The report does not provide an exhaustive overview of all individual occupations or educations. Instead, we try to assess more general questions, such as whether we find signs of changes in the relationships between education and occupation over time or linked to business cycles, as well as whether data for advertised positions can give us a detailed picture of demand for different types of qualifications.

We find that the concentration of education within occupation (and of occupation within education) varies significantly between occupations. The concentration is highest within licensed occupations (occupations that require a specific education) and other professionally oriented occupations, which are dominated by one or a few educations. The concentration is lower in a number of service-providing occupations, which have no clear "typical" education.

The share of employees in concentrated occupations has fallen somewhat since 2003 but does not vary significantly with the business cycle (measured as registered unemployment). The distribution of concentration is also approximately stable from 2015 to 2019, regardless of economic conditions, both for all employees and for new employees. There may be changes in the degree of concentration within individual occupations or educations, but overall, we find no signs of changed relationships between education and occupation. We also do not find that businesses in periods of a tight labor market react by hiring people with other types of education in a given occupation.

In Statistics Norway's projections of employment by education, it is assumed that the educational composition within industries develops steadily, and as in the recent past. The analyzes in this report do not provide any basis for concrete changes in these projections.

The distribution of occupations among advertised positions at industry level corresponds well with the distribution of occupations among employees at aggregate industry level, while it is more difficult to link advertised positions and new hires at enterprise level. This may be due, among other things, to differences in the advertised occupation and the occupation of the new employee, as well as variation in the time between the advertised position being taken off the market and the start of the new employee. We also find that the number of new hires is higher than the number of advertised positions. This discrepancy has become smaller over time, but it still appears that some positions are not advertised via formal channels.

The overall correspondence between advertised positions and new hires suggests that advertised positions can contribute to a more comprehensive measure of demand for labor. But this requires a better understanding of practice related to the advertisement of positions, and how this varies between industries and professions, among other things.