Stephen McCarthy (p), Fatemeh Naqavi, Daniel Jonsson, Muriel Beser Hugosson, Anders Karlström

Sandy Dall'erba

The City of Stockholm anticipates substantial new office space will be needed over the next 20 years to accommodate population growth, and is currently developing a plan to guide new office development. This paper describes a scenario planning exercise conducted in collaboration with the City of Stockholm using a work location and dynamic scheduling model to simulate and analyse scenarios for the spatial distribution of new office developments from a transportation perspective.

Scenarios are developed using a four-quadrant matrix with one axis corresponding to concentration of new development (concentrated vs. dispersed) and the other to distance of new development to the city centre (central vs. peripheral). The dispersed-peripheral quadrant is determined to be politically and economically infeasible. For the remaining three quadrants, realistic numerical scenarios for the zonal distribution of new jobs are created with input from City planning officials.

We have developed a dynamic discrete choice activity-based model, SCAPER, in previous work, which we apply here to each scenario using a simulated 2040 population. The model includes a work location choice component which uses the scheduling model to create an accessibility measure feeding into the work location model, so agents working in newly developed areas have a geographic distribution of home locations that respects both travel time to work and opportunities for doing free-time activities. The travel behaviour model is iterated with a traffic simulator to produce a stable travel time/demand result for each scenario.

The model produces disaggregate results that can be analysed on several geographical levels. The scenarios produce marginally divergent effects on transportation (e.g., mode share, travel times) at the city-wide level; differences for workers in the newly developed areas are more substantial. We analyse our results from the perspective of social welfare, environmental effects and spatial segregation, which are policy impacts important to the City that we have developed analysis tools for. We also exploit the advantages of SCAPER as a scheduling model that endogenously considers trip-chaining and activity durations, for example examining differences between scenarios in trip-chaining behaviour.

As well as examining the base scenarios, the paper explores potential scenarios in which many more Stockholm residents work from home, resulting in depressed work-based travel and less need for new office development. The paper will offer input to the City’s planning process and serve as a case study in how dynamic scheduling models can integrate with evaluation processes in planning policy.

Mahmoud Arbouch (p), Eduardo A. Haddad, Vinicius A. Vale, Inácio F. Araújo

Stephen McCarthy

There is a long tradition in transportation planning research in quantifying remoteness in the context of accessibility indicators. By considering the network structure of an existing system and the performance of its components, it is possible to assess the socio-economic impacts of changes in the physical characteristics of specific links that affect critical locations. Nonetheless, the link between changes in accessibility and their socio-economic consequences frequently relies on using parameters estimated econometrically under partial equilibrium frameworks. More recently, modeling integration based on links between transportation networks and computable general equilibrium (CGE) models has been gaining attention from the research community. In this paper, we add to this research trend by examining the case of remote regions in Morocco. The physical transportation network in the country is particularly interesting for understanding the economic costs of isolation. We calibrate a spatial CGE for Morocco and integrate existing interregional trade flows to the transportation infrastructure to simulate the potential higher-order economic impacts of a road project improving the access of regions in the Southern part of the country.

Sandy Dall'erba (p)

Mahmoud Arbouch

We adapt a Ricardian general equilibrium model to the setting of U.S. domestic agri-food trade to assess states’ vulnerability to adverse production shocks and supply chain disruptions. To this end, we analyze how domestic crop supply chains depend on fundamental state-level comparative advantages – which reflect underlying differences in states’ cost-adjusted productivity levels – and thereby illustrate the capacity of states to adapt to and mitigate the impacts of such disruptions to the U.S. agricultural sector. Based on the theoretical framework and our estimates of the model’s structural parameters obtained using data on U.S. production, consumption, and domestic trade in crops, we undertake counterfactual simulations to characterize the welfare implications of counterfactual scenarios depicting disruptions to (1) states’ agricultural productive capacity, and (2) interstate supply linkages. Our results emphasize that the distributional impacts of domestic supply chain disruptions hinge on the extent of individual states’ agricultural productive capacities, and that the capacity for states to mitigate the impacts of adverse production shocks through trade relies on the degree to which states are able to substitute local production shortfalls by sourcing crops from other states.