Few real-world situations fit perfectly the conditions of the standard economic approach to environmental
externalities learned in one’s first graduate environmental economics course and
presented, for example, in Baumol and Oates (1988). However, motor vehicles are quite likely
an especially poor fit. Their distinctive features offer opportunities and—more commonly—
create difficulties for fashioning approaches to environmental policy, especially policy respecting
conventional pollutants.
The users are primarily households, rather than firms. This means there are millions of
potential polluters, rather than thousands as in the case of stationary sources. This obvious fact
has had several ramifications that have profoundly shaped policy. First, emissions and emission
reductions cannot be directly measured for every vehicle, nor, for that matter, can vehicle use.
Vehicle use and vehicle emissions can only be determined indirectly by inferences from surveys
of households or (for emissions) of vehicles.1 Planners and regulators have had to rely heavily on
modeled rather than measured emissions. In the United States, for example, the U.S. Environmental
Protection Agency’s (EPA) MOBILE model is now used not only for planning but for enforcement
as well, a development that has had unfortunate consequences both for the model itself
and for policy. Second, these millions of motorist-polluters vote, which has made it very
difficult to implement effective emissions regulations directly on vehicles. Instead, most of the
work in reducing emissions has been done by emissions standards for new vehicles.
Responsibility for emissions is shared among a number of parties. The actions of many different
parties can affect motor vehicle emissions. Manufacturers design vehicles of different sizes
and types, and install the engine and emissions-control system. Refiners can produce fuel with
highly variable emission potential, particularly with regard to hydrocarbon (HC) emissions for
gasoline engines and sulfur for diesel fuel. Actual emissions, of both conventional pollutants and
greenhouse gas pollutants, vary greatly with how the vehicle is driven. Responsibility for vehicle
maintenance and emissions-system repair is shared between mechanics and motorists.
The upside of this diffusion of responsibility is that there are many potential regulatees, and
many possible interventions in the system. On the other hand, the more policies, the greater the
difficulty of choosing among them. Various actors can more easily point to others as the real
problem. And actions taken by one actor can be undone by the behavior of others. For example,
motorists’ failure to maintain their vehicles can compromise the integrity of manufacturerinstalled
emissions-control systems. The observed pattern of responsibility is in part a product
of the institutional structure that governs motor vehicle production, ownership, and use. Thus,
one possible approach to motor vehicle emissions policy is to consider ways to alter this structure.
For example, manufacturers could be required to maintain pollution control over the lifetime
of their vehicles; or in-use vehicle emissions control could be bid on and contracted by one
party, who would then be responsible for a target level of emissions reductions. These approaches
have been suggested (Harrington and McConnell 2000 have reviewed alternative property rights
structures) but have not been tried.