Depends on Your Definition of Subsidy

Ok, mere hours into my first day and I already have requests. Commenter dWj has a good question:

If we (Obama, Bloomberg, right-thinking Americans) ever get true carbon pricing and congestion pricing, does that mean we should quit subsidizing public transportation? Note that public transit creates negative externalities; we subsidize it because it's a consumption-substitute for activities with far more negative externalities. If we correct that, the subsidy becomes perverse, right?

His point, I believe is this -- transit is cleaner than driving but not cleaner than walking or biking. If we continue to subsidize transit after correcting for the relative cleanliness of transit over roads, then aren't we penalizing the greenest travelers of all?

Maybe. There are a couple of things to consider. One is that there might be other positive externalities to transit worth subsidizing. Transit might facilitate density, which can increase metropolitan productivity via agglomeration effects. Another point is that what might widely be perceived as "subsidy" could also just be a case of direct beneficiaries contributing to the funding of the system. Businesses along transit lines are often asked to contribute to transit funding (and they sometimes volunteer to do so) based on the increase in value the new transportation capacity creates. If you define subsidy as anything other than farebox revenues, then this would count as subsidy. But I don't think there's anything perverse about getting businesses to invest in infrastructure from which they directly benefit.

But there's another issue that's well illustrated in this post by Chris Bradford. Say we internalize all the externalities and price all the roads and rail lines properly, only to find that the system as a whole is bringing in a lot of revenue -- enough to pay for new capacity, in fact. And this will happen in growing cities; eventually transportation demand will outstrip available infrastructure, such that the city as a whole would be better off by building new roads or rails. Say that planners and engineers determine that the best way to add capacity is via a new heavy rail subway line. They use the excess revenue to build it, and then price the new line at the congestion reducing level, which will initially be zero. In other words, there's a subsidy involved; drivers are helping to pay for takers of transit. Even after riding habits adjust, the congestion reducing fare on the new line might be less than the operating costs. That's not perverse. When a system is considered as a comprehensive network, this will sometimes be the efficient result.

In any case, this is just a mental exercise. I wish we lived in a world where roads and parking and accident risk and carbon emissions were all appropriately priced, but I doubt we'll see that anytime soon. Until then, I'm happy to advocate for transit subsidies, however understood.