Transitions on the labour market

Fewer transitions to work in 2021

Work & training
Fewer transitions to work in 2021


The corona crisis has a serious impact on working people, especially in a number of sectors. This is what emerged from the transitions on the labour market of Statbel, the Belgian statistics office. If we compare the labour market status of the first quarter of 2020 with the labour market status of the first quarter of 2021, 9 out of 10 employed persons (90.8%) are still at work. A high percentage, although it is lower than the previous quarters. Especially people who worked in the ‘horeca’ sector have not yet returned to work in the first quarter of 2021.

For the unemployed we see that people are staying unemployed longer: 44.9% of the unemployed in the first quarter of 2020 are also unemployed in the first quarter of 2021. The regional differences are striking here: in the Flemish region, 38.4% of the unemployed persons are still unemployed after one year. In the Walloon and Brussels-Capital Regions, the figures are 48.6 and 48.7% respectively.


2020 will always be the year of the corona crisis. We can also see this in the longitudinal statistics of the Labour Force Survey: more unemployed people have entered the statistics in the past year because the unemployed are less likely to find work and employed persons are slightly more likely to lose their jobs. Statbel publishes today the transitions on the labour market between the first quarter of 2020 (i.e. just before the crisis breaks out) and the first quarter of 2021.

The comparison is complicated by a change in the definition of employed persons in the new European framework regulation: persons who have been temporarily unemployed for more than three months (‘long-term temporarily unemployed’) are now counted among the unemployed or inactive, and no longer among the employed - depending on the answers to the questions on job search and availability. In the first quarter of 2021, it is estimated that 80,000 long-term temporarily unemployed people are counted among the inactive (and to a lesser extent the unemployed). More details on the changes can be found in this note: (Changes to the Labour Force Survey (LFS) in 2021)

Methodological information

The figures presented here are the results of the Labour Force Survey (LFS), a survey harmonised at European level. The definitions regarding employment and unemployment that are used are those of the International Labour Office (ILO) to allow international comparison. We distinguish three ILO statuses on the labour market: employed, unemployed and inactive. The definitions applied are available here:

Please note that temporarily unemployed persons are temporarily absent from work and are counted as employed.

The Labour Force Survey is a continuous survey, which means that the sample is spread evenly over the 52 weeks of the year. The selected respondents answer a questionnaire mainly related to their activity in the course of a given reference week. The respondents participate four times: they participate for 2 consecutive quarters, then don’t for 2 quarters, and then participate again for 2 quarters. This way, we can observe what the labour market status of a given respondent is in a given quarter, and a quarter and/or a year later: e.g. is someone who is unemployed still unemployed in the next quarter and/or year?

So, if one speaks of a particular status in a particular quarter, it is by definition the status in the reference week. If one indicates to work in the reference week of quarter Q and in the reference week of quarter Q+1, they are counted twice as employed. There are, of course, a number of cases that were unemployed in the meantime, for example, but this is beyond the scope of our data.

The quarterly transitions are the sums of weighted observations of respondents who participated in the successive quarters (e.g. 2019Q4-2020Q1, 2020Q1-2020Q2).

The quarter-specific annual transitions are the sums of weighted observations of respondents participating in the same quarter of two consecutive years (e.g. 2019Q1-2020Q1).

The annual transitions are the means of four quarter-specific annual transitions for two successive years (e.g. 2019-2020).

Respondents who did not participate in one of two waves (= interviews) cannot be taken into account in this analysis. Respondents in the longitudinal sample are in both quarters at least 15 years old and at most 74 years old.

The longitudinal sample is calibrated to the estimated distributions of ILO labour market status per age, gender, region, level of education and nationality in the start and end quarters.

The published figures are based on the Labour Force Survey. They are no exact figures but approximations based on the extrapolation of a random sample from the Belgian population. This must be taken into account when interpreting the results. When the unweighted number of people is lower than 30, data should be interpreted with caution.