2016 年 46 巻 2 号 p. 199-212
Provincial cities are at a disadvantage in attracting industrial plants due to a low degree of industrial agglomeration. However, they have an advantage in low wages and land prices. We analyzed how plant location preferences vary according to firm types （that is, industry sector and firm size） by a conditional logit model. We then discuss the type of firms more （or less） likely to be attracted to small provincial cities.
First, we used the likelihood ratio test to assess whether plant location preferences are statistically different among firm types. Second, using the estimated results, we calculated how much each locational factor must change to compensate for the decrease in location probability caused by a reduction of one extra unit in the industrial agglomeration.
The results showed that large sized firms compensate for a reduction in the industrial agglomeration with larger changes in locational factors （example, wages and land prices） than small sized firms. This implies that the policy costs （example, recruitment and investment subsidies） for attracting large sized firms to provincial cities is more expensive than for small firms. Additionally, firms belonging to processing and assembling industries compensate with smaller changes in wages, job-to-applicant ratio for high school graduates, and a length of highway rather than firms belonging to basic material industries. On the other hand, firms belonging to basic material industries compensate with smaller changes in industrial parks than firms belonging to processing and assembling industries.
JEL Classifications: R10, R11, R12, R15
