The impact of the congestion charge on the retail business in London: an econometric analysis

On 17 February 2003, London introduced a pioneering congestion-charging scheme. The impact on traffic was sudden and dramatic. For example, car movements within the congestion-charging zone (21km2) were reduced by about 30%. However, the impact on other activities, such as retail business within the charged zone, is taking longer to become evident. Many unusual events happened in close succession in 2003 making it difficult to isolate the effect of any individual event on the retail sales. This paper describes a detailed study on the impact of the congestion charge (CC) on the John Lewis Oxford Street (JLOS) store, which is one of the biggest retail stores within the charged zone. The impact is estimated using descriptive analysis, forecasting methods and econometric models. The analysis is based on weekly sales data for the period January 2000 to January 2004 (three years before the CC and about one year afterwards) relating to JLOS and five other John Lewis stores in and around London. A seasonal autoregressive integrated moving average (SARIMA) model was fitted, showing that actual sales match predicted sales up to 19 January 2003 for the JLOS store but that actual sales fall consistently below predicted sales thereafter. Clearly, something changed for the JLOS store during the week after 19 January 2003 with no sign of recovery to the end of the time series. In order to isolate the effects of a range of factors on sales, econometric models were fitted. The drop in sales at the JLOS attributable to the CC was found to be 5.5% by the time-series model and 8.2% by the panel model.