Collection of job ad data:
Representative annual random samples of job advertisements drawn from the press are available for the period since 1950 and since 2001 also from the Internet. The press survey is based on a random sample of some 70 advertising papers and newspapers published in German-speaking Switzerland that is stratified according to region and circulation. Approximately 30 titles from French- and Italian-speaking have been included since 2001. The Internet survey collects ad data from the websites of a representative panel of businesses with locations in Switzerland. The initial sample – stratified by business size and industry – comprised some 1,000 businesses listed in the Swiss Business and Enterprise Register (BER). The sample was expanded to include 1,500 businesses in the course of panel readjustment in 2010. From 2006 on, the Internet survey has additionally covered the largest online job portals in Switzerland. In this process of expanding the sample base, the size of the annual ad samples has been increased from initially 500 job ads published in the press to now nearly 4,000 ads published through all advertising channels that are significant today. As a result of switching from a retrospective to a prospective survey mode in 2001, the job monitor provides current job market data on a yearly basis that are fully compatible with the full range of datasets dating back to 1950. This makes it possible to analyze developments in the job market with minimal delay and distinguish between long-term trends, business cycles, and new developments.
The SJMM Scientific Use File contains, first, the complete corpus of job ad data in raw text format, which can be used for text-based analyses. Job ads are defined as advertisements of job openings for gainful employment irrespective of the number of hours, duration of the employment relationship, and type of employment. Second, the SUF contains a dataset of information regularly provided in the ads, describing the job opening, the person wanted, and the business advertising the job. Standardized procedures are used to organize this information into categories that are appropriate for statistical analyses (coding). A high reliability of coding is ensured by providing clear, comprehensible, and comprehensive rules for coding on the one hand and intensive training of coding staff on the other to make sure that the coders interpret and apply these rules consistently.
Beginning in 2001, SJMM has conducted a standardized business panel survey parallel to the ad survey on a yearly basis. One of the purposes of this annual panel survey is to assess the growing or diminishing importance of various advertising media in order to adjust the set of media included in the ad survey as necessary so that it corresponds with businesses’ advertising practices.