LW
= unconditional elasticity:
how much does demand vary if wage rate changes
and choice of profit maximising output also changes?
LW
|
Y
= conditional elasticity:
how much does demand vary if wage rate changes
but output is kept constant?
Conditional elasticity reflects the possibility of factor
substitution between inputs of capital and labour
LW
=
LW
|
Y
+
LY
.
YW
Unconditional
LW
is sum of substitution and scale
effects
Scale effect is multiple of output elasticity of demand
for labour and elasticity of output w.r.t. the wage rise
Empirical values for demand elasticities
Cahuc & Zylberberg quote Hamermesh (1993)
Unconditional
LW
= 1.0
Conditional
LW
|
Y
= 0.3
It can also be shown that:
LW
|
Y
= - (1- s)
where s is share of labour in total cost and is the
elasticity of substitution of labour and capital
Given conditional elasticity of 0.3 and known share
of labour of 0.7 in most advanced countries then
elas. of substitution = 1.0
Implications of these estimates
Unit value of unconditional elasticity implies the
wage bill is constant along the demand curve
This suggests a fairly large rate of trade-off
between jobs and wage rises for union bargaining
(in right to manage model)
Large trade-off for government with imposition of
minimum wages in competitive markets
Large difference between conditional and
unconditional elasticity means that the scale effect
is a significantly big component of the adjustment
to any wage cost increase
Unitary elasticity of substitution means that Cobb-
Douglas production function is an acceptable
approximation for analysing aggregate production
Adjustments in the short and long
run with asymmetric costs
Distinguish between no. of workers and hours per
worker in labour demand
Changing employment incurs hiring or firing
costs (which are not equal, thus asymmetric)
Changing hours of work incurs payment of
overtime hours or compensation for short time
Distinguish between workers on permanent
contracts and those on temporary contracts
Temporary workers generally have lower claims
to redundancy pay
Some temporary workers are hired at higher
hourly rates e.g. agency staff
Empirical evidence on adjustment
Over the business cycle hours per worker adjust
more than no. of workers => labour hoarding
This phenomenon reflects higher adjustment
costs for workers than for hours
In the US: hiring costs > firing costs
In France: firing costs > hiring costs
Net impact on jobs/unemployment of rise in
hiring and firing costs is difficult to estimate:
Rise in hiring costs is expected to reduce
demand for workers
Rise in firing cost encourages labour hoarding
Expected profit per worker is reduced by
increase in firing costs so this reduces hiring
Relative adjustment rates for
temporary and permanent staff
Evidence for Spain: Benito and Hernando in Oxford
Bulletin Econ. Stat. 70(3) June 2008
Data 1985-2001 - rapid rise in share of temp. staff
from 9% mid 80s to 20% by 2001
Wage elasticity of demand all workers = - 0.4
Wage elasticity for permanent workers = - 0.0
Wage elasticity for temporary workers = - 2.1
Output elasticity of demand all workers = 0.18
Output elasticity for permanent workers = 0.11
Output elasticity for temporary workers = 0.63
Capital-skill complementarity
Source: Griliches 1969
Divide workers into two types
skilled S and unskilled N
As before capital K has price of R
Wage of skilled worker is Z, unskilled is W
How do relative demands for skilled and unskilled
change if W, Z or R change?
Hypotheses:
NK
>
SK
as the unskilled are
more easily substituted for capital than skilled
and
KN
>
SN
as capital is more easily substitutable
than is skilled labour with unskilled
Empirics of capital-skill complementarity
Griliches shows that the relative demands for
factors can be expressed by the following two
functions:
(1) S/N = f (W/Z, R/Z)
with signs of coefficients +ve and ve
Rise in W/Z causes normal relative price effect
(switch to cheaper substitute)
Rise in R/Z shows bigger switch from K to N than
from K to S, so S/N falls
(2) S/K = f (W/Z, R/Z)
with signs of coefficients ve and +ve
Rise in W/Z causes bigger switch between K and
N than between S and N, so S/K falls
Rise in R/Z gives normal relative price effect
Estimates and implications of
capital-skill complementarity
Griliches estimates using two databases for the
US in 1950s and 1960s confirm the signs of all
four coefficients
What are the implications?
