Production econometrics and transport demand modelling in Southern and Northern Sweden

Detta är en avhandling från Stockholm : KTH Royal Institute of Technology

Sammanfattning: This thesis consists of three main parts. The first and most important part, in terms of effort and time spent, is devoted to the estimation of the importance of accessibility for production at the firm or plant level using three different econometric estimation approaches. The results could have implications for the calculation of "wider" economic benefits of transport infrastructure, stemming from agglomeration externalities (e.g., scale economies). There are both methodological and result-wise conclusions that can be drawn from this research: methodologically, first, using unbalanced firm-level data requires the use of proxy variables to account for (initial) firm-specific unobserved productivity effects, and non-random exit from the dataset. Second, there are unsolved theoretical problems when applying an essentially aggregate approach to productivity analysis on disaggregate data, viz., relating to the existence of aggregate production functions, and to the aggregation of productivity from a disaggregate level to a more aggregate level in a spatial framework.Result-wise, clear productivity differences are presented, when comparing firms in the same time period but in different locations with different accessibility. However, it is not possible in this dataset to detect increased productivity for representative firms stemming from the opening of the Öresund link. It is therefore discussed whether the reason for this result could be the inappropriateness of output measures in a competitive business environment, where a large portion of the benefits are gradually transferred to consumers and thus remain unmeasured. Other, more comprehensive structural approaches to econometrics, including the demand side of the economy, are also recommended.The second part of the thesis treats an unjustly neglected area of transport research: the validation of transport demand models. These transport models are for example used to calculate the new traffic patterns and changes in accessibility from a transport infrastructure investment like the Öresund fixed link, around which most of this thesis orbits. The third and last part, written with two co-authors, deals with the "vulnerability" of the road network, in terms of effects on the travel time delays of the users when a link is disrupted. The calculated indices of importance and exposure could also be seen as extreme forms of accessibility, especially when there is no alternative route besides the one that is cut-off.