af40's Blog

A number of years ago I read a fascinating book by Columbia political scientist Robert Jervis entitled 'System Effects'. One of the book's major premises was that oftentimes policies initiated by governments aimed at ameliorating a specific problem end up having unintended consequences that might cancel whatever positive gains those policies may have been intended to have. One example that I have come across concerns bicycle helmet laws. One study out of Britain a few years ago found out that drivers who passed by bicyclists wearing helmets tended to drive a little less attentively, passing closer to those wearing helmets than those not wearing them. Perhaps subconsciously, automobile drivers felt they could get away with more reckless driving when bicyclists had more protection. But this may translate into a higher incidence of car-bicycle traffic accidents, undoing some or all of the benefit of bicycle helmet laws. 

Or, take Jon Corzine's proposed- arguably massive- toll hike on several major New Jersey roads. At face value, the measure assumes that if X amount of drivers spend Y amount on tolls each year, and, say, that toll is quadrupled, and assuming that X amount of drivers still drive on those roads each year, then the state of New Jersey should end up with 4*Y toll revenue. A welcome bounty for our fiscally irresponsibe state, right?

Well, not exactly. The problem with Corzine's plan is that it assumes that the context within which the toll hikes take effect- what Jervis referred to as the "system"- will remain unchanged even after a new policy is initiated. Or, in the toll example, that the "X" amount of drivers will remain the same. But will it?

Probably not. Since drivers are not mindless automatons, and have choice (a wonderful catch-all word that often confounds economists), they may rationalize that alternative modes of travel may offer as much benefit (for economists, utility), and thus this will reduce "X" by some finite amount.

Translation: fewer drivers are likely to drive on toll roads if the tolls reach a certain high level.

What is that level exactly? Recently, two academics named Peter Swan and Michael Belzer from Ohio's Wayne State University conducted a study of just that question, and found out- no surprise here- that drivers tend to opt for free, non-toll roads when they have an alternative.  In their estimation, a three-fold increase on the Ohio Turnpike would divert about 40% of traffic to "free" roads. In essence, this would mean that the actual revenue the government or toll operator would collect would not be a multiple of 3, but rather something only nearly half as much, about a factor of 1.8.  And this would be on a major highway that is arguably a vital conduit, not simply a minor road that a driver can easily avoid.

Why is this important for New Jerseyans? Well, for one, the major roads that Corzine has targeted- including the New Jersey Turnpike and Garden State Parkway- are not the only modes of transportation across the state. I frequently drive to north New Jersey via the Parkway, but I can instead opt for Route 206, a narrower road that takes me about half an hour longer but that costs me nothing in tolls.

The Wayne State study is very useful because it tracked changes in traffic following changes in tolls, exactly the type of information that might have been useful in Trenton, had our lawmakers been wise enough to consider the unintended consequences of their proposal.

But the consequences don't end with diverted traffic and lost state revenue. Let's consider a hypothetical scenario where traffic on NJ Route 206 increases by some 50% as a result of the Corzine toll plan. Since the road is almost at full capacity during peak hours, this is likely only to increase congestion. Some drivers, of course, might opt back on the Parkway if this happens, but some might find the toll prohibitive (especially once it increases eight-fold) and opt for the free road.  The higher traffic on 206 will mean more problems for local city governments. For example, there may be a higher incidence of traffic accidents and medical emergencies, a direct cost for local governments. Pollution is also likely to increase locally.  More lanes may be added to accomodate the added traffic- damaging the local environment.  All of these unintended consequences have direct financial costs associated with them, which serve to tax local governments. There might be some added local business revenue too, but how much is in question, in particular because commuters are not usually local business patrons (with the notable exception of stopping to buy gas at gas stations). And it is entirely conceivable that the higher cross-municipality commuter traffic will reduce local business patron traffic.

I have already stated in a previous blog on this issue that high tolls may cause businesses to lose revenue and even go out-of-state or into bankruptcy. This is itself an unintended reduction in the tax base.

The basic fact is that a major increase in tolls across New Jersey is likely to bring only a fraction of the estimated revenue to Trenton.  It may even, ultimately, be completely revenue-neutral, if enough drivers are diverted to free roads and costs and economic losses are high enough.

Given that so much is at stake here for so many, wouldn't it have been more responsible for Jon Corzine to look for ways to solve the real problems behind New Jersey's financial woes, in particular its runaway spending? Not addressing the underlying problem may simply facilitate future government fiscal imbalances, on top of the financial costs of such ill-conceived "bandages" as the toll hike proposal.