Monthly figures on the labour market

Statbel DataLab: new statistics, methods and data sources beta version

Monthly figures on the labour market– August 2020

DataLab
Monthly figures on the labour market– August 2020

Sharp decrease in the employment rate in August

As Belgian statistical office, Statbel wants to map the impact of the Covid-19 crisis on the labour market by providing fast, indicative figures. After the months of April and May in which employment fell, we recorded an increase in the number of people employed in June and July. In the meantime, a turnaround seems to have started, because preliminary results from the Labour Force Survey for the month of August this year show another decline in the number of people employed in August. Above all, we see a sharp decline in the number of employees, while the number of self-employed remains fairly stable.

The employment rate of people aged 20-64 in August is estimated at 68.8 %, the lowest level since the beginning of the health crisis. For people aged 55 and over, the employment rate is still increasing, but in the other age groups, it is falling sharply compared to the level of previous months. It is mainly in Flanders that the employment rate is falling. In Wallonia, the decline in the percentage of people employed in the population aged 20-64 is limited and in Brussels, in August, there was even a slight increase compared to the previous month. The employment rate of people aged 20-64 in August is estimated at 61.6 % in Brussels, 72.3 % in Flanders and 65.0 % in Wallonia.

ILO unemployment rate stabilises at 6.7 % in August

While the ILO unemployment rate[1] of people aged 15-64 still showed a downward trend in March and April, it started to show an upward trend since May. In July the ILO unemployment rate suddenly rose sharply from 5.4 % to 6.7 %. In August, the figure stabilises at 6.7 %. In Brussels, after a sharp rise in July, we are once again seeing a fall in the unemployment rate, while in Flanders and Wallonia we are seeing a slight increase in the unemployment rate. The unemployment rate of people aged 15-64 amounts to 13.4 % in Brussels, 5.0 % in Flanders and 7.8 % in Wallonia.

Significant increase in the number of inactive people

The sharp fall in employment, combined with a stabilisation of unemployment in August, is accompanied by a substantial increase in the number of inactive people. In August, 4,382,000 persons were inactive compared to 4,237,000 in July.

However, this still remains well below the levels of April and May this year, when the number of inactive people was even higher. But during that period, the number of ILO unemployed was much lower than in August. This had to do with the fact that at the beginning of the Covid-19 crisis many people without a job did not actively look for a job, or were not available within two weeks to start working, because they had to take care of the children, for example. According to the ILO definitions, this group was included among the inactive. Since the summer, however, it seems that part of that group is both actively looking for a job again and is available again to start working within two weeks. Once these two conditions have been met, they will no longer be counted among the inactive, but among the ILO unemployed.

If we compare with August last year, we now see more inactivity among both men and women. Especially in Flanders and among the group of medium-skilled people (upper secondary diploma), the increase is the largest compared to last year. Also, the group of inactive people who are available but not actively looking for work remains fairly large and the level is also significantly higher than in August last year.

Holidays the main reason of absence

Even more than in March, when we were in lockdown in only two out of four reference weeks, the crisis had a significant impact on hours worked in April and May. In April 2020, more than 44 % of employees had worked less than usual or had not worked at all during the reference week for which they were interviewed. This was 2.1 million employed people. In May we already saw a first positive effect of the implemented lockdown easing measures and the number dropped to about 1.6 million. In June, this number fell even further to 1.1 million employed persons. In July and August, the number of employed persons indicating that they have not worked or have worked less than usual rises again to 1.6 and 1.5 million respectively, but the holiday period obviously plays an important role in this.

On the one hand, there is the group of people who indicate that they did not work at all during the reference week. In August 2020, they were 1.1 million, roughly the same number as in August last year. The main reason of absence is holidays. This is the case for 813,000 persons with a job. This number is 115,000 units below the number in August last year. The second main reason for not having worked at all during the reference week is illness, accident or temporary disability. In August, there were 165,000 persons in this group. For 59,000 people, temporary unemployment is the main reason for absence during the whole reference week. This number is just under 9,000 units lower than in July.

On the other hand, there is the group of people who indicate that they worked less during the reference week. In August 2020, there were 355,000 persons in this group, i.e. almost 100,000 fewer than in August 2019. At the height of the crisis in April 2020, there were 760,000 persons in this group. Holidays are the main reason for working less than usual (172,000 persons). This number is about half as low as in August last year. Temporary unemployment comes next, with 51,000 persons. This figure is 19,000 units lower than in July, but still significantly higher than before the crisis.

