The importance for the economy of a good transportation infrastructure has long been acknowledged. However, there is another interesting link between transport and economic growth. Transport has become intimately connected to economic activity. It is not only anymore about the transport of commodities and finished goods, people also travel to get to work, clients and suppliers need to be visited. Since 2005, the US Bureau of Transportation Statistics publishes a monthly transportation services index (TSI), now stretching back to 1979. It is based on a survey of both physical quantities transported and turnover of transportation services firms[Lahiri et all. (2003)]. It concerns the development of commercial transport, and does not relate to traffic flow measurements. Data on transport by road, rail, air, water and pipelines are collected. These are then combined into two separate indices, once concerning freight transport and another concerning passenger transport. These are then aggregated into the overall TSI. The relevancy here is that this transport services index has been shown to have a clear and strong relation with economic activity in the US. Lahiri et all. (2003) find that the TSI reliably identifies peaks and troughs in the US Business Cycle, with a lead of about 6 months. They also conclude that the freight index has a stronger link to the business cycle than the passenger transport index, and is therefore the more reliable indicator. Lahiry and Yao [2012] consider the usefulness of the transport services index for the US index of coincident economic indicators. They find a good coherence between the historic business cycle chronology and the TSI evolution. In their analysis, the TSI actually performs better than several of the current indicators included in the coincident index. Including the TSI would enhance the performance of the index of coincident indicators, not in the least as it suffers less from revisions than some other economic indicators. These results indicate a potential important role of traffic data as an economic indicator. In the Netherlands, an extensive network for the monitoring of road traffic is in place. Different government entities have placed sensors to monitor traffic flow relevant to them, such as the state highway agency, provinces, and municipal governments. An important initiative was started to collect all these data in a central database, the NDW, to enhance their value by completeness and to make the data accessible to a wider audience. Thus it contains traffic flow data from highway sensors, but also from provincial roads, commercial parks and inside cities. Also, efforts keep being made to include more sensors in the database, thus extending its coverage. The database used to contain mainly data from the densely populated western part of the country, but now covers sensors in most of the country. All this means that the data in the NDW database are able to give a fine grained picture of road transport activity in the Netherlands. This is what makes it valuable for this study. A countrywide road transport index would make a fine business cycle indicator for the whole economy, potentially even a leading one. However, using the data from individual sensors in a certain area, it is possible to construct local indicators of transport activity. The thesis here is that these can then be used as indicators of local economic activity. This would be highly valuable, as relatively little information is available on regional economic developments, certainly in the short term. Different parts of the country have different economic structures, and can therefore be expected to react differently to general business cycle developments. A fast and reliable regional indicator of economic activity would be of great use for policymakers and local firms alike. Traffic flow data have the potential to fulfil that role. They become available almost instantaneously, if that would be desired, and they are “hard” data, i.e. count data of real events. The link to economic activity is intuitive. And also very important, there are no revisions to traffic flow data, it gives an immediate final estimate. This study will take the Eindhoven region in the south of the Netherlands as a test case. There are several reasons for this. The amount of data is just too large to do a proof of concept on the whole of the country. The Eindhoven region is not only geographically well defined (no overflow into another urban area), but also has an economic structure which is highly focused on manufacturing. This means that outcomes form the manufacturing industry survey, which asks firms about turnover development and expectations of other aspects of business, can be used as a benchmark for testing the plausibility of the regional traffic flow data as a regional economic indicator. This would be less clear cut for a region with a more mixed economic structure, for example with a large central government presence. If the Eindhoven case shows that there is a clear link between local traffic intensity and economic activity, it becomes plausible that traffic intensity indicators for other regions should be credible and useful indicators of local economic activity. The data used in this study were obtained from the National data warehouse for traffic information (NDW, http://www.ndw.nu). The aim is to collect data from all government road traffic monitoring operations in the Netherlands. It thus contains data from sensors on highways (Aroads), provincial roads (N-roads) and urban road networks (B-roads). This extensive coverage means that all types of traffic are captured, and different selections for different uses can be made. Unfortunately for the purpose of economic analysis, the database goes back only to 2010 or later for most sensors. (Author/publisher)
Abstract