17 / 20
17(20)
5
Conclusion
Monitoring and screening have been shown to be empirically important in
reducing the (ex post) moral hazard in social insurance programs. It is easy
to show theoretically that ex ante or deterrence effects of the monitoring
could also be important in reducing the take up rate of the programs. The
empirical support of ex ante or, when it comes to monitoring, deterrence
effect are however, basically non existing. In the tax evasion literature
Kleven et al. (2011) have shown that the deterrence effect on self reported
income on audits the year before could be substantial. This results support,
in general, the results from “threat of audit” letters in the same literature
(see Coleman (1996), Slemerod et al (2001) and Hasseldine et al (2007)
and Kleven et al. (2011)).
Our result supports the results in Engström et al. (2007) who sent out
“threat of monitoring letters” in the Swedish temporary family insurance.
Their results showed that the parents who received the “threat of
monitoring” decreased their use of the benefit by 13 per cent. Our results
show that individuals respond on information of (increased) screening.
The advantage with our study is that we use the normal SSIA monitoring
routine. This routine has, furthermore, been stable over a longer period.
One result from our study is that individuals on average update the risk
of being detected of misusing the insurance. This means that there exists
an equilibrium effect from monitoring. Our intent-to-treat estimates show
on average a 1.4 per cent decrease in the benefit days one year after
assignment to monitoring. Given that only around 7 per cent of the parents
are directly contacted (they could however also receive information on the
monitoring from employers or day care) the estimate in Engström et al.
(2007) is not unreasonable high.
The consequences of misusing the parental leave system are very mild
and the degree of monitoring is quite low. In addition, our intent-to-treat
estimate is most likely biased toward zero. Even so, we found an economic
significant effect of monitoring on later take up rates. This means that
there exists deterrence effects that are larger and more important in
programs with larger sanctions and in programs with a higher degree of
monitoring.