Detecting dynamic processes of agglomeration in the product life cycle and sudden changes using space-time clustering analyses

Agglomeration economies are dynamic in nature since they are based on cyclic processes and dynamic interactions among agglomerative factors. However, most studies have only investigated the static phenomenon of agglomeration and spatial concentration and have ignored the possible changes. This dissertation hypothesizes that agglomeration economies are more obvious as both spatial concentration and the rapid change in the amount of new firm activity are considered simultaneously. In other words, the dissertation suggests that time-space clustering better represents the existing agglomeration economies than spatial concentration alone. In particular, time-space clustering is observed during the initial stage of the product life cycle and sudden changes of historic accident and discovery. For the empirical test, three methods, the Jacquez test, the space-time K-function and Kulldorff's scan statistic, were applied using the ES-202 data of the Columbus MSA. The study discovered that knowledge based industries and FIRE services represent the agglomeration process in the Columbus MSA due to the product life cycle and sudden change and that industry restructuring is taking place from traditional manufacturing industries to knowledge based industries and FIRE services in the study area. In addition, using the space-time K-function, the study identified the threshold of time-space clustering in which the agglomeration process is taking place.