MONETARY POLICY REPORT | SPRING 2023 (box 2)
(authors: Michal Franta, Jan Vlček)
A wage-price spiral is most often defined as rapid, simultaneous growth in nominal wages and prices over several successive quarters.[1] It involves a process of interaction between the two variables: growth in prices leads to growth in wages, which in turn causes prices to rise, and so on. Given the generally higher rigidity of wages, the direct transmission between wages and prices tends to be dampened and the situation does not usually turn into a perfectly escalating spiral of ever-increasing wages and prices. Even so, it leads to an undesirable long-lasting rise in inflation.
The risk of a wage-price spiral can be analysed using an index measuring joint growth in wages and prices. Its elevated level in the Czech Republic over the last year implies concurrent high growth in wages and prices (see Chart 1). The level of the index does not, in itself, tell us much about the risk of a wage-price spiral. For this purpose, the conditional 95th percentile of the distribution of the index outlook[2] at the policy-relevant one-year horizon is estimated by quantile regression. The percentile is conditional on the previous observed realisation of the index (its persistence), on macroeconomic variables and also on monetary policy.[3] An index value exceeding the 95th percentile indicates materialisation of the risk of high growth in wages and prices.
Chart 1 – The wage-price spiral index is at a historical high
index in %; contributions in pp
Note: The index is defined as a weighted average of year-on-year growth in the average nominal wage and the CPI index. Historical standard deviations are used as weights.
Growth in the upper end of the index distribution, represented by the 95th percentile, implies an increase in the risk of high index values, i.e. a higher probability of a wage-price spiral. This is significant mainly in a situation where this percentile is above its “neutral” level. The neutral level is an index that is consistent with the 2% target and productivity growth plus two historically observed standard deviations.[4]
Both the value of the index and the upper bound of its distribution reached historical highs last year (see Chart 2). The calculated index of the risk of a wage-price spiral was above the estimate of the conditional 95th percentile in the course of 2022. Last year, this percentile also moved above the neutral level consistent with the 2% inflation target. Both suggest a major risk of a wage-price spiral, and the calculated index exceeding the 95th percentile was indicating the materialisation of this risk.
Chart 2 – The risk of a wage-price spiral rose last year
index and conditional 95th percentile in %
Note: The grey area (2023) is the one-year-ahead forecast for the 95th percentile estimated on the basis of the observed historical data. The light red areas denote the periods in which the ex-post calculated index exceeded the 95th percentile, which was simultaneously above the neutral level.
The breakdown of the contributions of the individual variables to the value of the 95th percentile (see Chart 3) reveals a strong effect of persistence (the lagged index values). The quantile regression results also show that the persistence of the upper percentiles is higher than that of the lower percentiles. This means that ceteris paribus the risk of a wage-price spiral remains elevated for longer than, for example, the mean of index outlook.
Chart 3 – The real economy and monetary policy affect the risk of a wage-price spiral, but the persistence of the index is also a major factor
percentile in %; contributions in pp
Note: The grey area (2023) is the one-year-ahead forecast for the 95th percentile estimated on the basis of the observed historical data.
The decomposition of the upper end of the index distribution also shows the effect of the business cycle, or rather the labour market situation. A positive output gap indicates inflation pressures (an overheating labour market) and leads to an increase in the 95th percentile. Similarly, depreciation of the real exchange rate implies inflation pressures via higher import prices and higher economic activity (exports in particular), hence this percentile increases. By contrast, an economic contraction or appreciation of the real exchange rate fosters a decrease in this percentile. Moreover, the historical relationship between monetary policy and the upper end of the index distribution suggests that monetary policy can significantly reduce the risk of a wage-price spiral. Last year, interest rates fostered a decline in the risk of a wage-price spiral.
The construction of the 95th percentile allows us to use the observed data to create a forecast for 2023 (the grey areas in Chart 2 and Chart 3). The forecast shows an increased risk of a wage-price spiral for the whole of 2023. The forecast for the percentile remains above its neutral level in the first half of the year, before falling slightly below it. The decline in the index is connected with restrictive monetary policy and, to a lesser extent, with appreciation of the real exchange rate. This will outweigh the effects of persistence and the previously positive output gap.
[1] The literature uses various definitions of a wage-price spiral. Examples include simultaneous growth in prices and wages in three out of four consecutive quarters (Wage-price spirals: What is the historical evidence? (external link) Jorge A. Alvarez, John C. Bluedorn, Niels-Jakob H. Hansen, Youyou Huang, Evgenia Pugacheva, Alexandre Sollaci, IMF Working Paper 2022/221) and nominal wage increases exceeding price increases (Are major advanced economies on the verge of a wage-price spiral? (external link) Frederic Boissay, Fiorella De Fiore, Deniz Igan, Albert Pierres-Tejada, Daniel Rees, BIS Bulletin 2022/53). This situation is often associated with major economic shocks and hence also large simultaneous changes in prices and wages. Theoretical models of wage-price spirals can be found in The wage price spiral (external link), Olivier J. Blanchard, Quarterly Journal of Economics, 1986/101(3), and Wage price spirals (external link), Guido Lorenzoni, Iván Werning, 2023.
[2] Quantile regression estimates the linear relationship between the explanatory macroeconomic variables and a percentile of the distribution of the response variable. The 95th conditional percentile shows how high an index value can be expected for given observed macroeconomic variables if we only consider the highest 5% of its possible future values. In other words, this percentile describes the risk of significant and concurrent increases in prices and wages for a given macroeconomic situation over a four-quarter outlook.
[3] The motivation for the choice of macroeconomic variables on which the 95th percentile is conditional is the standard specification of the wage and price versions of the Phillips curve. The selected variables are the output gap and the gap in the real exchange rate against the effective euro area. The output gap is taken from the CNB forecast in the Winter 2023 MPR and the real exchange rate gap is calculated using the HP filter. Monetary policy is proxied by the 3M PRIBOR nominal interest rate.
[4] The distance of two historically observed standard deviations is due to the fact that 95% of the observed values for a normally distributed variable (the neutral level of prices and wages) lie within a distance of less than two standard deviations.