The average actual working time per week is about the same as in August last year

Because of holidays, the average actual working time per week in the holiday months is usually below that of the other months. In July 2020, persons employed worked on average 26 hours per week in their main job. In August, this average working time is 26.7 hours, which is even slightly higher than the average working time in August last year, which was 26.4 hours.


[1] The ILO unemployed are all the people who do not have a job, are actively seeking work and are available to start working within two weeks.

Methodologische info

De hier gepresenteerde cijfers zijn resultaten van de Enquête naar de Arbeidskrachten (EAK), dit is een enquête die geharmoniseerd is op Europees niveau. De definities over werkgelegenheid en werkloosheid die worden gehanteerd zijn die van het Internationaal Arbeidsbureau (IAB), waardoor een vergelijkbaarheid van de resultaten op internationaal vlak wordt gewaarborgd. De gehanteerde definities bevinden zich hier: https://statbel.fgov.be/nl/themas/werk-opleiding/arbeidsmarkt/werkgeleg….

Merk op dat tijdelijk werklozen tijdelijk afwezig zijn van hun werk en tot de werkenden worden gerekend.

De enquête naar de arbeidskrachten is een continue enquête, wat wil zeggen dat de steekproef gelijk verdeeld is over de 52 weken van het jaar. De geselecteerde respondenten beantwoorden een vragenlijst die hoofdzakelijk betrekking heeft op hun activiteit in de loop van een gegeven referentieweek. De resultaten op maandbasis kunnen beschouwd worden als het gemiddelde van de maand.

Voor de opmaak van de maandelijkse indicatoren werden de antwoorden van de respondenten voor een bepaalde “maand” (d.i. een set van 4 of 5 opeenvolgende volledige kalender- of referentieweken, bv. maart 2020 bestaat uit referentieweken 10-13) gekalibreerd op basis van volgend model: Prov*Sex*Agecat + Regio*Educat3c waarbij Prov= provincie, Sex= geslacht, Agecat= leeftijdsgroep per 5 jaar, Regio= 3 gewesten, Educat3c= onderwijsniveau (laag, midden, hoog). Hiervoor worden twee frames voor berekening van kalibratietotalen (benchmarks) gebruikt:

Uit het Rijksregister: populatiecijfers op de eerste dag van het kwartaal, volgens de kruising Prov*Sex*Agecat.

Uit EAK: geschatte populatiecijfers volgens de kruising Regio*Educat3c, m.a.w. een geschatte verdeling van het onderwijsniveau per regio; hiervoor baseren we ons op gekalibreerde EAK-steekproeven voor de 4 meest recent beschikbare kwartalen (bv. voor januari-maart 2020 zijn dat de kwartalen 2018T4 t.e.m. 2019T3).

De gepresenteerde cijfers zijn geen effectieve populatiecijfers, maar benaderingen die gebaseerd zijn op de extrapolatie van een toevalssteekproef uit de Belgische bevolking. Bij de interpretatie van de cijfers dient hiermee rekening te worden gehouden. De resultaten die hier gepresenteerd worden zijn indicatieve resultaten op maandbasis en zijn onderhevig aan grotere toevalsschommelingen dan de resultaten op kwartaal- en op jaarbasis omdat ze gebaseerd zijn op een twaalfde van de steekproef op jaarbasis. Voor maart 2020 gaat het om ruim 8600 respondenten. Kleine aantallen en kleine verschuivingen doorheen de tijd dienen dan ook met de nodige omzichtigheid geïnterpreteerd te worden omdat ze gebaseerd zijn op de antwoorden van een beperkt aantal respondenten.

De in dit bestand opgenomen maandelijkse statistieken zijn experimentele statistieken en worden geproduceerd met het specifiek doel tot monitoring van de Coronacrisis. Het is belangrijk voor ogen te houden dat het steeds gaat om voorlopige cijfers, die geproduceerd werden op een eerste versie van de gegevens en waarbij de snelheid primeert op de volledigheid en kwaliteit van de binnengekomen gegevens. De maandelijkse cijfers zullen dit voorlopig karakter behouden tot na de publicatie van de officiële kwartaalresultaten.

Meer info over de enquête naar de arbeidskrachten is te vinden op:

https://statbel.fgov.be/nl/themas/werk-opleiding/arbeidsmarkt/werkgeleg…

https://statbel.fgov.be/sites/default/files/Over_Statbel_FR/Analyse_eak…