要旨：The standard approaches to studying industrial agglomeration have been in terms of summary measures of the "degree of agglomeration" within each industry. But such measures often fail to distinguish between industries that exhibit substantially different spatial scales of agglomeration. In a previous paper, Mori and Smith (2014) proposed a new pair of quantitative measures for distinguishing both the scale and degree of industrial agglomeration based on an explicit method for detecting spatial clusters. The first, designated as the global extent (GE) of industrial clusters, measures the spatial spread of these clusters (within a given country) in terms of the areal size of their essential containment, defined to be the (convex-solid) region containing the most significant subset of these clusters. The second, designated as the local density (LD) of industrial clusters, measures the spatial extent of individual clusters within their essential containment in terms of the areal share of that containment occupied by clusters. The central purpose of the present paper is to apply these two measures to the manufacturing industries in Japan, and to demonstrate how they can be used in combination to distinguish both the relative scale and degree of agglomeration exhibited by cluster patterns for each industry. In addition, the information provided by this pair of measures (GE, LD) is systematically compared to that of the most prominent summary measures currently in use. Finally, it is shown that these measures also support certain predictions of new economic geography models in the sense that shipping distances for establishments in each industry tend to be negatively (positively) correlated with the GE (LD) measures of agglomeration in these industries.

要旨：An empirical finding by Trefler (2004, AER) and others that industrial productivity increases more strongly in liberalized industries than in non-liberalized industries has been widely accepted as evidence for the Melitz (2003, Econometrica) model. We show that under fairly standard assumptions, a multi-industry version of the Melitz model does not predict this relationship. Instead, it predicts the opposite relationship that industrial productivity increases more strongly in non-liberalized industries than in liberalized industries.

要旨：This paper investigates whether agglomeration discourages households’ fertility decision making. A theoretical prediction shows that higher costs of education and housing for rearing children in denser areas have negative effects on fertility. In an empirical analysis, we exploit social survey data to control for households’ unobservable characteristics. Furthermore, we control for costs of migration, which also discourages fertility. We offer robust evidence that agglomeration measured by population density discourages fertility. We also find that rural-urban migrants have fewer children than stayers in rural and urban areas, suggesting that households who move to urban areas are strongly affected by costs arising from not only agglomeration but also migration. On the other hand, there is no significant difference in the number of children between urban-rural migrants and stayers in urban and rural areas.