MONETARY POLICY REPORT | AUTUMN 2023 (box 3)
(author: Adam Ruschka)
The CNB introduced the Labour Utilisation Composite Index – LUCI – in 2017 and extended it two years later.[1] In recent years, the Czech economy has been hit by a series of shocks which have greatly complicated macroeconomic analysis and forecasting, and some of the approaches used in the past have become obsolete. As the labour market plays a major role in domestic inflation pressures, it was also useful to revise the method for calculating its composite indicator. This box briefly describes the impacts of the revised method on the paths of the LUCI and its components.
The first motive for changing the method was that the volatility of most of the time series had increased substantially over the previous few years. The sensitivity of the computation to sharp movements therefore needed to be revised to give more intuitive identification of trends and especially cycles in the recent past. The resulting LUCI thus better matches the expert view of observed developments on the labour market. The second major motive for changing the method was the effect of the 2021 census in the Czech Republic on some labour market time series. The modified Labour Force Survey weighting scheme has given rise to significant changes in those series. The changes, however, are purely statistical, do not reflect a structural change in the labour market[2] and hence carry no fundamental information. For this reason, the series were replaced by alternatives not subject to these issues.[3]
The new LUCI computation method is nonetheless essentially the same as the previous one. The trend and cyclical components of a total of 28 time series of labour market variables are simultaneously separated. The composite information from all the cyclical components – de facto their first common component – then offers a comprehensive view of the cyclical position of the labour market, expressed at the aggregate level by the LUCI.
The changes described above have changed the entire profile of the LUCI (see Chart 1). The qualitative story, though, remains similar (except for the results in the recent past). Even with the new method, we identify labour market slack until 2006 and also after the great economic and financial crisis. The interpretation of the strong overheating of the labour market in 2017–2020 is also unchanged. By contrast, with the onset of the pandemic and the other shocks we see only a slight slackening of the labour market under the new method. The labour market therefore does not converge markedly towards the steady state. This interpretation is more consistent both with the time series and with other analytical sub-indicators of labour market tightness, such as the Beveridge curve. Unlike in the previous forecast, there is no repeated tightening of the labour market in the outlook either.
Chart 1 – The LUCI revision pertains mostly to the recent past, when the economy was hit by a series of shocks
index; y-axis shows standard deviations
As can be seen from the lower figure in Chart 1, if we were to apply the new LUCI computation retrospectively to the data that were available when the previous forecast was drawn up (Monetary Policy Report – Summer 2023), we would assess the current labour market situation almost identically using the revised method as we do now. The overall labour market situation has changed little since then, as strong employment growth, a still low unemployment rate and continued willingness of employers to hire new staff are outweighing the marked fall in real wages. This is keeping the labour market tight.
The new computation better reflects the macroeconomic evolution of the labour market, including from the perspective of its components (see Chart 2). With hindsight, we can say that the previous method attributed all the large movements of the LUCI (in the past and in the outlook) to demand for work to too great an extent, primarily to the detriment of employment and wages and costs. In the new LUCI, the contributions of its components better reflect the observed paths of labour market variables. The new LUCI therefore more faithfully describes the current labour market situation and continues to offer an important overall assessment and partial information for forecasting purposes.
Chart 2 – The paths of the contributions of the various categories are now more consistent with each other and with the overall labour market story
index; y-axis shows standard deviations
[1] The construction and reasons for the creation of the LUCI were described in Box 1 in Inflation Report IV/2017. The extension of the LUCI was presented in Box 2 in Inflation Report IV/2019. A more detailed description of the LUCI will be published as a CNB working paper in the coming months.
[2] For more information, see the official opinion of the CZSO (in Czech only).
[3] The new series include unemployment series of various lengths published also by the MLSA, moreover before the CZSO (LFS statistics).