Skip to main content

2015-10-09 | project

The Social Cost of Beijing's driving restrictions program: A stated preference approach

In this project we set out to find data that policymakers need to decide whether to retain and replicate and refine their driving restriction programs: reliable estimates of the costs these programs impose on households and the incidence of these costs across socioeconomic strata.

Over the past four decades, vehicle fleets in developing countries have grown at 6 percent per
year, double the rate for developed countries. As cars, trucks, and buses
have proliferated, so too have attendant negative externalities. Today, vehicles are a leading
source of local air pollution in developing countries, contributing more than 90 percent of
emissions in some cases. In addition, they generate 13 percent of
global emissions of greenhouse gases. And annual deaths from
accidents in developing countries average almost 1 per 100 vehicles, a rate three times higher
than that for industrialized countries.

We will take a number of steps to design a CV survey that is policy relevant, plausible, and
meaningful to the respondent. First, in designing the survey, we will consult with researchers and
policy makers specializing in transportation in Beijing. Second, we will conduct multiple focus
groups with representative drivers subject to the driving restrictions program. Finally, we will conduct at least one pilot survey before the full-scale survey is administered. A pilot is needed to
test the CV survey and develop a preliminary vector of WTP bids.

In CV surveys, respondents are asked about their willingness to pay (WTP) a certain amount of
money for a change in the level of a nonmarket good. As noted
above, in our case, this good will be an exemption from Beijing’s driving restrictions program.
CV payment questions can be formulated in different ways, including as open-ended (how much
would you pay for a change in the level of good Y?) and closed-ended (would you pay amount
$X for a change in the level of good Y?). Closed-ended questions are preferred because of their
incentive compatibility properties (Carson and Groves 2007). We will follow this standard and,
as a result, will not directly observe respondents’ WTP for an exemption, only whether or not
they are willing to pay a certain amount.