Nearly 1 million people enter Central London between 7:00am and 10:00am. 85% of these trips are made by public transport and 12% by private transport (pre-congestion charge). Transport for London states that the goals in implementation of the congestion charge were four-fold. These objectives were: (i) to reduce congestion in Central London; (ii) to make improvements to bus services; (iii) to improve journey time consistency; (iv) to distribute goods and services more efficiently. It also served to decrease pollution, accidents, increase passengers per car, and increase both bicycle and pedestrian safety.
London’s program works as follows. Motorists driving in Central London between 7:00am and 6:30pm are required to pay £10. Exceptions are granted to certain groups including motorcycles, taxis, disabled vehicles, alternative fuel vehicles, emergency vehicles and buses. In addition, local residents receive a 90% discount. Enforcement is a major concern to policy-makers. Video cameras around the city record license plate numbers and match it with the list of paid vehicles. Fines levied range from £40-$120 if not paid.
Costs and Inefficiencies
The two most standard criticisms of congestion pricing schemes are well examined by Evans (1992). The first is that congestion pricing may redistribute welfare from road users to the rest of the community. Additionally the benefit from congestion pricing may not be Kaldor-Hicks efficient in that the net benefit may not be enough to compensate those harmed by the policy. Evans concedes that should some portion of the revenue raised by the charges be used to improve road infrastructure, that a net welfare gain could (and likely would) result. The second criticism is that the implementation of congestion pricing would set up “perverse incentives for governments.” Evans argues that because governments are monopoly suppliers of roads, congestion pricing allows the possibility of monopoly exploitation. This problem could be eliminated by transferring a lump-sum payment, a function of previous revenue raised by the charge, to all registered drivers in and around the congested area.
There are several areas in which the London congestion pricing scheme can be considered inefficient. First, the fee is levied irrespective of how many miles the vehicle drives in Central London. Because a driver pays a lump-sum upfront, the marginal cost of driving within the concerned area is equal to what it would be without a congestion charge once inside the congestion zone. An addition problem is the lack of flexibility in terms of both location and time. It would be more efficient to vary the charged amount based on the time and location of the vehicle. As the goal is to reduce inefficient driving, it would be desirable to target trouble areas and times more effectively.
The high overhead costs of the system were criticized heavily before IBM took over administration in late 2009. The government had originally chosen the London based Capita Group to administer the system despite Capita’s inferiority in comparison with international companies. A comparison between Capita Group and a similar system administered in Stockholm shows Capita to be more expensive and to have a higher failure rate than their Scandinavian counterparts. Some studies suggest that almost half of the revenue gained by the program were spent on Transportation for London costs and payment to Capita Group. The accuracy of the Capita system was regularly questioned initially but failure rates showed a clear decline.
A 2005 Guardian article noted that foreign registered vehicles incurred 65,534 penalty charges from 2003 to 2005. However, due to a combination of driver ignorance and cross-border enforcement problems, only 1,993 of the 65,534 culprits paid (roughly 3%). This means that Transport for London is owed £9.5 million for those two years alone. This problem identifies both equality and efficiency issues with the system. In order to be most effective, the system must be able to collect debts from foreign cars or the congestion charge will not be applicable to them. Additionally, it is undesirable to have one segment of the population that is not responsible for the charge as this places an undue burden on the remainder of the population.
A 2007 econometric analysis found evidence to suggest that the congestion charge harmed retail businesses in London. The estimation shows with clear statistical significance that the congestion charge was responsible for decreases in retail sales (defined as stores selling clothing and furniture) even after controlling for seasonality, tourism, sales, several consumer price indices, and underground closures (notably the central line). The developed model suggests the congestion charge caused a 5.5% decrease from expected retail sales. Frustratingly, the model does not include gas prices as an explanatory variable. This omission should clearly call the results into question. Exclusion of gas prices may result in an upward bias of the estimation of the influence of the congestion charge on the decision to drive. Additionally, broadening the dataset to include businesses outside of the paper’s narrow definition may mitigate the negative effects shown by the model. For example, London delivery businesses reported delivery times dropped by 50%. Only 9% of Central London Business owners reported a negative impact from the charge. Further, when the charge was only £5, a savings of 17 minutes per employee earning average London wages would break even before other benefits are considered.