John Gibson (p), Yi Jiang, Bambang Susantono

Federica Galli

There is increasing interest in assessing whether growth of big cities has effects that differ from effects of growth of secondary towns, especially for impacts on poverty. It can be difficult to study these issues with typical sub-national economic data for administrative units because urban growth often occurs outside of the administrative boundaries of cities. An emerging literature therefore uses remote sensing to measure patterns of urban growth without being restricted by limitations of data for administrative areas. We add to this literature by combining remote sensing data on night-time lights for 41 big cities and 497 districts in Indonesia with annual poverty estimates from socio-economic surveys, using spatial econometric models to examine effects of urban growth on poverty during 2011-19. We measure growth on both the extensive (lit area) and intensive (brightness within lit area) margins, and distinguish between growth of big cities and of secondary towns. The extensive margin growth of secondary towns is associated with lower rates of poverty but there is no similar effect for growth of big cities. The productivity advantages of big cities and concerns about agricultural land loss to expanding towns and cities may imply that urban growth patterns favoring big cities are warranted, while on the other hand these new results suggest, from a poverty reduction point of view, that policies to favor secondary towns may be warranted. Policymakers in countries like Indonesia therefore face difficult trade-offs when developing their urbanization strategies.

Zoltán Egri (p)

John Gibson

The issues of regional income inequality have been prominent in both academic research and economic and territorial policy ideas over the last two to three decades. The topic has aroused particularly intense interest in Central and Eastern Europe and Hungary in the period of regime change and beyond.
The overall aim of the presentation is to analyze the effects of the territorial income inequality affecting Hungary in the period following the global crisis that started in the United States in 2007-2008.
The territorial entity of the study is the Hungarian settlement (3154 pieces).
Using spatial econometric methods (spatial and LISA Markov-chains, kernel function estimation, etc.), we describe the main features of income inequalities. With our analyzes, we point out the spatial club convergence at the municipal level, and the forces shaping the phenomenon.

Christoph Hauser (p), Gottfried Tappeiner, Janette Walde

Zoltán Egri

Social trust is increasingly seen as an important determinant of economic growth and social prosperity in regions and nations. Even in a comparatively homogeneous area such as Europe, there are stark sub-national differences in levels of generalized trust. It is thus of crucial importance to identify the driving forces of regional trust and analyze the dynamics of its formation. The present paper considers these issues based on three waves of the European Values Study. Evidence is provided to demonstrate that values of regional trust remain substantially stable over an approx. 20-year period and are modified only through spatially correlated random noise processes. This finding is consistent with additional analyses identifying slow-moving factors that are responsible for the spatial distribution of trust scores and are buried deep in the cultural background of a society. Hence, in spite of its economic significance, social trust does not appear to be amenable to political intervention in the short to medium term.

Federica Galli (p)

Christoph Hauser

In the last two decades, scholars started to expand classical stochastic frontier models (SF) in
order to include also some spatial components. Indeed, firms tend to concentrate in clusters,
taking advantage of positive agglomeration externalities due to cooperation, shared ideas and
emulation, resulting in increased productivity levels. Thus, producers cannot be regarded as isolated entities and the hypothesis of cross-sectional independence underlying the basic SF models must no longer be considered valid. Until now scholars have introduced spatial dependence in SF models following two different paths: evaluating global and local spatial spillovers affecting the frontier function or considering spatial cross-sectional correlation in the inefficiency and/or in the error term.
The model proposed in this work combines the two different modelling approaches obtaining a full comprehensive specification that introduces four different sources of spatial cross-sectional dependence. Specifically, we introduce the spatial lag of Y and of X to capture global and local spatial spillovers affecting the frontier function and we also add a spatial structure to the inefficiency and to the error term to capture respectively behavioural and environmental spatial correlation. Hence, we obtain a spatial Durbin stochastic frontier model for panel data introducing cross-sectional dependence both in the inefficiency and in the error term (SDF-CSD).
The most appealing feature of our model is that it allows to capture global and local spatial spillover effects while controlling for spatial correlation related to firms' efficiency and to unobserved but spatially correlated variables. The SDF-CSD model can be estimated using maximum likelihood techniques, modifying the estimation procedure suggested by Orea et al. (2019) in order to consider the endogeneity deriving from the inclusion of the spatial lag of the dependent variable. Implementing some Monte Carlo simulations, we show that our spatial estimator is able to distinguish between frontier and error-based spillovers considering sparse spatial weight matrices (as binary contiguity or truncated inverse distance matrices). Finally, an application to the Italian agricultural sector is provided.