Changes in employment localization and accessibility: The case of Switzerland between 1939 and 2008 - Recomposed Data

Abstract

The main goal of this work is to study the spatial repartition of jobs, in particular the non- agricultural ones, and to try to find a relationship between the evolution in their repartition and the concomitant evolution of accessibility. In order to study the spatial repartition of jobs through time, only two sources can be chosen. The first is the population census, which currently records, for all active persons, an array of information regarding their activity: their learnt and current profession, the location of their job place at the communal level at worst, the economic specialization of their company of institu- tion, their activity rate and status, in all, a wealth of information which would cover nicely our work. However, the data from population census suffers from two fundamental flaws with re- gard to our thematic. The first is that it is population based. It will then give us all information needed about the people who work, but not quite about their workplaces. In particular, all in- formation pertaining to the size of their companies, whether they are controlled from the out-side or not, and many similar information are lost to the researcher. Furthermore, only Swiss residents are asked about their working places, which would exclude the cross-border workers and thus severely underestimate the economic size of border areas which rely heavily on them. The greatest default of the population data, though, is that information about the work place location is given only from 1970 on – before 1970, all professional and activity information is solely referred at the residential place, which makes it impossible to research older phases of job location deployment. The second possible source is the business census. Unlike the population census, the business census uses the establishment, a localized part or whole of a company, as its base unit. Thus, all the information pertains to a company or a branch which is unequivocally lined to a given site, a necessary information when looking after the job repartition through time. Business censuses have been held in Switzerland since 1905, which gives us the temporal depth of information we are looking into, and give results at the communal level, which gives us the territorial detailwe’re basing our study on. However, business censuses have some shortcomings. The first is that it doesn’t include in its count the primary sector, which was censed separately for most of the period under review. Thus, data must be reconciled in order to obtain a global view of employ- ment in Switzerland. Secondly, and more annoyingly, for a long period of time business censuses were censing only entities active on the free market, excluding, for instance, most public services and administrations and even some liberal professions which were not thought to be of business orientation, notably in the health and law services. Lastly, while the first edition of the business census in 1905 was very detailed, some subsequent editions were less so and up to 1965 in- cluded, economic activity wasn’t detailed at the communal level.