Comparing private urban land use and public policy

In comparing the returns from different land uses, the key element appears to be balance. Developers need at least a 30% rate of return in order to acquire financing and build. The public sector requires a 20-30% rate of return in order to overcome the deadweight loss incurred by taxing buildings (24%) and the transaction costs of taxation and public expenditure (which vary). Deadweight loss means loss of economic output from distortions caused by taxation.

1. Historic Rehabilitation. Although the returns are much lower than surface parking, a 40% private return is enough to build. A -20% return for local government makes the historic tax break a win compared with demolition for surface parking or abandonment, though still no path to financial success. Of these four, this is by far the most balanced.

2. Surface Parking. With a private rate of return a whopping 240%, surface parking is a clear winner for investors, while a -50% public rate of return means that much of that profit is a transfer from building owners and renters, and those who work and shop in the area. Despite the public costs, this is by far the most profitable choice given to urban land owners in the study area.

3. Subsidized Development. This offers a similar return to private developers to Historic Rehabilitation at 40%, but at a staggering public cost of -3000% annual return of investment. It is unsurprising that local government is reluctant to pursue this route.

4. For Profit Development. The private return here is the lowest that we examined, offering only 20%. This is below the amount required for financing. The potential return to local government exceeds 1000%, but this revenue is a pipe dream. 

This analysis suggests that a fifth choice is necessary that better balances outcomes between private investment in construction and maintenance on one hand and public investment in services on the other. I call this technique a Build District, and I will lay out this option in the next installment.