Analysis of behavioral changes due to the Stockholm Congestion Charge Trial

Abstract: The Stockholm Congestion Tax Trial is one of the few projects where an automatic system was implemented to tax road vehicles in order to reduce congestion. The taxation period lasted from January until July 2006 and vehicles going in or out of the inner city during daytime were taxed with 10 to 20 SEK (1.07 to 2.14 Euro). The project included improved public transportation and more park & ride sites. The trial was limited to a half year and the public was able to vote on the congestion tax afterwards. A large research package was set up in order to inform everybody about the effects of the tax. The referendum was positive and the congestion tax is permanent today, but the data collected for research purposes is not fully analyzed yet. This thesis focuses on travel behavior of the population in Stockholm County and uses a household panel survey with 24,000 individuals. The waves took place in September 2004 and March 2006.The main idea is to use the advantages of the panel survey and keep track of how the individuals behaved during the two waves. The analysis is done separately for each trip purpose and examines the change of mode choice during the second wave in comparison to the individual’s choice before the congestion tax. This can be even specified for different socio-demographic groups.The cross-sectional comparison shows that avoiding trips to and from the inner city is not an option. The number of trips of the individuals passing through the congestion tax cordon is reduced to a lower extend by seasonal influences than the trips in all other parts of the county. Therefore the inner city has become more attractive as a origin or destination or the trips passing through the cordon can not be substituted by trips free of tax.Public transportation is the mode which benefits most from the congestion tax in terms of modal split. The importance of it increases in the whole county, but the gain is a lot higher for trips through the cordon. The modes foot and bike are not a relevant alternative for trips through the cordon because of long distances in general and cold weather during the survey. The panel analysis shows that many more people doing cordon passages to work, home and shopping change from car to public transportation in comparison to the rest of the county. Changes from public transport to car occur less often for cordon passages than for trips not being taxed.Certain socio-demographic groups adapt differently to the congestion tax. Compared to men, up to 30 percent point fewer women continue to use the car depending on the trip purpose. This may be related to personal income, but the variables household income has proofed to have a smaller impact. Only really poor households use the car a lot less often and the highest income group tends to continue to drive by car.The main limiting factor of the analysis is the seasonal and weather influence because of unusual cold weather during the second wave. It is not possible to take this completely into account because the control group of the cordon passages has different trip characteristics.The thesis provides better knowledge how a congestion tax affects the behavior of the people. It allows checking the reasonability of transportation modeling results and shows that certain socio-demographic variables should be used if modeling is done for areas with a different composition of the population than Stockholm.