Delivering supermarket shopping: more or less traffic?

Abstract

Car travel for food and other household items represents about 40% of all UK shopping trips by car, and about 5% of all car use. In the past 10 years, there has been an extremely rapid growth, albeit from a very small base, of home delivery services for such shopping. As this sector has developed, there has been increasing concern about the growth in delivery traffic, countered by the hope that services are reducing personal car travel. The balance between such tendencies is critical to the overall impact of services. The present paper examines a range of international evidence, including the results of nine modelling assessments. The evidence suggests that with realistic levels of take‐up, a direct substitution of car trips by van trips could reduce vehicle‐km by 70% or more. More complex shopper behavioural responses will occur, but, according to available empirical evidence, overall traffic reductions are still probable. Meanwhile, the benefits of services could be maximized by use of appropriate cost structures, new types of delivery location, less polluting vehicles, greater cooperation or out‐sourcing by retailers, and measures to encourage greater consumption of local produce.

Notes

Correspondence Address: Sally Cairns, ESRC Transport Studies Unit, University College London, Gower Street, London WC1E 6BT, UK. Email: [email protected]

To the author's knowledge, information is correct as of January 2004, although the home delivery sector develops so rapidly that there might soon be anachronisms.

A direct consideration of environmental costs and benefits is not included since transport and environmental impacts are not always straightforwardly correlated and a full exploration of the relationship would require further research.

However, all distances have been converted into kilometres to facilitate international comparison.

Results are based on an annual survey undertaken by BMRB International of approximately 1000 adults in the UK aged more than 15 years of age.

As trade has grown, Amazon has invested in both national Internet sites and national warehouses.

The model uses assumptions relating to the transfer of proportions of total grocery sales in the UK, but only examines the effects on travel for the most affluent 40% of households. This is based on the assumption that all home shoppers will come from the most affluent 40% of households.

At the time Palmer's work was undertaken, Asda and Sainsburys were both concentrating on the ‘fulfilment centre’ model.

Later simulations looking at the costs of reception box and delivery box solutions also examined the impact of varying levels of customer demand.

For various reasons, it was not possible to obtain the originals of these studies (in English).

As noted above, transport and environmental impacts are not necessarily directly related. Hence, only the principal conclusions of the studies are given as an interesting parallel with some of the direct work on transport impacts.

This is surprising, since deliveries to the workplace or a pick‐up point should allow for greater consolidation of orders. The most likely reason is that the warehouses were assumed to serve considerably wider areas than the stores, so that groceries were being divided into individual loads further from their final destination, creating inefficiencies. A similar effect was observed in Palmer's work.

In a study of tele‐workers, Gould and Golob (1997) showed that on days when people worked from home, on average they spent about 1 h less on work and associated travel while spending 46 min on out‐of‐home discretionary activities (including travel for those activities). In contrast, those working away from home spent about 1 h more on work and associated travel and 37 min on out‐of‐home discretionary activities (including travel). This implies that when working from home, only about 15% of ‘saved time’ was converted into out‐of‐home activities. This fact highlights that people who gain time will not automatically wish to convert all of it into extra out‐of‐home activity. However, the particular focus of the study of Gould and Golob makes it difficult to draw more specific conclusions about the overall traffic impacts of home shopping.

Information was obtained from J. Millar, Director of The Food Ferry (personal communication, 1998).

This trip frequency is considerably greater than that currently reported in the UK, although this might be due to cultural differences and the date of the research.

In the Lex Transfleet (2001) report, the same question was asked—but only to home shopping motorists (a total of 317 people). Opposition was significantly lower (at 24%), whilst support was higher (37%). However, given who was asked, the answers are less representative of general public opinion.

This applies to the delivery services from stores offered to customers living within a 30‐mile (about 50 km) radius. John Lewis also operates an Internet store, ‘John Lewis Direct’, which operates separately, and it has an alternative system for organizing deliveries.

Currently, the range of grocery retailers in the Local Collect service is limited. Sainsburys and Tesco were both involved in the trial of drop‐off boxes. It is unclear whether any groceries were delivered as part of the Nottingham trial.

Additional information

Notes on contributors

SALLY CAIRNS

Correspondence Address: Sally Cairns, ESRC Transport Studies Unit, University College London, Gower Street, London WC1E 6BT, UK. Email: [email protected]