Greater harmonisationThe new approach to the international labour force survey is the result of a European regulation, which came into force on 1 January 2021. The aim of this regulation is to facilitate greater harmonisation between social statistics in the countries of the European Union, including the unemployment rate. Working with the national statistical offices of the EU member states, CBS has been intensively involved in developing this new regulation.
Registers‘This process included extensive discussions with Eurostat and the International Labour Organization in Geneva about the content of the questionnaire and its implementation,’ says Eric Schulte Nordholt. In his role as international coordinator of social statistics at CBS, Schulte Nordholt has been involved in the development and implementation of the renewed Labour Force Survey. One development that CBS insisted on was using variables from existing registers to reduce the length of the questionnaire. ‘One of the things we brought to the table during those European consultations was our knowledge about how to do this effectively. We also shared our experience of reweighting data when response rates lag behind in certain subgroups.’
Key indicatorOf all the personal and household surveys conducted by CBS, the Labour Force Survey is one of the most important, Schulte Nordholt points out. ‘The unemployment rate is published monthly and is always eagerly awaited, both at national level and in the European context. In combination with underlying data on factors such as the composition of the labour force, it represents a key indicator for policymakers. Labour market statistics are crucial input for the coordination of European economic policy.’
Core questionnaireOver the years, the way in which member states conduct their labour force surveys has become increasingly harmonised. The regulation that came into effect on 1 January of this year formalises several aspects of how the labour force survey should be conducted. ‘Previously, states had more freedom in the wording of questions and the order in which they were put to respondents. Now a core questionnaire has been established, to which all member states are required to adhere. This makes the results of all those national surveys more comparable.’
Higher response rateCBS took advantage of the European harmonisation process to further optimise its own Labour Force Survey. ‘For example, we switched from a household survey to a personal survey. The majority of the questions put to the respondents are now about themselves and far fewer concern other members of their household,’ says Martijn Souren, process manager of the Labour Force Survey at CBS. ‘This makes the questionnaire shorter. The survey starts by offering respondents the opportunity to complete the questionnaire online. We have seen that online respondents now take about 11 minutes less to complete the survey. The shortened duration and the new “look and feel” of the questionnaire are having a positive impact on the survey’s online response rate, which has risen from 25 percent to 38 percent. This increase boosts the efficiency of CBS’s data collection,’ Souren says.
Postponement due to coronavirusCBS’s publication of the results from the redesigned survey was delayed by six months due to the coronavirus pandemic. ‘Doorstep surveying is important in order to achieve sufficient response rates in all subgroups, but this simply wasn’t possible in the last six months.’ The organisation was unable to draw on past experience when it came to assessing the impact that a lack of doorstep interviews would have on the new-style survey. This experience was available for the previous version. ‘For this reason, CBS decided to extend the period in which it continued to use the old design in parallel with the new. This will ensure continuity of outcomes,’ Souren observes.
Mix of methodsThe overall response rate in the Netherlands held up well despite the pandemic because, unlike some EU counterparts whose statistical research is more reliant on face-to-face interaction, CBS uses a broader mix of data collection methods. Souren outlines the Dutch approach: ‘We always begin by asking respondents to complete the survey themselves online. Only then do our interviewers call or schedule a home visit to conduct the survey. The fact that we do not rely entirely on face-to-face interviews has proved to be a great advantage. Even during the Covid crisis, this enabled us to conduct a large number of personal and household surveys.’
Deviations in the figuresThe figures on the Dutch labour force in 2021 as presented by Eurostat will deviate from the official CBS results. Whereas CBS will continue to publish the results from the old-style survey well into 2021, Eurostat will be publishing data on the basis of the new design as from 15 July 2021. The figures presented by Eurostat have been given ‘preliminary’ status because they do not take into account the absence of home surveys. In addition, the overall size of the response from the new design is still much smaller than it was for the old design. This is because, unlike statistical offices in other member states, CBS did not switch to Eurostat’s new survey design in one go.
‘For the comparability of our national figures over time, we decided it was not desirable to present respondents with a different questionnaire during this cycle,’ Souren explains. ‘In our case it will therefore be another year before Dutch outcomes will be calculated on the basis of a full dataset from the new harmonised questionnaire design. Until then, the Eurostat results will deviate from the official CBS figures, which are still based on the old design.’
Preliminary results from the first dataset of the new design suggest that in response to the new mandatory mode of questioning, young people in particular are more likely to report being employed. In addition, the new questionnaire asks what type of contract they have, a question that is more likely to reveal the temporary nature of a contract than its equivalent in the previous questionnaire, which asked respondents if they were in permanent employment. Meanwhile people without a job are more likely to be categorised as unemployed in the new design; they are more likely to report that they are looking for work and able to start work in the near future.