An Agent-Based Simulation of the Swiss Labour Market : An Alternative for Policy
Assuming a higher initial-wealth of 15 and a lower strict value of 0.98 , the effect of the “ incremented subsidy ” disappears . This implies that , according to the model , subsides are not effective if the labor market is more flexible , and therefore policymakers should invest in different policies ( i . e ., support for job search ).
4.2 - Job displacement effects
A second experiment concerns the analysis of the treatment effects on the individual and the macro-economic level . From a theoretical perspective ALMPs cause an increase in the size of the labor force ( allowing more people enter in the market ) and this involves more competition in the labor market and could produce the opportunity for firms to dismiss “ nonsubsidized ” workers and to replace them with “ subsidized ” workers ( Froy & Giguère , 2010 ). The simulation assumes therefore that the subsidy “ improves ” the history of the applicant , increasing his hiring probability .
As in the previous experiment worker agents are divided into two groups , one of treated and one of non-treated workers ; the agents from the non-treated group do not receive subsidies . The experiment is divided into two phases . In a first simulation round , no subsidies are given to neither the treated workers nor to the non-treated workers , then in a second round , the agents that belong to the treated group are subsidized . Finally , the last differences between the employment rate of treated and non-treated agents for both round are compared .
The user input values are set according to the values contained in Table 8 .
In this case , after 1,000 steps , results show a negligible difference in the average final employment , showing that the macro effects , as the total skill level of the applicant or his human capital level , exceed the signal effect of the subsidy on the workers ’ history . This result , which is contradicted by empirical evidence , constitutes a point for further improvements of the model .
Table 8 — Experiment Job displacement — parameters
Parameter
Value
Job displacement : NFirm [ 15 ] NWork [ 50 ]
initial _ wealth |
[ 10 ] |
strict |
[ 0.99 ] |
thrsprev [ 4 ] memory-size [ 5 ] number-strategies [ 10 ] thrsflex2 [ 0.7 ] thrsflex3 [ 0.5 ] JobDisplaceSub [ 0 ] foreign [ 0 ]
Table 9 — Experiment Job displacement — effects
Job displacement effects
First round treated workers 4.61 %
non-treated workers
Second round
Unemployment rate
4.62 %
treated workers 4.57 %
non-treated workers
4.68 %
26