Page 227 - ePaper
P. 227
Chapter 4: Restoring Convergence between Member States in the EU and EMU
1
Box 3: Estimating the pass-through of changes in the nominal unit labour cost ( )
The starting point of the empirical analysis is the assumption that in the long run output prices are in line with the nominal
unit labour cost ( ). However, in markets characterised by imperfect competition and imperfect information, the output prices
2
are not automatically fully aligned with nominal unit labour costs due to, inter alia, menu costs ( ), administered prices ( ),
4
3
or backward-looking ‘rule of thumb’ price setting ( ). Moreover, the state of the business cycle (i.e. fluctuations in effective
5
demand compared to potential output) may put demand-push inflationary pressures on prices ( ). Within such an economy,
6
prices may over- or undershoot their equilibrium values in the short- to medium-run so that output prices will only converge
gradually towards the nominal unit labour cost ( ).
7
Specifying these adjustment channels and regressing quarterly changes in output prices on a set of explanatory variables (includ -
ing changes in the nominal unit labour cost, past price changes and past divergence between output price and nominal unit
labour cost) ( ), yields estimates that are in line with the hypothesis that output prices adjust with a lag to changes in nominal
8
unit labour costs. Subsequently, the point estimates can be used to project the path along which prices converge to the new
equilibrium in response to changes in nominal unit labour cost (keeping all other factors constant) — as shown in Chart 27.
More particularly, Chart 27 shows the impact of an (exogenous shock in the) nominal unit labour cost after two quarters and then
one, three, five and ten years — for the euro area Member States for which the data are available (for other Member States the
dataset needed for the estimation is not available). It would be beyond the scope of this chapter to take into account feedbacks
of changes in output prices and nominal unit labour cost on the rest of the economy, such as nominal interest rates, exchange
rates, etc. Moreover, it should also be recognised that to the extent that the effects of cuts and increases in nominal unit labour
cost are not symmetric in price adjustment, the simulated results in Chart 27 may overestimate the adjustment speed of prices.
( ) More technical details are to be found in Annex 1.
1
( ) More specifically, it is assumed that unit labour cost and price levels are co-integrated.
2
( ) See, for instance, Mankiw (1984).
3
( ) Which are in the short run not necessarily disciplined by market forces.
4
( ) See, for instance, Calvo (1983).
5
6
( ) As well as inflationary pressures on nominal unit labour cost via its impact on nominal wages and productivity — requiring the use of instrumental
variables estimation techniques.
7
( ) Note that the analysis in this note is limited to the Member States of the euro area (for which the data are available). This section does not analyse the
price level at the level of the euro area as a whole. At that level, the price level is aligned (in the long run) to developments in the supply of money and
demand for real money balances.
( ) Using harmonised, seasonally and working-time adjusted, quarterly Eurostat data of the Member States for which the data are available, covering the
8
1995a1–2013q2 period, and applying instrumental variables estimation techniques.
Chart 27: Adjustment path of output prices after a permanent
cut in nominal unit labour cost — total economy
0
-20
-40
-60
-80
%
-100
-120
-140 Year 1/2 Year 1 Year 3 Year 5 Year 10
-160
DE PT SI CY ES AT FR BE NL LU FI IT EE SK
Source: DG EMPL estimations using Eurostat data.
Notes: nominal unit labour cost is compensation per employee adjusted for productivity. No data
available for IE and EL.
225
1
Box 3: Estimating the pass-through of changes in the nominal unit labour cost ( )
The starting point of the empirical analysis is the assumption that in the long run output prices are in line with the nominal
unit labour cost ( ). However, in markets characterised by imperfect competition and imperfect information, the output prices
2
are not automatically fully aligned with nominal unit labour costs due to, inter alia, menu costs ( ), administered prices ( ),
4
3
or backward-looking ‘rule of thumb’ price setting ( ). Moreover, the state of the business cycle (i.e. fluctuations in effective
5
demand compared to potential output) may put demand-push inflationary pressures on prices ( ). Within such an economy,
6
prices may over- or undershoot their equilibrium values in the short- to medium-run so that output prices will only converge
gradually towards the nominal unit labour cost ( ).
7
Specifying these adjustment channels and regressing quarterly changes in output prices on a set of explanatory variables (includ -
ing changes in the nominal unit labour cost, past price changes and past divergence between output price and nominal unit
labour cost) ( ), yields estimates that are in line with the hypothesis that output prices adjust with a lag to changes in nominal
8
unit labour costs. Subsequently, the point estimates can be used to project the path along which prices converge to the new
equilibrium in response to changes in nominal unit labour cost (keeping all other factors constant) — as shown in Chart 27.
More particularly, Chart 27 shows the impact of an (exogenous shock in the) nominal unit labour cost after two quarters and then
one, three, five and ten years — for the euro area Member States for which the data are available (for other Member States the
dataset needed for the estimation is not available). It would be beyond the scope of this chapter to take into account feedbacks
of changes in output prices and nominal unit labour cost on the rest of the economy, such as nominal interest rates, exchange
rates, etc. Moreover, it should also be recognised that to the extent that the effects of cuts and increases in nominal unit labour
cost are not symmetric in price adjustment, the simulated results in Chart 27 may overestimate the adjustment speed of prices.
( ) More technical details are to be found in Annex 1.
1
( ) More specifically, it is assumed that unit labour cost and price levels are co-integrated.
2
( ) See, for instance, Mankiw (1984).
3
( ) Which are in the short run not necessarily disciplined by market forces.
4
( ) See, for instance, Calvo (1983).
5
6
( ) As well as inflationary pressures on nominal unit labour cost via its impact on nominal wages and productivity — requiring the use of instrumental
variables estimation techniques.
7
( ) Note that the analysis in this note is limited to the Member States of the euro area (for which the data are available). This section does not analyse the
price level at the level of the euro area as a whole. At that level, the price level is aligned (in the long run) to developments in the supply of money and
demand for real money balances.
( ) Using harmonised, seasonally and working-time adjusted, quarterly Eurostat data of the Member States for which the data are available, covering the
8
1995a1–2013q2 period, and applying instrumental variables estimation techniques.
Chart 27: Adjustment path of output prices after a permanent
cut in nominal unit labour cost — total economy
0
-20
-40
-60
-80
%
-100
-120
-140 Year 1/2 Year 1 Year 3 Year 5 Year 10
-160
DE PT SI CY ES AT FR BE NL LU FI IT EE SK
Source: DG EMPL estimations using Eurostat data.
Notes: nominal unit labour cost is compensation per employee adjusted for productivity. No data
available for IE and EL.
225
