Association for Public Policy Analysis & Management

Policymakers and politicians often face the trade-off between meeting a policy objective and accounting for its distributional impact. While economists would suggest a price-based instrument to address an externality efficiently, politicians have to evaluate it based on its effect on the median voter. If the policy is regressive, then they would resort to other alternatives. One such case of disagreement is the License-Plate-based Driving Restriction. The policy is commonly implemented in developing countries to deal with transportation sector externalities, namely congestion and air pollution. Based on the last digit of the license plate number, a vehicle is restricted on particular days of the week. The focus of the current paper is to compare the distributional impact of this driving restriction policy to two price-based instruments - the vehicle mile tax (VMT) and a cordon charge, using a mode choice model.

The results of the paper lend support to the argument proposed by politicians for their choice of the driving restriction policy over a price-based instrument. If both policies lead to the same reduction in total car trips and in the absence of a revenue recycling mechanism or any indirect effect on driving behavior from time saving, the loss in consumer surplus from a price-based instrument is higher for all income segments (high-, middle-, and low-income) than the driving restriction policy. The license-plate-based driving restriction policy does impose a compliance cost on vehicle owners, particularly the middle class who own a single-vehicle and cannot afford multiple car ownership to circumvent the restriction. However, it is lower compared to the loss suffered under a VMT. The loss under a cordon charge is less than the scenario with VMT as the former effects only commuters in the downtown area of a city. This finding supports the proposal of policymakers in Beijing to implement a cordon charge in the downtown area to reduce congestion.

The policy has been implemented in Latin American cities like Santiago and Sao Paulo, Mexico City, and Beijing, China. Even though the distributional impact is the main argument proposed by a politician for their choice of the policy, a limited number of studies have explored its validity. Using data from the 2012-13 Santiago Metropolitan Travel Survey, a mode choice model was analyzed to empirically examine this argument in the case of Santiago, Chile. The conditional logit model and its properties were used to estimate willingness to pay for time savings and the consumer surplus under each policy scenario.