4 ms·
I think parking capacity probably reduces to a local scheduling problem [1]. If it does, it reduces to a 3-sat problem and this would mean optimum parking is NP
by brudgers 8y ago
I think parking capacity probably reduces to a local scheduling problem [1]. If it does, it reduces to a 3-sat problem and this would mean optimum parking is NP-complete. Most of the time a parking space more than 400m away is irrelevant [2] and when it comes to where a person parks relative to their home the acceptable radius is probably 100m or less...Des Moines is cold in the winter, hot in the summer, and gets storms at the scale of the Great Plains several times a year.
The marginal cost of a parking space is low. The heuristic for structured parking has long been $10,000 per space. Over a 30 year building life cycle that's less than $1/day. Private surface parking is a few cents and on street public parking is nothing...and often so is even that expensive structured parking because the $1/day of capital costs makes parking a potentially lucrative business. This means that the most cost effective way to avoid scheduling bottlenecks is with capacity that far exceeds the average rate or median demand.
To put it another way, all that parking is what a working system looks like. Sure Des Moines might not need 1.6 million parking spaces -- but it's probably not far from what's necessary to provide required for reasonable livability at the density and geography [3] of Des Moines. Recovering all those parking spaces for development doesn't do a lot for Des Moines. It's not severely space constrained by an ocean like NYC or SF and so the economic value of conversion is not lower.
To put it another way, land use for parking in Des Moines (or other cities) largely reflects an economic equilibrium in the real-estate market. When there's too much, it gets converted to other uses over time...and when regulations are impediments, major local real-estate interests get those regulations relaxed/modified/removed. [4]
[1]: The size of the 'local' may vary, but it's much much smaller than Des Moines and difficult even down to the scale of on-street parking for a single urban block.
[2]: Based on the heuristic of 1/4 mile reasonable walking radius in urban environments.
[3]: Consider that the Des Moines river runs through the city and a parking space on one side of the river is not generally fungible with a parking space on the other side. The river complicates public transit by bottlenecking surface transit with bridges and subsurface transit with tunnels.
[4]: Institutions with teams of $600/hr real estate lawyers and fifty year investment horizons.
- al2o3cr 8y ago> It's not severely space constrained by an ocean like NYC or SF and so the economic value of conversion is not lower. is somewhat contradicted by > Most of the time a parking space more than 400m away is irrelevant This is the space constraint the article focuses on: if there's enough surface parking, it pushes destinations far enough apart that you need to move your car to go from one to the other. Re: marginal cost, there are plenty of effects that impose additional capital and/or opportunity cost on converting between parking and not-parking uses - everything from long leases for companies that operate parking (10+ years sometimes) to basic constraints like "putting up a building requires a large chunk of up-front spending". Those effects make "economic equilibrium" slow-to-impossible to reach.