Suppose capital becomes cheaper through time,
for example due to innovation in producer
durables, then R/Z falls
In (2) this will cause production to become more
capital intensive even relative to skilled workers
But in (1) this will cause the demand for skilled
workers to rise relative to the demand for
unskilled workers
See next lecture for more on this topic
Theory - Minimum wages, labour
demand and supply
Competition v. Monopsony
Minimum wages (MW) in perfectly competitive labour markets -
face immediate trade-off of jobs and wages along the demand
curve
Under Monopsony (or Oligopsony) standard model shows can
be some increase in employment instead
(MW creates horizontal supply curve with constant MC)
To maximise employment under Monopsony MW is set at the
competitive market wage
Any higher MW faces trade-off along demand curve
Additional theory relating to labour market frictions
Reactions to rise in wage can be increased labour supply, more
intensive job search and a fall in turnover
Also efficiency wage theory suggests rise in productivity of
those in work leading to higher employment
Employment as minimum wage rises
Employment
On Supply Curve On Demand Curve
Minimum Wage = Minimum Wage = Minimum Wage
Monopsony Wage Competitive Wage
Empirical studies of effects of minimum
wages (quoted in Boeri and van Ours)
UK Stewart Economic Journal (2004) found no adverse
effects on low wage employment
Problem that his study was too early? (UK MW was set at
very low rates in 1999-2001)
US Card and Krueger AER (1994) compared workers in
fast food in New Jersey and Pennsylvania
Found NJ employment in these establishments rose when
MW was increased in this state
Suggests there was monopsony despite rapid turnover in
this sector (or efficiency wage effect reduced turnover)
EU Dolado et al. Economic Policy (1996)
Mainly found negative effects on employment
Effects in Europe worse for young people
Minimum wages in G5 countries (2005)
Note - Germany has no national minimum wage
Approx. 70% of its workers are covered by wages set in
collective bargaining agreements
In all four countries in table MW coverage is 100%
Country Minimum Wage to
Average Production
Workers Wage (%)
Min Wage per
hour in Euros
(PPP exch. rates)
France 52 7.51
Japan 40 4.15
UK 39 6.40
US 31 3.48
UK National Minimum Wage
- rates since 1999
Age group 22+yrs 18-21yrs 16-17yrs
Apr-99 3.60 3.00 none
Jun-00 3.70 3.20 none
Oct-01 4.10 3.50 none
Oct-02 4.20 3.60 none
Oct-03 4.50 3.80 none
Oct-04 4.85 4.10 3.00
Oct-05 5.05 4.25 3.00
Oct-06 5.35 4.45 3.30
Oct-07 5.52 4.60 3.40
Oct-08 5.73 4.77 3.53
Changes in Minimum Wage rates
Age group 22+yrs 18-21yrs 16-17yrs
99-08 59% 59%
99-04 35% 37%
04-08 18% 16% 18%
Adjusting for inflation
Real rise since 1999 in Adult Min Wage is +31%
Next increase is due 1st October 2009
Announcement expected in week of 11th May
Questions for policy now
Having raised MW by > 30% in a decade should
its real rate remain constant now?
Index linking could avoid MW rising above the
competitive wage for low-paid workers
In the present severe recession has the
competitive real wage fallen? (Seems likely)
If so should the real value of MW be reduced to
maximise employment?
Might be sensible to keep nominal rates constant
or even reduce them given near zero inflation
Implies trade-off between equity for the employed
and for the unemployed
Minimum wages and gender equality
Robinson article in Ox. Bull. Ec. & Stats. 2002
Many women work in part-time jobs with low wages
Expect MW to affect women more than men
Can we observe Male-Female wage differential
falling when UK MW began?
Study by Robinson (2003) compares evidence for
period 1995 to 2000 (MW intro. was April 1998)
Robinson uses data from UK Labour Force Survey
which is very large database
Able to track changes at all points in wage
distribution
Problem here (as with Stewart) that this was the
period of very low level of MW
Gender pay gaps in hourly wages
raw and adjusted for characteristics
Men and women adults; characteristics include
region, industry, marital, education,union, temporary
Comparison hourly rates for adults in 3
rd
quarter
Robinson
Table 3
Raw gap Adjusted
gap
1997 -0.308 -0.160
1998 -0.323 -0.177
1999 -0.303 -0.171
2000 -0.277 -0.145
Findings MW and gender wage gap
Gender pay gap rose between 1997 and 1998
but fell again from 1998 to 1999
No net change over two years in raw gap, but a
rise in gap adjusted for characteristics 1997-99
Gender pay gaps (raw and adjusted) then fell
between 1999 and 2000 by 2.6 & 3.2 (log) points
respectively
Simulation results show a rise in MW would
raise F/ M gender pay ratio by 3 points from 73.7
at 3.60 to 76.5 at 5.00
This rise in MW has now happened so = good
news for low paid women?
But what trade off with jobs? Back to labour
demand curve!