Do Employees’ Sickness Absences React to a Change in Costs for Firms? Evidence from a Natural Experiment*

• DOI: http://doi.org/10.1111/sjoe.12335
• Date: 01 April 2020
• Published date: 01 April 2020

Scand. J. of Economics 122(2), 553–581, 2020

DOI: 10.1111/sjoe.12335

Do Employees’Sickness Absences React to a

Change in Costs for Firms? Evidence from a

Natural Experiment*

Ren´eB¨oheim

WU Vienna Universityof Business and Economics, 1020 Vienna, Austria

rene.boeheim@wu.ac.at

Thomas Leoni

Austrian Institute of Economic Research (WIFO), 1030 Vienna,Austria

thomas.leoni@wifo.ac.at

Abstract

Weanalyse the impact of a social security reform that changed the costs incurred by firms due to

sickness absences. The reform abolished a compulsory insurance for firms, which insured them

against the wages paid to sick blue-collar workers.During the first year after its introduction, we

estimate that the reform resulted in about 6.3 percent fewer sickness absences, and in about 8.6

percent fewer absence days.We do not find evidence for changes in hiring or firing, and we find

only limited workforce composition changes.We do not find spillover effects on the absences of

white-collar workers. Robustness checks confirm these results.

Keywords: Absenteeism; moral hazard; sickness insurance

JEL classification:I38; J22

I. Introduction

Sickness absences lead to significant productivity losses, and consequently

reduce incomes and profits (Allen, 1983; Coles and Treble, 1993; Brown

and Sessions, 1996; Barham and Begum, 2005; Osterkamp and R¨ohn,

2007). Firms in most OECD countries are at least partially insured against

the direct costs arising from their workers’ sickness absences. In some

countries, the amount or period of statutory sick pay is limited. For example,

*Support from theAustrian National Research Network “Labor Economics and the WelfareState”

is gratefully acknowledged. We are also grateful to Stuart Adam, Mike Brewer,Natalia Danzer,

Marco Ercolani, Martin Halla, Helmut Rainer, Rudolf Winter-Ebmer, Martina Zweim¨uller, and

seminar participants in Colchester, Engelberg, Essen, Linz, London, Munich, and N¨urnberg

for their valuable comments. Bj¨orn Fanta and Clemens Kozmich provided excellent research

assistance. R. B¨oheim g ratefullyacknowledges CESifo Munich’s hospitality.R. B ¨oheim is also

associated with the Johannes Kepler University Linz, CESifo (Munich), IZA (Bonn), andWIFO

(Vienna).

C

2018The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨oreningen

f¨or utgivande av the SJE/The editors of The Scandinavian Journal of Economics.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution

and reproduction in any medium, providedthe original work is properly cited.

554 Costs of sickness absences

Norwegian firms need to pay workers their full wages during the first 16

days of a sickness absence, and social security pays the wages thereafter

(Markussen et al., 2012). In other countries, firms are refunded for the

costs incurred when their workers are sick. For example, in Germany, firms

that have fewer than 30 employees have part of their costs refunded by

an insurance fund (Ziebarth and Karlsson, 2010). Depending on how such

insurance is organized, firms might have an incentive to exert very little

effort in monitoring or in preventing absences, which in turn might lead to

more sickness absences.

Economic research has devoted considerable effort to determine if sick

pay leads employees to “adapt their work-absence behavior” (Johansson and

Palme, 2005, p. 1880), and to remain absent from work without actually

being sick.1The growing body of literature on absenteeism has identified

several other determinants of sickness absence.2

However, there is little evidence on how changes in the costs incurred

by firms affect sickness absences. Westergaard-Nielsen and Pertold (2012)

find that a voluntary insurance scheme for small companies in Denmark

led to more short-term sickness absences, possibly from lower levels of

monitoring. Fevang etal. (2014) analyse a Norwegian reform that exempted

employers from refunding sick pay only for pregnancy-related absences.

They show that this exemption led to approximately 5 percent more sickness

absence spells of pregnant women.

We exploit a reform of the social security system in Austria to provide

causal evidence for the effect of employers’incur red costs on their workers’

absences. The reform provides a unique situation that allows us to study

whether employees’ absences changed after an exogenous variation in

firms’ costs. The mandatory insurance insured firms against the direct

costs incurred because of their blue-collar workers’ sickness absences until

1There is substantial evidence that both the availability and the level of sickness benefits

influence the extent of absenteeism. Johansson (1996), Johansson and Palme (2005), and Hall and

Hartman (2010) find such a moral hazard for Swedish workers. Similarly, Ziebarth and Karlsson

(2010) and Puhani and Sonderhof (2010) show that a reduction in the replacement rate for sick

workers in Germany substantially decreased sickness absences and hospitalization days. In the

United States, where workersare not universally covered bysickness insurance, numerous studies

investigate the moral hazard associated with workercompensation schemes. Krueger (1990) and

Hirsch et al. (1997), among others, find that these insurance schemes have large incentiveeffects

on workers’behaviour,although more recently, Bronchetti and McInerney (2012) havechallenged

the prevailing wisdom that workers are highlyresponsive to changes in benefit levels.

2These include: workers’ health status (Delgado and Kniesner, 1997) and gender differences

(Ichino and Moretti, 2009); social norms, peer-group dynamics, and infection effects on the

workplace (Drago and Wooden, 1992; Ichino and Maggi, 2000; Barmby and Larguem, 2009);

working conditions, work arrangements, and work contracts (Darr and Johns, 2008; Riphahn,

2004; Ichino and Riphahn, 2005; Dionne and Dostie, 2008); business cycle effects on labour

force composition (Askildsen et al., 2005); and doctors’ behaviour (Markussen et al., 2011).

C

2018The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨oreningen

f¨or utgivande av the SJE/The editors of The Scandinavian Journal of Economics.