Labour market flexibility research seminar

1
  HOW COSTLY IS DOWNWARD NOMINAL RIGIDITY IN THE
UNITED KINGDOM? Jennifer C Smith University of
Warwickjennifer.smith_at_warwick.ac.uk Labour
Market Flexibility ResearchSeminar, London, 15
December 2004
2
Overview of presentation

• Introduction and motivation
• How extensive is nominal wage rigidity?
• What are the costs of nominal wage rigidity?

3
Introduction and motivation

• What is downward nominal wage rigidity?
• Why is it a problem?
• real wages too high, unemployment
• Why might it arise?
• fairness, legal/institutional reasons, money
  illusion
• Data British Household Panel Survey (BHPS)
  1991/2-2002/3
• stayers aged 16-65, representative sample
• wage data

4
Annual pay growth
Source BHPS 1991/2-2002/3
5
Annual wage growth
Source BHPS 1991/2-2002/3
6
Nominal wage rigidity in the UK

• On average over 1991-2002, 8.3 of employees who
  remained in the same job from one year to the
  next report no change in their weekly pay from
  one year to the next.
• Rigidity in hourly pay is 3.9.
• Rigidity among hourly-paid stayers is 15.3, on
  average over 2000-2002.
• Similar proportions of workers are missing from
  the lower tails of the distributions asymmetry
  in the pay growth distributions roughly matches
  the sizes of the zero spikes.

7
Rigidity and inflation

• Downward nominal rigidity should rise as
  inflation falls.

Nominal pay growth
8
Rigidity and annual inflation
Source BHPS 1991/2-2002/3
9
Rigidity and inflation

• P(rigidrt) a b(medianrt) ert (1)
• 11 standard regions, 1992-2002.
• Expect blt0.
• Find b-0.514 for weekly pay.
• Find b-2.416 for basic hourly wage rate.

10
Rigidity and inflation

• The rigidity-inflation relationship might be
  non-linear.

11

• The median ranges from roughly 3 to 8.

Source BHPS 1991/2-2002/3
12
Rigidity and inflation
13
The unemployment cost

• Under downward nominal wage rigidity, the real
  wage will be (permanently) too high. This will
  lead to (permanently) higher unemployment.
• The Phillips curve relationship between real wage
  growth and unemployment will be non-linear there
  will be a variable trade-off between real wage
  growth and unemployment.
• This trade-off will be most negative when
  inflation is high.
• As inflation falls, downward nominal rigidity
  holds increasingly more workers wages too
  high.
• Because of this, at low inflation, unemployment
  would need to rise by more to generate a given
  fall in real wages.

14
The unemployment cost
Dw
Steep Phillips curve at high inflation
D(DwL)
Flat Phillips curve at low inflation
D(DwH)
U
DU
15
Non-linear Phillips curve

• Dwrt a bUrt c DptUrt ert (2)
• 11 standard regions, 1992-2002
• Expect clt0 at higher inflation, the real wage
  growth-unemployment trade-off becomes more
  steeply negative.
• Without allowing for non-linearity, results
  suggest that a doubling of the unemployment rate
  would reduce pay growth by around 1 percentage
  point.
• In general, results do not indicate any
  significant non-linearity. Indeed, positive
  interaction coefficients are found in some
  specifications.
• The maximum effect found is not large. It
  indicates that when regional inflation is 4, a
  doubling of the unemployment rate would reduce
  real hourly basic wage rate growth by 1
  percentage point. If regional inflation fell to
  0, a doubling of unemployment would result in
  less downward wage pressure real wage rate
  growth would fall by 0.7.

16
Concluding remarks

• Data are consistent with the existence of
  significant downward nominal wage rigidity in the
  UK.
• More workers wages appear to be held rigid, the
  lower is inflation.
• But this does not seem to have a substantial
  macroeconomic cost in the form of higher
  unemployment.