E-Business Impacts for Urban Freight: Results from an Australian Study

E-Business is expected to dramatically change the way business is conducted internationally, nationally, within states and at the local area level. Moreover, these changes are very likely to happen well within the planning time frames required for provision of transport infrastructure and services. E-business is defined as including e-commerce, either between Businesses to Business (B2B) or Business to Customers (B2C), and the adoption of electronic technology within businesses. This paper presents some results from a study commissioned by the Australian National Transport Secretariat (NTS) to assist Australian business and government pro-actively address the transport issues arising from e-business. The resulting working papers will be used to establish a research framework for identifying policy and planning levers to maximize benefits to Australia from national and global e-business activity. The study sought to investigate three principal questions on e-business impacts: how will the transport task change; what will be affected; and how can the transport system respond? Current literature suggests that growth in e-business stems from the combined existence of market demand, suitable enabling technology, and skills and familiarity in management/users/ industry/government. The results of the study suggest that e-business will have implications for urban freight including higher levels of demand for goods and services, increased requirements for logistics distribution, changes in location preferences and improved transport network performance. Smith, Ferreira and Marquez 2 INTRODUCTION E-commerce continues to grow exponentially. Worldwide e-commerce revenues, worth less than $US20 billion in 1997 are estimated to reach $US5300 billion by 2005 [1]. Growth in business applications of electronic communications within companies which, together with e-commerce between companies (B2B) or between business and consumers (B2C), make up e-business, is even more significant. E-Business is dramatically changing the way business is conducted internationally, nationally, within states and at the local area level. This can impact transportation directly as transportation organizations adopt new practices and indirectly as businesses and consumers change their demand for freight and travel. The magnitude and speed of changes increase the urgency of planning infrastructure and services in line with changing needs. The Australian National Transport Secretariat, representing state and federal governments, commissioned research into potential e-business impacts to answer three principle questions: • how will the transport task change; • what will be affected; and • how can the transport system respond? The scope of the study, carried out between October 2000 and May 2001, was quite wide as it sought to answer these questions for urban and rural regions across Australia. The study coverage included freight transport, passenger transport and associated issues ranging from direct impacts on land uses to associated impacts on environmental, social and economic fronts [2]. While some of the findings are location specific, the challenges encountered in forecasting the impacts of e-business on transportation in view of rapidly changing developments should be of general interest. This paper discusses some issues arising from ranking and rating the likely impacts of e-business on transportation. It uses the particularly significant urban freight results of the study as an example as they may be of relevance to numbers of cities across the world. While Australia has a land area of 7.7 million sq. km, approximately the size of continental USA, with a population of only 19 million as of December 2000, 65% of that population lives in 6 major cities and the three east coast cities, Brisbane, Sydney and Melbourne have growing populations from 3 to 4.5 million. The majority of the population of the cities live in low density suburbs, in common with North American cities. There is a gradual increase in apartment developments in inner city areas and land costs and rents are rising with proximity to city centers, in common with European cities. Base line information for the NTS study was collected including a review of literature, appraisal of available data for on-going assessment; information about trends in e-business, and current and expected transport impacts. Interviews were conducted with stakeholders and experts in planning agencies, e-business companies and transport/logistics firms across Australia, as well as international experts. These fed into the ranking and rating process. The paper first summarizes issues relevant for urban freight arising from literature. It then describes our approach to ranking and rating and describes the issues involved in ranking impacts and in the associated need to derive quantitative estimates of impacts for ranking. The approaches are then illustrated in a table rating impacts of ebusiness on urban road transport and some estimated impacts on urban freight and urban shopping deliveries. Finally, we propose a set of steps to provide ranking and rating results of sufficient accuracy and sufficient detail to inform decisions at local, state and national levels. E-BUSINESS AND THE URBAN FREIGHT TASK Changes in the transport task due to