Market potential and the employment growth of knowledge-intensive services: comparing different geographical resolutions

In order to empirically understand the geographical reach of different markets, it is vital to use an appropriate geographical resolution. Using too large observational units risks hiding the interesting relationships within the regional boundaries. In this study, we aim to investigate and compare similar analyses performed on different geographical levels, with a special focus on innovative industries. Accessibility to markets, services and infrastructure is thought to be major determinants of the potential for economic development and welfare of a region. Earlier empirical research establishing the relationship between agglomeration forces and regional growth typically includes a measure for accessibility or market potential as an explanatory factor. The geographical scale that conventional accessibility measures operate on is usually on the level of municipalities or similar, even when theory suggests that a more disaggregated scale is desirable. Most often the reason for this is limitations in available data. In many cases, the researcher is left with a geographical level based on administrative borders. Analyses on more disaggregated levels allow the researcher to better pinpoint the actual accessibility that each firm faces. In order to shed light on the importance of these issues, this paper utilizes an exploratory approach to investigate the relationship between the spatial distribution and growth of knowledge-intensive services (KIS) and the accessibility to economic activity (market potential). We use regional employment growth in KIS as a proxy for regional innovativeness. The relationship is estimated on two different geographical levels using Swedish data. The more conventional model is estimated with the 290 municipalities in Sweden as the units of analysis. In the Swedish context, this represents the geographically smallest administrative level. In the more novel model, we use the 298 so-called SAMS areas of Jonkoping County in Sweden. Our results show that the detailed level is particularly important for the analysis of the growth of the more advanced sectors of the economy, in our setting, the high-tech knowledge-intensive services.