Economic Thought and Practice, , 2026.
Original scientific paper
https://doi.org/10.17818/EMIP/2026/33
FOUR WAYS OF MEASURING THE PREDICTIVE POWER OF THE EMPLOYMENT EXPECTATIONS INDICATOR IN FORECASTING LABOUR MARKET CHANGES – WHICH ONE IS THE BEST?
Mirjana Čižmešija
orcid.org/0000-0002-9873-7291
; University of Zagreb, Faculty of Economics and Business, Croatia
*
Marita Novoselec
orcid.org/0009-0005-6950-4324
; Croatian Bureau of Statistics, Croatia
* Corresponding author.
Abstract
Monitoring and forecasting labour market developments are essential for macroeconomic policymaking and strategic planning. The crises of 2008 and 2019 highlighted the importance of expectations held by managers and consumers, alongside objective economic indicators, with the Employment Expectations Indicator (EEI) playing a particularly significant role. As the European Commission’s standard approach, which compares changes in the EEI with annual employment growth, does not yield satisfactory results for Croatia, this paper proposes three alternative measures of EEI change. For each model, confusion matrices were constructed and relative accuracy measures were calculated to assess the ability to predict the direction of employment change across different forecasting horizons (h = 0, 1, 2, 3). The results indicate that the most accurate specification of the EEI aligns with its theoretical interpretation: values above 100 signal managerial optimism, while values below 100 indicate pessimism. The analysis uses monthly data covering the period from January 2019 to August 2025.
Keywords
labour market; Employment Expectations Index; Business and Consumer Confidence Surveys; confusion matrix; measures of accuracy of forecasting the direction of change
Hrčak ID:
348640
URI
Publication date:
1.7.2026.
Visits: 0 *