e-business can have particularly significant impacts in urban areas because many urban areas are already suffering traffic congestion and environmental stress from urban air pollution and noise. Of special interest are (1) the expected growth in traffic due to growth of business to business transactions and (2) the potential decrease in traffic due to on-line shopping replacing urban shopping trips, and virtual goods and services Smith, Ferreira and Marquez 3 replacing physical goods and services. Suggested factors which might influence such changes are briefly reviewed below to support the later discussion of likely significant effects. Business to Business As Figure 1 shows, by far the largest growth in e-commerce is in business to business transactions and this also applies in urban areas. E-business technologies have allowed re-engineering of the entire logistics chain from a pull supply chain model where customers demand goods, which are built to order and quickly dispatched. This contrasts with the traditional push supply chain: forecast demand, produce the product for stock and warehouse close to the customer. Pull systems result in the movement of smaller quantities more frequently just-in-time (JIT) hence lower inventory holdings. However such systems are unforgiving and “on time intact” delivery is of higher importance The Internet increases “reach” and “speed” of communications. While “speed” helps just-in time response, “reach” allows sourcing from further afield either across the country or across the city. Access to these distant markets may be either by direct web contacts, or via e-portals, really just electronic versions of a central market. These can be vertical, within an industry group. Examples in the transport industry include car industry and railroad portals. However possibly more important in the urban context are horizontal portals used by groups of companies to buy “high volume low value” goods such as office supplies. Such group tenders may result in delivery to firms outside the sellers usual market catchments. Access to wider markets and a wider range of products together with increased capacity to design products to customer needs result in increased consumption. Thus there are more trucks because there are more packages, but also because of a change in delivery techniques. Same day and next day delivery can mean less than truckloads are carried. These increases will be additional to the general increase in freight with city population growth and general economic growth. A study in the Netherlands suggested a total increase in freight traffic of 38% by 2005 [4] Freight industry productivity improvements can have some compensating effects. For example, the carrier, often aided by e-business tools, may be able to efficiently make multiple deliveries, either of one type of product to many addresses or many types of product to a single address. However, this will usually include the use of larger vehicles. Hassall [5] gives an example of an urban rigid truck operator reducing the fleet from 800 to 750 trucks between 1995 and 2005 despite a 21% task increase by moving from 9 tonne to 22.5 tonne vehicles. Over the same period a Light Commercial Vehicle operator reduced the fleet from 1550 to 1400 despite a 6% increase in task, due to a change from 1 tonne to 2 tonne vans. Business to Consumers There has been of particular research concentration on affects of electronic home shopping for supermarket goods because of the potential large effect of high levels of trip substitution. Approximately one in five household trips is for shopping and supermarket shopping is more repetitive than other types of shopping [6]. Complete replacement of car-based shopping trips by Internet ordering and fully loaded van-based delivery can reduce distance traveled by up to 70 to 80%. However, complete replacement is unlikely. A UK study has estimated that electronic home shopping will reduce car based shopping travel by 10% by 2010, offset by a 0.5% increase in delivery traffic by 2010 [7]. However, the Dutch study [4] concluded that home shopping in that country will increase vehicle mileage, because in the urban Netherlands, van deliveries will often substitute for cycle or foot trips. The two studies indicate the sensitivity of the outcomes to the assumptions made and the need for more empirical evidence. Overall effect depends upon what is bought on-line and who buys it. This is likely to vary over time with gradual adoption of new shopping patterns for different socio-economic and cultural groups within a city. Effect also depends on the delivery pattern of the urban freight vehicles. It can be shown theoretically [8] that if delivery from a single depot on the city outskirts replaces shopping trips to local supermarkets, the total urban kilometers will increase. While in the future impacts of groceries might be significant, at the moment most deliveries are of discretionary goods such as gifts and homewares. Delivery trucks from firms like Federal Express and the United Smith, Ferreira and Marquez 4 Parcel Service (UPS) carry e-commerce purchases directly to households in larger numbers than ever before. The high growth in Internet commerce is leading to an increase in demand for delivery vehicles. Growth of 15.3 percent and 25 percent for small/medium and larger vans respectively, have been