Economics blogger Ryan Avent has a modest proposal for D.C.: In order to allow for more capacity in areas where there’s demand for it, impose a one-time “floor tax” on all stories that a developer would like to build over the existing height limit. He’s on the fence about whether to charge different rates in downtown, office-oriented districts vs. more residential areas, but the principle remains the same.

We’re on the same page about the need to allow taller buildings, but I’m going to go ahead and disagree with this idea that it’s enough to simply accept as much cash as a developer wants to pay in order to build higher, which is a very dismally scientific way of looking at how the city ought to be shaped. As much as I think the argument about a building being “out of scale” with its surroundings is overplayed, the city should make these kinds of planning decisions intentionally—if we decide that it’s okay for buildings to be higher in a given area, then there shouldn’t be a financial disincentive to take advantage of the new rules.

Avent’s proposal is akin to Ed Glaeser‘s idea that we should get rid of zoning restrictions and just allow developers to pay whatever we deem the public cost to be. The thing is, governments make rules in order to achieve ends that can’t be achieved financially. If you believe in a “human-scaled” city, then the wealthy shouldn’t be able to undermine it. If you don’t, then everybody should be able to build as they see fit. (I happen to believe in something in between).

Instead of cash payments, I’d rather see height exceptions granted for excellent architecture, as the Zoning Commission is likely to do for Burnham Place (though buildings will still not be able to breach the 130-foot height limit). That way we can use rules to achieve a more beautiful city and a more competitive one at the same time. Under Avent’s plan, I can imagine developers spending money they might have otherwise put into great architecture on just adding a few more stories—which would just be a lost opportunity.