Analysis of spatial concentration and dispersion

The aim of this chapter is to present a number of statistical approaches to the study of the spatial concentration and dispersion of economic activities. Traditionally the problem of the spatial location of economic activities has been approached by looking at the distribution of the agents within geographical partitions such as administrative units, regions and municipalities: the so-called ‘mesoeconomic’ approach. Such an approach will be considered in Sections 2 and 3 of the present chapter, with an explicit consideration of the problem of the spatial correlation among units. However, a mesoeconomic analysis is fundamentally undermined by the ‘modifiable areal unit problem’ (or MAUP; see Arbia, 1989) in that any conclusion depends intrinsically on the specific geographical partition chosen and it could be very different if it referred to a different partition or to the single economic agent. To tackle this issue a second approach has been increasing in popularity in recent years, where the phenomena of concentration are analysed on a continuous (rather than discrete) space, an approach that finds its theoretical justifications in the earlier contributions of Hotelling (1929), Christaller (1933), Palander (1935), Lösch (1954) and Isard (1956) and in the studies of Paelink and Nijkamp (1975) and Paelink and Klaassen (1979) in the 1970s. In Section 4 we will review a set of spatial statistical models derived from the so-called ‘point-pattern analysis’ that built up a new spatial microeconomic approach to the analysis of spatial concentration of human activities.