reported. UPS, Federal Express, and the United States Postal Service (USPS) have all increased the size of their delivery fleets considerably over the past few years — the USPS’s fleet alone now numbers nearly 200,000 ground vehicles; UPS’s fleet, nearly 150,000 ground vehicles [9]. Many of these purchases are additional to rather than substitutes for in store shopping. E-tailers who went out of business in the “dot com” collapse are being replaced by traditional retailers who will, in most cases, want to continue trading from their store fronts. Such businesses aim to sell additional products on line rather than replace store visits. Such increased consumption affects demand for freight across the city from factory (or port for imports) through to final delivery to customer Virtual Goods and Services E-materialisation is the term used to describe the substitution of online products for physical products. Ematerialisation is significantly impacting transfer of music, software, money and information. However transfer of information is likely to be most important for urban freight. Downloaded music is predominantly a complement to CD sales and software sales is not considered high volume. Electronic funds transfer provides convenience and saves on armoured van cash collection and deliveries. But the myriad of documents attached to B2B or B2C transactions from bills to legal documents and transported across cities generate significant numbers of courier and postal delivery trips. In the USA there are trends towards decreasing physical mail per capita as information is moved electronically. Reducing the volume of paper bills mailed to consumers will also reduce the environmental impacts associated with paper production, printing, and transportation. Bills account for 60 percent of the first-class mail delivered by the US Postal Service. The average household receives 10 to 12 recurring bills each month, for an annual total number of more than 15 billion bills [10]. Financial transactions performed via the Web require far fewer material resources and none of the energy involved in moving information stored on paper to and from the home or office. It is estimated that electronic billing saves approximately 50 to 75 cents per bill in envelopes and postage, and another $1 in handling costs [11]. According to Forrester Research, about 47 percent of US households now use some form of automated payment for at least one bill, and an additional 17 percent say they would do so if the option were available [10]. Forecasters estimate that by 2001, 535 million bills will be presented electronically, up 275 percent from 1997. RANKING AND RATING POTENTIAL IMPACTS Both the rapid speed of change and the changes in society due to the information revolution in general, poses special challenges for forecasting the future of e-business and assessing its impacts. The traditional processes of collecting information over time, describing a base case and projecting of future trends, or assessing change scenarios, encounters problems in all stages. These challenges apply to studies of impacts on the economy in general but are greater in studies of impacts in specific industries, such as freight and transport. More detailed information is required and more detailed outputs are expected. Almost always such studies are looking for sufficiently detailed insights on future conditions to allow plans to be put in place. This requires predictions to be linked closely to time frames and specifics of localities. Assessing Degree of Importance The initial task for the analyst is to define importance. This requires consideration of • Quantity versus Quality: Is the principal indicator the magnitude of the impact (quantity) or the significance of the impact (quality)? • Criteria: Is the impact manifested in additional vehicle kilometres traveled, person trips, cost per kilometre, energy used, greenhouse gases generated, tonne kilometre, extra road infrastructure required, funding required or all of these? Will additional criteria be required? Smith, Ferreira and Marquez 5 • Stakeholders: Importance will vary with viewpoint. For instance, extra air travel will have limited impact on urban road authorities. • Where and When? Even assessing what is happening here and now is difficult in a time of rapid change. Our work to date has noted both differences in impacts over time and variability of impact with place. • Thresholds: E-business changes leading to strains on capacity will be subject to threshold effects. Demand may be satisfied up to some saturation level and then an entirely new facility may be needed. A related issue is the approach to thresholds of certain facilities, such as warehouses or road links. A one percent increase in road traffic outside peak periods may not cause any problem, but such an occurrence during peak travel may result in gridlock. • Urgency: The order and speed in which resources are allocated correspond closely to the priority attached to a given area. Resources may need to be allocated first to a less environmentally deserving area because opportunities, frequently economic in nature, may be lost entirely otherwise. Clearly, people will attach different degrees of importance to issues according to who they are, where they are and when they are considering the matter. From this viewpoint every stakeholder at every time in every