It’s also dangerous to assume local land use changes amount to altered regional forms

ATTENTION CONSERVATION NOTICE: Changes in land use do necessarily amount to a change in regional form. I don’t think there are any downsides to trying land use strategies, but there are downsides to overestimating the climate effects of the strategy.

So I ruffled some feathers yesterday with my post, with twitter smarties telling me what’s what and calling the argument ridiculous and summarizing my post: “local land use change too hard! federal/state carbon taxes easy!”

Here’s a more accurate summary: “federal/state carbon taxes broad-based and potentially immediate! Regional form long-term and harder than people think!”

And so by emphasizing the latter and giving up on or underemphasizing the former, we are screwing up by not holding policy-makers at all levels accountable for both! Federal and state officials have punted the political hot potato of climate policy to cities, and I think we’ve allowed it–in fact, we have encouraged it by overemphasizing what cities might be able to accomplish. Cities are engines for change, we urbanists love to say. And they are. But cities and regions exist in larger policy contexts we should not ignore.

That’s enough exclamation points for now.

That said, the first tweet has an assumption in it that embodies a problem for me: the connections between local land use and regional form feels intuitively straightforward. What settlement density as a climate policy relies on, to make big changes, is regional form–aggregate changes that alter the commuting environment so that people get out of their cars.

It should be the case that higher permitted densities built parcel-by-parcel, development-by-development should accumulate into better regional form over time. After all, development-by-development is how we got sprawl, right?

Sort of.

As the research on sprawl matured, it became evident that sprawl was a bunch of different things: yes, tract housing and bad subdivisions were one set, but there were other things, as well: changing employment geographies and regional fragmentation were others, and the connections between them and single-use or restricted density zoning are more difficult to suss than intuition allows. Sure, single-use zoning isn’t helping. But how much of all this do local land use changes undo now that a great deal has been done?

To me, the move from the parcel or development level to the regional level is empirically and theoretically a lot like the move from micro to macro in economics. We know all these marginal decisions accumulate to trends in demand, etc, but even though we know this, the connections have been slow in emerging and damned hard to predict, let alone create policy for because the number of variables and units of analysis start to proliferate past what we can reliably measure or model. Or, necessarily, perceive.

Benjamin Ross chided me on Twitter and said, Hey, I wrote a book on this (He did, and it’s here, and you should buy and read it) and I’ve not read it yet (I bought it!), so maybe Benjamin has the handle on it we need, and I’m just behind. Jeffrey Sellers and team has some nice empirical research on regional political fragmentation, but they don’t get to the parcel level. Marc Schlossberg at the University of Washington has some very cool measurement approaches, but it’s a grounded approach and doesn’t systematically lay out a theory for how the parts accumulate into a whole.

The question becomes: Is US urban development not so path dependent that we can overcome aggregate phenomenon like existing regional form (and thus commuting patterns, and thus fuel use, and thus climate change) with better local land use choices (parcel-by-parcel)?

It’s a good question, and from my perspective, we don’t really have much to lose by trying it. Absolutely, let’s all change zoning right now. Or get rid of it and rely on nuisance law. I’m in! Big fan of graduated zoning that unbundles land from structures, and of Henry George, right here am I.

But I do think we do have a lot to lose by emphasizing this strategy and not much else.


1) We won’t all try it. Regional governance, despite years of discussing it, is weak and usually advisory, with jurisdictions internal to the region usually being adversarial towards each other and towards the regional governance body. Does it do the fight against sprawl any good if four out of ten jurisdictions change their zoning and growth controls and the other four raise their middle fingers and do not, thus taking as much of the dumb subdivisons as possible? Perhaps it’s marginally better than all 10 of the jurisdictions having bad zoning practices, but are four enough to move the dial? I suppose they could be if the four included the biggest jurisdictions like the city of Los Angeles. Even then, the potential for leakage on the policy is itself a reason to worry that it won’t work.

This is not an issue with single-jurisdiction regions, and those might do great. Horray! How much fuel consumption do those account for?

2) How much urbanization remains to be done in the US? We are at 81 percent now. This is a serious question. If the US changes its immigration policy towards greater inclusion, the answer could be “a lot.” If Milennials start forming households soon, the answer could be “quite a bit” in raw numbers but not percentages and who cares about percentages if the raw numbers stack up because percentages don’t consume fuel and people do. If Millenials in the US go the way of Europe and Australia, the answer could be “eh.”

The first is something we might force if we figured out how. That’s not shaping up to be a short, easy process, either. But if we sort it, all to the good.

For places with rapid urbanization, hey, no contest. Pick smart land use and growth regimes. But for the US?

Two strikes me as an exogenous change that we really don’t control. Redistributing people from the urban fringe works exactly how if we don’t have policies that pinch them in some way for being out there right along with our infill developments if we can’t secure infill developments fast enough to change the relative prices? In shrinking regions, we might have the right combo of relative prices, so ha! Shrinking towards density is a great idea. But let’s not act like shrinking is any less about distributing spoils than other issues in politics, facing the same fragmented political landscape that land use changes currently do.

I come back, again and again, to the good empirical work done by Shlomo Angel’s team at NYU. They have a nice book out on their work examining human settlement across the globe, and they have a lot to say about densifying.

Now, models with the limited number of units that cities provide are always a problem. But Angel is a competent empiricist. They caveat what they need to.

In this and prior research, the one variable that always comes out as significant and negatively related to the size of the urban footprint is always petrol taxation. It’s not zoning regimes, it’s not land use law, it’s not planning. It’s petrol taxation. Now Angel et al still have lots of ideas for how to manage land use towards density. But that petrol taxation variable is a big deal.

I’m not saying that carbon and petrol taxation are easy fights. I’m saying they are fights worth having.

I’m also saying that by turning away from those fights in favor of land use, we have created a policy environment for squirmish after squirmish after squirmish, and on and on in the fragmented political environment of US regions where winning the squirmish one part of the region can be readily undermined elsewhere in the region.

And these squirmishes are harder than they should be because the people involved in them do not face policy environments that make their preferred lifestyles more expensive than they are right now.

Tomorrow I have to teach, but on Thursday, I will make the opposite argument for why, in the US at least, we should try to supply density before we try to change carbon taxes or petrol taxes. I’m not sure I buy it, it’s worth talking about.