Thursday, August 7, 2014

LEVs in HOVs?

A LEV is a "low-emission vehicle," usually referring to a hybrid electric-gas vehicle. HOV stands for "high-occupancy vehicle," and refers to the lanes on highways set aside for vehicles with multiple riders. But what happens when LEVs with single occupant are allowed in the HOV lanes?

The fundamental problem here is that for the goal of reducing auto emissions, letting single-occupant LEVs use the HOV lanes imposes congestion costs on carpoolers. After all, the intention behind encouraging LEVs is to reduce carbon emissions. The primary reason for HOV lanes is to provide an incentive for carpooling and ride-sharing and thus to reduce traffic congestion. But as Antonio Bento, Daniel Kaffine,  Kevin Roth, and Matthew Zaragoza-Watkins point out in their paper, "The Effects of Regulation in the Presence of Multiple Unpriced Externalities:  Evidence from the Transportation Sector," allowing LEVs in the HOV lanes adds to congestion in those lanes, which has costs in terms of time and emissions. Their article appears in the most recent American Economic Journal: Economic Policy (2014, 6(3): 1–29). The AEJ: Economic Policy isn't freely available on-line, but many readers will have access through library subscriptions.

Bento, Kaffine,  Roth, and Zaragoza-Watkins write (footnotes and citations omitted): "Recently, in an attempt to reduce automobile-related emissions, policymakers have introduced policies to stimulate the demand for ultra-low-emission vehicles (ULEVs) such as gas-electric hybrids. A popular policy, in place in nine states and under consideration in six others, consists of allowing solo-hybrid drivers access to high-occupancy vehicle (HOV) lanes on major freeways. In this paper, we take advantage of the introduction of this policy in Los Angeles, California to study interactions between multiple unpriced externalities. ...  "Beginning August 10, 2005 and ending June 30, 2011, owners of hybrid vehicles achieving 45 miles per gallon (mpg) or better were able to apply for a special sticker that allowed them access to HOV lanes regardless of the number of occupants in the vehicle."

The authors had access to detailed data on the cars travelling on Los Angeles freeways. The basic analysis of the study is a "regression discontinuity," which basically means looking at whether their is a discontinuous change in traffic levels at the time the policy began. " So how do the benefits of lower emissions from more use of LEV cars and the costs of greater congestion in HOV lanes balance out?

Assume that every LEV observed n the HOV lane during the rush hour was purchased only because of this policy--which of course will overstate the benefits considerably--they find that the benefits in terms of reduce emissions are worth about $28,000 per year.

On the other side, the primary cost arises because the increase in travel time during the morning peak on the HOV lane is 9.0 percent and is statistically significant at the 1 percent level; this effect corresponds to an increase of travel time of 2.2 minutes." Of course, this is 2.2 minute per driver in the conga line of LA traffic in the HOV lanes during morning rush hour. Assuming a value of time of about $21/hour for others in the carpool lane accounts for a rise in congestion costs of about $3.3 million per year. This is partly offset by $1.7 million in reduced congestion times in the other traffic lanes, and by the gains in reduced driving time for the hybrid drivers now allowed in the the HOV lanes. But the costs still heavily outweigh the benefits.

Consider that when you allow LEVs in the HOV lane, the only time this provides a  positive incentive is if the driver is avoiding congestion in the other lanes. Unless the HOV lane is completely free-flowing, which is often not true in Los Angeles, adding more cars to that lane will add to congestion for those drivers. They write: "While adding a single hybrid to any HOV lane at 2 am creates zero social costs of congestion, adding one daily hybrid driver at 7 am to a very congested road in our study area (the I-10W) generates $4,500 in annual social costs. On these exceptionally congested roads, HOV lane traffic may be up to 30 percent above socially optimal levels, implying significant congestion costs from allowing hybrid access."

The authors then calculate the costs of reducing various air pollutants (greenhouse gases, nitrogen oxides, and hydrocarbons) by letting LEVs drive in the HOV lanes. "Our findings imply a best-case cost of $124 per ton of reductions in greenhouse gas emissions, $606,000 per ton of nitrogen oxides (NOx) reduction, and $505,000 per ton of hydrocarbon reduction in the most optimistic calculations. These costs exceed those of other options readily available to policymakers."

This analysis shows how economics clarifies the underlying reality of policy choices. Allowing LEVs in the HOV lanes does not require an explicit outlay of funds, and so often appears cost-free to local traffic authorities. Bento, Kaffine,  Roth, and Zaragoza-Watkins write: "Further, a policy that was perceived as “free” was far from free. We find that it costs carpoolers $3–$9 for every $1 of benefit transferred to hybrid drivers."

If anyone was to propose that carpoolers should pay a special tax, with the money to be sent to those who buy LEVs, everyone would question their sanity. But this policy of allowing LEVs in the HOV lanes has the actual economic effect of taxing those in the carpool lane--not in terms of money, but in terms of time--and transferring gains to those who buy LEVs. In effect, it's a policy that tries to pay for reductions in emissions by increasing the costs of traffic congestion. It's muddled thinking.

(Full disclosure: The AEJ: Economic Policy is published by the American Economic Association, which also publishes the Journal of Economic Perspectives where I have worked as Managing Editor since 1986.)