place needs a personal impacts rating scheme. Certainly we believe that tailored schemes are needed and will suggest processes for putting them in place. However, we still aim for some general rating of issues of importance here to allow discussion. Setting Criteria To address the problem of comparing ‘small apples with large oranges’ we need some common terminology for comparison across all modes and for all authorities. We have adopted the change in vehicle trip numbers as a common measure of changing activity. Percentage change in trip numbers, rather than absolute trip numbers is the relevant metric as it needs to be compared to the current situation to estimate likely impact on stakeholders. We rate on a scale the impact of changes on different stakeholders. We realise that many stakeholders are not interested in the number of vehicle trips per se. Rather they are concerned about persons carried or tonnage carried, transit or delivery times and a whole series of financial issues from cost per unit to bottom lines. However, it is difficult to encompass all such issues in a simple table. Instead, we base qualitative assessment of impacts on stakeholders as impacts on the issues of interest to such stakeholders due to increases (or decreases) in vehicle trips. For example, our rating of the impacts to greenhouse gas (GHG) targets from any increase in particular types of trips, are based on the expected change in the current GHG impacts of such trips. The ratings are based on assessment over the course of the project and especially informed by discussions with stakeholders and review of information. We have chosen to tabulate them in terms of the source of the trips (for example, shopping deliveries) rather than outcomes, such as urban trips. While the outcome will be of more interest than the source for most transport purposes, noting the source makes the rating assessment more transparent. In numbers of cases, it will be the cumulative effect of trips from different sources that make the total impact on the transport system. These overall effects are addressed later in making impact estimates. Where and When? The base year for this study is 2000 and changes are expected within five years from 2001 to 2005 and those expected in 2006-2010 are compared with the situation in 2000. In particular, increases in 2006-2010 are interpreted as increases over 2000. Additionally, within the first category, we needed to include some changes that have taken place / are taking place in 2001. The cities considered in our study are multi-centered cities with a major CBD (central business district), and minor CBDs surrounded by low and medium density suburbs. Good or Bad? Impacts may be important because they offer advantages, such as extra service; or disadvantages, such as increased pollution, or in some cases, both. We have accommodated this by using signs in the rating system; plus (+) for advantage and minus (-) for disadvantage. Such signs require some assumptions of what stakeholders will see as Smith, Ferreira and Marquez 6 disadvantage. For example, we assume extra traffic will be viewed negatively by road authorities and local governments. How Much and How Many? A major challenge of this study has been ‘putting some numbers on impacts’. An assessment of the transport impacts of e-business saying the impact will be ‘big’ would expect the immediate response ‘how big?’ Yet, as noted, ‘how big’ depends on ‘when’, ‘where’ and ‘who’ is asking. Both data collection and assessment processes are needed to evaluate impacts at detailed local level, regional level, state level and a national level. In the absence of such measures we have attempted to provide ‘ball park’ estimates of the extent of some important changes in transport demand and supply. Overall impact depends both on the degree of change and the size of the base population subject to change. For instance, a one percent increase in the number of shopping trips per person in a large city can have a significant impact on total kilometres traveled, whereas 100 percent increase in the number of trips to market by organic produce suppliers would not. Final estimates of additional activities included estimates of actual change now underway and estimates of potential change. They are estimates of extra activity per year for stated years. Cumulating the figures over a five or ten year period could be misleading in view of their indicative nature. INDICATIVE RATINGS & ESTIMATES FOR AUSTRALIAN CITIES The main areas where e-business is forecast to have a significant impact on the demand and the supply of urban freight are summarised as follows: • Higher levels of demand for goods and services due to wider choices and lowering of business transaction/administration would bring about increases in travel by light commercial vehicles (LCV) for local centre/home deliveries, occurring mainly at off-peak periods. Hence there will be little impact on urban congestion. Consequently, relatively higher levels of GHG and decreased air quality are expected given that LCV have low standards for emissions. This presents a rationale for incentives to convert LCVs to alternative fuels such as electric power or natural gas. • Increased pressure on logistics/distribution tasks from increased expectations or requirements from customers for greater flexibility and greater levels of reliability