Stanford Research Communication Program
  Home   Researchers Professionals  About
Archive by Major Area


Social Science

Natural Science

Archive by Year

Fall 1999 - Spring 2000

Fall 2000 - Summer 2001

Fall 2001 - Spring 2002

Fall 2002 - Summer 2003




The Benefit of Roads to Companies

Tom Petersen
Department of Infrastructure
Royal Institute of Technology (KTH), Sweden
June 2003

For what good are roads to companies? We know that roads make it easier to access different places faster, cheaper, and with greater versatility. Many roads and nice and useful places constitute an urban area. I try to measure the benefit that companies are experiencing from roads, railways and other transport infrastructures in the new Øresund region of Sweden.

My field of study is the interaction between the transport infrastructure and the location of companies and households. Companies are dependent on workers and raw materials for their operations, and on households as consumers for their products. Households, in turn, are dependent on companies and organisations for work to earn money, and for the provision of useful goods and services. Moreover, both are dependent of the transport system to transport people and goods. The specific area that I study is the new Øresund region, between the south of Sweden and Denmark. It was "created" by the construction of a major bridge that connected the capital of Denmark with the third largest city of Sweden. It comprises some 3.5 million people.

Research about the benefits of road investments to society has been ambiguous. In the 1990's, there was a debate in the U.S.A. about the decrease in economic productivity, with one possible cause being too little public investments in, for example, interstate highways. Many studies were published, some contradicting each other. Mostly, the data used in these studies was on a large scale, like the county or state level. In contrast, I am using data on individual companies (called microdata) and a finer geographic resolution.

Accessibility is a very widespread concept in the world of transport planning. It expresses how many "attractive" sites there are around and how easily they are accessed. For companies, this means easier and faster access to skilled labour, intermediate products, and information, as well as to the supply market, where the company sells its products. High accessibility is also important for another reason: it implies higher flexibility. Accessibility can be measured in many different ways, but it can always be given some kind of number. However, this number is relative; there is no "absolute accessibility"! Now, if the data on accessibility explains some of the variation in the productivity of companies, meaning how efficient they are in terms of production, then we can say that the transport system has some importance for the establishment, development and growth of companies. We might also be able to say how important, in comparison with other production factors.

Another way of looking at the benefits to companies of the transport system is to investigate the so-called dynamic effects. For example, if several companies in the same branch are located in the same area, they could benefit from each other. These benefits are not only from the location itself, with its composition of other companies, skilled labour, consumers and subcontractors, but also from the mere adjacency between companies. Perhaps you could think of it as a kind of spillover of different kinds of information. In this way we get a self-reinforcing process of more and more companies in one branch attracting each other to a spot that becomes more and more attractive (until some physical limit is reached). This so-called cluster effect results in more and more companies concentrating in this area.

One great difficulty about the data is to adjust for self-selection of the companies. For example, if we want to study the effect of a "treatment" on a subject, in this case the treatment of "high accessibility", we have to control for the selection process of the subjects, which might not be completely random. It could be that it is the already productive firms that are located where there is high accessibility (possibly for other reasons). But what we want to know is if the accessibility alone improves the productivity. Another difficulty is the presence of spatial correlation, or if data in one area is correlated with data in adjacent areas.

Using data on the companies, I have found that in the construction sector, productive and unproductive companies are indeed clustered within a radius of about 2 km. For distances above 2 km, there is no detectable clustering. But on the other hand, when compensating for self-selection, I find no significant difference in average productivity between companies in high-accessibility areas and companies with low accessibility. The conclusion is that there is need for further research!