the working class to redistribute part of the higher wages to the 
unemployed. So the microquestion of why wages are not flexible 
downwards is not so easily separable from the macroquestion of the 
effects of a decrease of real wages on unemployment. (This view seems 
to me supported by the recent experience of Swedish trade unions 
which accepted cuts in real wages of, I understand, around 30%, in 
order to permit the restructuring of the economy necessary to maintain 
full employment in the face of tougher world competition; the labour 
movement will often be ready to trade off real wages for employment, 
when it has reason to believe that such a tradeoff really exists.) In 
other words, neoclassical economists are not wrong in believing that, if 
a (reasonable) decrease of real wages would ensure full employment, 
then the downward rigidity of real wages would be something of a 
puzzle. Where they are wrong is in believing that the premise of this 
hypothetical statement is true: why in this they are wrong cannot then 
be left out of an appraisal of the present state of labour economics. 
(As to your reply that a sufficient decrease of real wages will 
always increase jobs by increasing the demand for domestic servants, I 
am still thinking about it. I think you would agree that even this 
increase of demand for labour encounters limits in the fact that the 
cost of domestic labour services cannot be indefinitely decreased by 
decreasing wages, because it also includes other costs - e.g. transport 
costs if the servants live elsewhere, or housing costs if they live with 
the employer - and because there is a ’subsistence’ lower limit to 
wages: servants must be in good health to be able to work and not 
transmit diseases to employers, etc.) 
Yours very truly 
Siena, 15 September 1990 
Dipartimento di Economia Politica, 
Universita di Siena 
Piazza S. Francesco 7, Siena 53100 
Italia