in the delivery of goods and services. • Increases in transport demand through increases in freight task due to wider choices of supplier/provider. Additional demand, mainly on road based transport, given its greater flexibility, level of service and ability for value adding services. Rail will be unable to compete in what will be increasingly more demanding market segments. • Location decisions by firms: Small centrally located freight terminals replacing large warehouses on city fringes. Consolidation trends into major centers; local distribution/collection centers using established ‘bricks and mortar’ (e.g. Seven-Elevens; petrol stations; newsagents; etc.). Possibly small warehouses for freight will emerge. • Dispersion trends (e.g. self-employment; teleworking; loss of jobs in white-collar services industry), leading to longer trips. Less reliance on the CBD for commuting thus conventional public transport (fixed route/radial) will suffer continued erosion in market share. Flexible public transport will be valued. Location decisions by individuals; there will be a need for more personalized types of transport. • Logistics operators increase productivity through better scheduling and routing software. This leads to cost reductions, less empty running and reduced emissions and GHG. Smith, Ferreira and Marquez 7 • Improved transport network performance through interfaces between Intelligent Transport Systems (ITS) components and data interfaces to B2B and B2C systems. This has the potential to reduce transit times and trip time variability through improved knowledge of transport demand, as well as improved real-time data on transport network performance. Safety will also be improved. Such co-ordination between ITS and e-business systems is likely to be implemented in practice in the medium term. Links between ITS systems and B2B alliances: more data/information on entire supply chain available to all, including ITS components, leading to reductions in delays, accidents, vehicle operating costs and emissions. The above impacts are illustrated in detail in Table 1. The importance of the impact on the transport task or the associated service offered of listed e-business changes to particular groups of stakeholders, are rated on a scale of ‘ + ’ for small but significant benefits to ‘ + + + + ’ for very significant benefits. In the column next to the source of ebusiness change, the change/s in the transport task due to that type of e-business is listed. We have taken the opportunity to include some stakeholders outside the transport area since governments want to take into consideration the interest of the business or the community in general in setting priorities. Urban Shopping Trips Estimates Urban Shopping Estimate Small overall increase in total travel, an extra 1.1 million kilometres by 2005. As reported in Smith et al. [13], only one of the transport experts interviewed during the study saw internet shopping opportunities as having an important impact on urban transport activity in the medium term. Moreover, more rather than less traffic is expected as use of e-shopping increases. Both these predictions will seem counter intuitive to the many people who think first of internet shopping when e-commerce is mentioned and expect reductions in trip making due to ordering online. Hence it is important to provide an estimate of e-shopping impacts on transport activity. Shopping Trip Numbers It is first necessary to estimate the total number of shopping trips available for conversion to e-shopping, together with an indication of the length of such trips. Melbourne’s Transport Research Centre [14] and New South Wales Department of Transport’s Transport Data Centre [15] both carry continuous area-wide surveys of household travel and identify shopping trips, as described in Smith et al. [16]. Estimating the amount of travel which might be foregone, if the shopping is done online, from either data set is complicated by two issues: • Multi-purpose travel shopping is frequently one stop on a journey, such as work to shops to home. The TDC approach is to record trip in terms of their destination purpose. This makes it difficult to extract twoway shopping trip distances. We have approximated the two-way distance as 1.8 the average one-way distance drawn from the travel survey data. • Commodities purchased are not recorded in travel surveys. Time spent shopping as a proxy has limitations. Instead, we approximated a proportion of shopping trips via a proportion of expenditure of goods from the Australian Bureau of Statistics Expenditure Survey [17]. Other sources of information such as the ACNielsen [18] shopping surveys could allow these estimates to be refined. Access and Inclination to use online ordering will govern the take up of these options. Our assumptions are based on current pricing patterns where goods purchased over the internet cost more. Thus e-shopping appeals to the time-poor but resource-rich members of the community. Similarly only a proportion of the population uses the Internet, and ubiquitous use will be some time off, even allowing for technology improvements offering computer free internet access. ACNielsen [18] found that 41 percent of 25 and 39 year-olds expressed interest in ordering groceries over the Internet, as compared to 10 percent of all main grocery buyers. Smith, Ferreira and Marquez 8