Economics of Bike/Walk infrastructure followup: Portland's example

From a previous post, the $60 million figure that Portland, OR spent on cycling infrastructure over the years was mentioned, and that the cost was around the cost of a single mile of 4-lane urban freeway.

Question: strictly from an economic point-of-view, was the money well spent, especially when we consider the fact that even in Portland, cycling mode share is around 5% of all trips. The answer is a resounding yes. Here's why.

From the 2010 census, Portland has a population of a little less than 600,000. If we assume that the typical family consists of four people and two cars, then there are around 300,000 cars. Five percent of that is 15,000 cars. Now, let's assume that roughly the time over which Portland spent the $60 million was the lifetime of these cars as a rough approximation. Each car costs at least $20,000 with the costs of fuel, insurance, repairs, maintenance, and so forth over the years. So that comes to $300 million - six times what Portland spent.

This simplified analysis did not include the benefits of a healthier lifestyle or increased social interaction. Nor did it include the costs of urban sprawl or pollution. The $20,000 per car figure is intentionally underestimated.

Now, infrastructure needs to be intelligently constructed. Efficiencies in what currently exists already need to be used - otherwise it won't be cheap. Designs like road diets on existing roads and the connection by low-speed secondary roads of separate neighborhoods are low-cost solutions that should be implemented wherever practical. New construction, which tends to be high cost, is difficult to secure in today's budget climate. Above all, infrastructure needs to actually go to places people want to go, and connect to/from neighborhoods and existing infrastructure to take advantage of the network effect.