William Davidson Institute Working Paper 487
Annex 3 Real Wage Rigidity Index Calculations (Nenovsky and Koleva, 2001)
The dilemma “unemployment - devaluation” underlies the issue of monetary
authority credibility. This basic theoretical assumption serves as a basis for modelling a
speculative attack against the currency board (Rivera-Batiz and Sy, 2000). Since real
wage rigidity is a major condition for absorbing possible shocks in the economy under a
static central bank, construction of the so-called real wage rigidity index is a key element
in the labour market analysis.
The basic structural model for estimating rigidity is proposed by Layard and al.
(1991) and is based on a system of equations describing wages and prices:
w - p = - c (u - hu-1) + zw (1)
zw = es + ew, (2)
where w, p, u are the logarithms of nominal wages, the price index and
unemployment respectively, u-1 is logarithm of unemployment with one lag, c and h are
the parameters for estimation, and zw reflects shocks on nominal wages (the sum of es -
technological shock, and ew - labour-supply shock). h measures unemployment inertia,
its hysteresis effect, and c shows the elasticity of real wages to changes in
unemployment. The rigidity index is calculated by replacing the estimated c and h in the
formula below:
11
RWR = (c (1 -h))-1 = (3)
c(1 - h)
Some of the weaknesses of the Layard model (a structural model) are overcome
by Vinals and Jimeno (1998) who generate the values of the parameters in question by
estimating two versions of BVAR model, relating real wage dynamics both with
unemployment level and unemployment dynamics rate. Their models show both mutual
responses to shocks and the decomposition of real wage response and unemployment
response, the average lag of the response to unemployment being ʌ .
1-h
39