Our conclusions are that the most important influences on unemploymentudcome from the following.ud(i) The longer unemployment benefits are available the longer unemploymentudlasts. Similarly, higher levels of benefits generate higher unemployment, withudan elasticity of around one half. On the other hand active help in findingudwork can reduce unemployment. So more “flexibility” may need to beudcomplemented by more intervention to provide active help.ud(ii) Union coverage and union power raise unemployment. But if wageudbargaining is decentralised, wage bargainers have incentives to settle forudmore than the “going rate”, and only higher unemployment can prevent themudleap-frogging. Although decentralisation makes it easier to vary relativeudwages, this advantage is more than offset by the extra upward pressure on theudgeneral level of wages. Thus, where union coverage is high, coordinatedudwage bargaining leads to lower unemployment.ud(iii) Conscious intervention to raise the skill levels of less able workers is anudimportant component of any policy to combat unemployment. Pure wageudflexibility may not be sufficient because it leads to growing inequality whichudin turn discourages labour supply from less able workers.udThus in these areas it is clear what types of reform are needed. If welluddesigned, such reforms might halve the level of unemployment in many countries.udBut there are three other remedies which have been widely advocated in bothudthe OECD Jobs Study and the Delors White Paper. These are: less employmentudprotection, lower taxes on employment, and lower working hours. Our researchuddoes not suggest that lower employment taxes or lower hours would have any longudterm effects; while the effects of lower employment protection would be small.ud(iv) Lower employment protection has two effects. It increases hiring and thusudreduces long-term unemployment. But it also increases firing and thusudincreases short-term unemployment. The first (good) effect is almost offsetudby the second (bad) one. The gains from flexibility are small.ud(v) Employment taxes do not appear to have any long-term effect onudunemployment and are borne entirely by labour. There may be some shorttermudeffects, but it is not clear that there would be any fall in inflationaryudpressure if taxes on polluting products were raised at the same time as taxesudon employment were lowered.ud(vi) Hours of work appear to have no long-term effect upon unemployment.udEqually, if early retirement is used in order to reduce labour supply, it isudnecessary to reduce employment pari passu unless inflationary pressure is toudincrease. While flexible hours and participation can reduce the fluctuationsudin unemployment over the cycle, they cannot affect its average level.
展开▼
