Smart growth failures

Gerrit-Jan Knaap for the University of Maryland presented some of his work on using incentives to manage development at SERC on Friday. Gerrit-Jan is director of the National Center for Smart Growth Research and Education so I had expected him to be fairly pro, but his research findings make for fairly gloomy reading for smart growth proponents.

Amongst other things, smart growth Maryland style involves spatial concentration of government expenditure in locally identified priority areas. The idea being that developers would then prefer these areas even if they were free to develop anywhere in the state. Even from the off, you would worry about the ability to seriously exclude non-priority areas from accessing priority expenditure (e.g. roads or schools). If the amounts involved were small, you might expect this to weaken incentives further. This appears to have been the case, with the programme having essentially no effect on the amount of development occurring inside priority areas.

The second programme Gerrit-Jan talked about has even deeper conceptual problems. Rural priority areas involved the State providing money to protect certain parcels of rural land from development. But coverage was patchy, so you would expect this to increase incentives to develop on neighbouring parcels that were not being protected if the benefits of protected open space outweighed the benefits of higher density development around you (because you know that your beautiful view is not going to get built on). Again, this concern appears to have played out in practice, with the scheme making very little, if any, difference to development in rural priority areas.

Only the last scheme that Gerrit-Jan talked about (prioritising regeneration areas) appeared to have a measurable effect - but the actual impact on re-development rates was tiny.

More details on the papers and the talk are available here.

I am sure that these findings are contested by the pro-smart growth lobby, but all in all, quite a negative message.