus, of course. It was unreasonable to expect consensus given that Austinites want different things. But the public process did identify the key areas of contention such as parking, missing middle housing, and compatibility regulations. These issues were endlessly analyzed, dissected and debated.
But although staff gathered public comment at every step, it never got any direction from Council. It was left to chart a path on its own, and it settled on a draft that made no one happy.
Council began to debate these core issues only after it received the draft from Planning Commission with Planning Commission’s scores of amendments. It was only at this point—supposedly, the final stage of the code adoption process—that Council began to wrestle with fundamental issues such as how much housing capacity the new code should provide. Council has adopted a strategic housing plan that calls for 135,000 units of new housing over the next 10 years. Should the new code and map permit exactly that number of units? Twice as many? Three times as many? How do we measure zoning capacity? It was clear from their initial discussion that Council members did not agree.
But this is perhaps the concept that frames every other issue. It is not merely a technocratic issue. It is a matter of philosophy. A code and map that provide the capacity for another 270,000 housing units (with the assumption that only a fraction of those units will be built) look completely different than a code and map that provide the capacity for exactly 135,000 units. The latter guarantees that fewer than 135,000 units will be realized, and that what is built will be expensive. A housing unit capacity analysis that is based on zoning constraints will yield a misleadingly high capacity. An honest analysis will consider constraints such as drainage and water quality regulations, which sometimes limit density as effectively as zoning. (We should also recognize that any planning that assumes there will be no demand for growth ten years from now is foolish, guaranteeing a housing shortage twenty years from now.)
It should not have been left to staff to decide this issue, or the other central issues. Ultimately, Council will settle these issues, either by deferring to staff’s recommendation or choosing another course. Council will settle these issues even by doing nothing—sticking with the status quo is a policy choice, too.
Council has directed the city manager to recommend a process for restarting the code. I have no idea what the city manager will propose. But it is crystal clear what the process ought to be. Council must provide direction on the key areas of contention before staff performs any more work on the code. Council should return to the conversations it began having shortly before it pulled the plug on CodeNEXT. It should debate zoning capacity, missing middle housing, parking, the density bonus program, and the like. It should then adopt a resolution providing general but unambiguous guidance on these issues. It will be up to staff to implement Council’s policy direction. That is staff's proper role.
We do not need one more minute of public input for Council to adopt such a resolution. It has all of the public input it needs. Indeed, the six-year public process was not a waste, even if it was frustrating. Thanks to the voluminous public input, the multiple drafts and maps, and Planning Commission’s recommendations, we can identify the key areas of contention.
I’ll be writing about what I think such a resolution should cover. Others can probably produce a better list than I can. But regardless of what is on the list, it is imperative that Council provide guidance now. It should not kick the mess back to staff for a “restart” on the false premise that one more round of public process will fix the problems with staff’s draft. Council, not staff, must own the land development code.