Angelika Krehl (p)

Mark van Duijn

This paper suggests an exploratory means to distinguish several urban spatial configurations from each other. Using the local Moran’s I permits the identification of statistically significant spatial employee densifications in different settings: entire regions and counterfactuals obtained by omitting the core city (‘donut model’). It is assumed that the more polycentric a region is the less influential the core city omission should be. One possible theoretical rationale behind this exploratory approach is the presence of agglomeration shadows, which mask much of the variation of employee densifications in the vicinity of dominating core cities. By identifying spatial clusters of employees with the help of the local Moran’s I it is shown that the method is a promising way to visualize agglomeration shadows. The identification of corona-like employee densifications in the ‘donut models’ furthermore speaks in favor of spatially effective agglomeration shadows. The analyses conducted for four German city regions on the spatial scale of 1 km² grid cells reveal that counterfactuals are a sensitive approach to identifying polycentricity in an urban region. They also highlight that the donut model approach is feasible to distinguish polycentricity from other spatial configurations such as multi- and monocentricity.

Rodger Barros Antunes Campos (p), Gustavo Pereira Serra

Mark van Duijn

Earmarking taxes is a common policy mechanism in Brazil. The Brazilian Federal Constitution of 1988 earmarks a large part of tax revenue to expenditures considered essential, such as education, public health and social security. In 1989, the government of the state of São Paulo passed Law 6.556, increasing the value added tax from 17% to 18%, and earmarking the incremental revenue to low-income housing policies. The Brazilian Supreme Court declared such a law illegal in 1997. We apply a non-linear Diff-in- Diff approach to evaluate the impact of this ruling on the housing deficit of the state. By comparing the situation pre and post ruling, we found that this decision increased the probability of a household face inadequate housing conditions (many families in one house; more than three persons in the same room; inadequate construction, etc.). Our study contributes to the debate on the role of earmarking in overcoming housing deficit.

Özge Öner (p)

Mark van Duijn

Do urban amenities, such as consumer services, shops, and arts & entertainment facilities, explain income sorting within a city? Using unique geo-coded micro-data from Sweden for 2001-2013, I estimate the likelihood of moving across the neighborhoods of Stockholm metropolitan region for individuals with different income levels to determine with respect to various types of urban amenities at the neighborhood level. The analysis also takes into account centrality, and various individual factors that may relate to within-city mobility. In line with previous theories for endogenous amenities, I find that average income in the neighborhood is an amenity itself, and the availability of amenities –over and above the size of the neighborhood- implies a push-factor for the low-income individuals and a pull-factor for the high-income individuals. Further analysis signals that density disincentives individuals to move to moderate and high-amenity neighborhoods, as it indicates a crowding out effect, whereas it appears to be a pull-factor for the neighborhoods with no or very few amenities, potentially signalling a tradeoff between urban amenities and the availability of large housing units.