Estimating matching functions when recruiting intensity matters

ABSTRACT

The article estimates matching functions – the relationships between new hires, job vacancies and unemployment – using monthly time series data for Sweden. The data contain rich information on types of job vacancies which provides information on firms’ recruiting efforts. The analysis shows that the aggregate flow of new hires is well explained by an augmented version of a standard matching function that incorporates data on firms’ recruiting intensity.

KEYWORDS:

• Unemployment
• vacancies
• job matching

JEL CLASSIFICATION:

• J63
• J64

Acknowledgement

Financial support from FORTE is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Vacancy statistics based on registers at the public employment offices have been available for decades and frequently used in empirical studies.

² The vacancy statistics are published by Statistics Sweden in ‘Konjunkturstatistik över vakanser’ (Job openings and unmet labour demand), whereas the employment data are available in ‘Kortperiodisk sysselsättningsstatistik’ (Short-term employment). An early discussion of survey-based vacancy statistics in Sweden is provided by Farm (1997).

³ The fifth category is labelled ‘vakanser’ (vacancies) in the Swedish statistics, whereas the sum of all five categories is labelled ‘lediga jobb’ (job openings). The terminology is slightly confusing since it implies a distinction between job openings and vacancies. We use ‘vacancies’ as a generic term.

⁴ A regression of the ratio of active to total vacancies on the unemployment rate (using seasonally adjusted quarterly data) yields a coefficient of −1.53 on the unemployment rate (absolute t-value = 2.7; R-sq. = 0.17). The procyclicality of recruiting intensity is consistent with the findings on US data presented in Davis et al. (2013) which are based on various indirect methods (absent direct measures of vacancy types).

⁵ It is notable that the share of the population born abroad and aged 25–54 increased from 15% in 2001 to 24% in 2015.

⁶ A classic paper by Holt and David (1966) begins by stating that ‘although the systematic collection of data on job vacancies goes back to the First World War, the vacancy concept has yet to be adequately refined, and a comprehensive theory of the labor market into which such a concept could be fitted has yet to be developed.’

Additional information

Funding

Financial support from FORTE is gratefully acknowledged.