Decline in sick leave continues in 1997
The average duration of sickness leave in Austria has declined considerably from 14.9 days in 1995 to an estimated 13.3 days in 1997, according to figures released in February 1998. Men are sick more often than women and for longer periods. A link between sickness levels and GDP growth or labour market conditions cannot be proven.
According to estimates by the Association of Social Security Providers (Hauptverband der Sozialversicherungsträger, HSV) released at the beginning of February 1998, the average duration of sick leave was 13.3 calendar days per worker in 1997. Final figures will be known only in June. The early estimate for 1996 was 14.7 days and the final figure was 14.0 days. Both the 1996 final data and the 1997 estimate are the lowest in well over a decade. This development continues a trend in evidence since 1991. During the period 1989 to 1991 the average annual duration of sick leave peaked at 15.3 days. Since then it has been falling, at first by roughly 0.1 days per year until it reached 14.9 days in 1995, and much faster since then. The data include all recipients of wages and salaries but exclude civil servants and most apprentices.
There is no bias towards particular days of the week: according to the HSV, days of sick leave were distributed equally across the week including the weekend and holidays. There is, however, a clear difference between men and women. Men consistently fall sick 10% to 15% more often than women, and their sickness spells are around 5% longer. As a result their average duration of sick leave per year has recently exceeded that of women by about 20%.
In a statement by the HSV's president on last year's estimate, the decline in sickness absence was linked to low GDP growth rates and adverse labour market conditions (quoted in Der Standard on 21 December 1996). However, this notion is not sufficiently confirmed by statistics. Correlations were run with real GDP growth (as defined nationally) and two labour market indicators, namely the rate of unemployment (as defined nationally) and the number of registered unemployed people per registered job offer. All coefficients show the expected signs, but GDP at 0.16 yields virtually no correlation. The two labour market variables do better at -0.45 and -0.60 respectively but, as with GDP, neither of them is significant. This does not rule out the possibility that a general feeling of economic insecurity, not captured in the data, may be contributing to the decline in the average duration of sickness. An alternative explanation could be an improvement in the health of the population.