Most of us do not live in abundant riches. Our nation, our states and our cities are confronted with problems and limited resources to throw at them. For every dollar thrown at a problem, there is some alternative use of that dollar that didn’t get funded. Economists call this opportunity cost. Among the many benefits of foresight for  public policy, an awareness of how trends may intersect in the future may prevent us from wasting scarce resources funding projects that will lose their value prematurely.

Recently, I’ve been doing some research on self-driving vehicles (SDVs).  Along the way, I’ve developed a strengthening sense that two trends that are headed for collision.

1. SDVs seem to be the best bet for achieving Personal Rapid Transit (PRT), primarily through car-share and for-hire models (Uber, Zipcar, Lyft, etc.)  Even SDVs that are individually owned may be monetized during down time by loaning them to a share service.  Except perhaps for those who reside directly on transit routes, PRT is a superior option relative to mass rapid transit (MRT).   PRT will deliver the person from origin to destination via the most optimal route, minimizing waiting and eliminating transfers.
2. Cities are spending billions on light rail installations.  For example, Houston recently expanded its rail system by adding a North line and a Southeast line at estimated costs of $143M and $125M per mile respectively.  An analysis in the Houston Chronicle estimated the cost of 8.9 miles of rail at $1.4B (http://www.chron.com/opinion/outlook/article/Gattis-MetroRail-the-good-the-bad-and-the-ugly-6237429.php).  Because buses operate on public roadways, bus-based MRT is flexible and can more easily adapt to need. But rail-based MRT is much more capital intensive, requiring expensive infrastructure to operate. Because of the capital involved, rail-based MRT projects are major bets with time horizons in excess of 30 years.

I also have two hunches or assumptions about the impact of car share models on the future:

1. They will provide an alternative to car ownership for the poor, even in transit-weak cities;
2. Because of 1), they will accelerate the adoption curve for SDVs by eliminating the back-end of the traditional vehicle life cycle.

While the earliest models are coming on-line now (Tesla, 2017 Mercedes E-Class), I think fleet models (like Google) will roll-out around 2020 in the first cities (San Francisco, Austin, etc).  I expect most luxury brands to roll-out SDV capabilities around 2020 and mainstream brands to follow around 2025.  If there were no changes in car ownership patterns, half-the vehicles in service will turn over in 11 years and most will be out of service in 15.  However, viable and cost-effective SDV car-share fleets could eliminate the need for the poor to own a vehicle and secondary car markets (used cars) could take a shock.  This would shorten the life-spans of non-SDV stock.  It is plausible that the public fleet could become majority SDV around 2030.

If SDVs realized the potential of PRT, then what will keep MRT ridership up, particularly on the fixed routes of light rail?  All the cities making light rail investments presumably are anticipating system lifespans beyond 15-20 years, but SDV delivered PRT could threaten those systems within that period. At great cost to society, light rail systems may lose economic viability prematurely. When cities are paying for expensive light rail projects long after PRT has taken their riders away, there will be fewer resources for public safety.  Urban planners need to keep in mind the potential impacts of SDVs when making grand pitches for rail.