Beijing has undergone rapid urbanisation from 1950-2010. The main built-up area expanded concentrically from 100 to 1210 km2. Its urban fringe became popular for new developments, some placed haphazardly. The government concentrated efforts in 'stopping concentric fringe expansion'. E.g. on neighbourhood and district levels, the government intended to densify roads, increase neighbourhood density and improve public transport. However, the effects remain unclear. The LUTI model is widely used to assess the effects of planning policies at entire city scale, and generate the whole metropolitan area's spatial pattern. However, it is insensitive to micro-level (neighbourhood or district) built form change. E.g. there are two policy scenarios with the same floorspace supply quantity, one case is that all houses concentrate around road nodes, the other that houses distribute sparsely. Given equal floorspace provisions, regardless of built form, the LUTI model will compute the same spatial patterns. However, in reality, different built forms will lead to different policy performances. Most LUTI models, to the best of our knowledge, are incapable of measuring this difference. Therefore, this paper augments existing LUTI models with a LUTI framework compliant built form component, allowing investigations into the impact of built form on whole city function. In our newly assembled model, we quantify the effects of built form change on route selection change. Then, the micro-level travel behaviour change is passed to the macro-level MEPLAN transport model (Echenique, 2004, 2011). Finally, we input the MEPLAN results, including generalised travel cost, distance and time, into a LUTI framework (Jin, Echenique, & Hargreaves, 2013) to simulate population distribution, changes on price level and spatial costs. We then apply this model to Beijing's case in order to: validate the micro built form component in the general LUTI framework; predict the impacts of alternative built forms in reshaping the city's fringe; gain insights into how built form at the urban fringe shapes the bigger scale spatial structure. We first calibrate a 130-zone recursive spatial equilibrium model using observed data for 1990-2010. Alternative built forms at the urban fringe for 2000 and 2010 that share the same floorspace provision are tested. These scenarios include densification around subway stations and decreasing road density in built-up area. The model results will show each area's spatial cost from its built form. The analyses suggest that under rapid transformative urban change, the built form has an indirect but significant impact on the economic performance of the entire city, through travel behaviour change. In densification scenarios, access trip length decreases, as does the spatial cost and price level. In road density restriction scenarios, spatial costs increase in the entire city. These results allow understanding of the impact to whole city economic performance from built form at the neighbourhood scale. They also provide insights into designing better built forms, as such design on neighbourhood and district scale are more feasible in practical implementation - not only for cities in emerging economies but also fast-growing cities in the developed economies that are revisiting the design.
