Bank of Israel Research Department analysis: The unemployment rate and its definition during the corona period
- The figure regarding the unemployment rate that the Central Bureau of Statistics is expected to publish will be significantly lower than the data on job seekers published by the Employment Service.
- The high number of employed persons who have been put on unpaid leave, and the activity limitations on the population and on the business sector, are creating a unique conceptual difficulty in defining and analyzing unemployment data at this time.
- The difficulty in definition is reflected in the differences between the IMF’s forecast and the Bank of Israel’s forecast regarding the expected unemployment rate in Israel in the second quarter.
Since the start of the corona crisis, the Israeli public has been exposed to an almost daily increase in the number of job seekers registering with the Employment Service. The number of job seekers (which are sometimes labeled “unemployed”) has crossed one million people, and accounted for more than 27 percent of the labor force on April 21, 2020. Most of the new job seekers registered due to being placed on unpaid leave (see Figure 1). Alongside this figure, the Central Bureau of Statistics publishes an unemployment rate once a month that is based on the Labor Force Survey it conducts among a few thousand sample working-age individuals. This document explains the differences in definition between the figures, and clarifies the effect of the special situation created as a result of the corona virus on how the Central Bureau of Statistics measures unemployment in the economy. In particular, the analysis assesses whether all the individuals that were placed on unpaid leave will be counted as unemployed.
An unemployed person is defined as someone who is not employed, is interested in working, is available to work, and is actively searching for work. In contrast, a person who is absent from work but receives payment (including from the National Insurance Institute), and is expected to return to the same employer, is considered an employed person who is temporarily absent form his job. Since most of those being placed on unpaid leave meet this definition, our assessment is that the data for March (which will be published at the end of April) will not reflect a significant change in the unemployment rate, and the data will show an increase in employed persons who are temporarily absent. Regarding April, and the second quarter as a whole, the picture is less clear, and also depends somewhat on how the surveyed individuals respond. The classification of a surveyed individual on unpaid leave depends on the question of whether he expects to return to work for the same employer. If the answer is yes, he will be classified as a temporarily absent employed person. If the surveyed individual’s assessment is that he will not return to work for the same employer, he will be classified as not employed. (It will be explained below whether he will also be considered unemployed.) It is important to note that this is not just a case of semantics. The economic behavior of a person who thinks that he is on temporary leave (even if forced) differs from that of a person who believes that he will not be able to return to his job.
Figure 2 provides a schematic of the structure of the labor market. The comments in red in each segment in the chart show the expected change in that segment in view of the crisis. For instance, the Central Bureau of Statistics will classify those not employed into two categories: unemployed, and those who are not participating in the labor force. The classification depends on the question of whether the surveyed individual (who is not employed) has actively looked for a job. This factor is expected to lead to a large number of newly non-employed people being classified as those who are not participating in the labor force (and are therefore not classified as unemployed). As a result, there will be a decline in the labor force participation rate, since someone who answers the labor force survey and states that he has not actively searched for work due to the corona pandemic (for instance because he believes that it is impossible to look for work due to movement restrictions, or that there is no point in trying to search) will be considered as not having searched and is therefore not defined as unemployed, but rather as someone who is not in the labor force. This phenomenon is temporary, and is relevant only as long as workplaces remain shut. As workplaces reopen, as began happening on April 19, 2020, answers of the type mentioned will become less valid. As such, in analyzing the data and forecasts relating to the period from March until the full reopening of businesses, it is important to also take note of the employment rate, which is defined simply as the number of employed persons out of the working-age population. The employment rate is not affected by the difficulty in classifying the non-employed as unemployed or as non-participants. In addition, it will be important to distinguish between the number of work hours of employed persons and the rate of those partially employed—and those among them who are employed part-time involuntarily—since this group as well may expand when workers placed on unpaid leave gradually return to their jobs.
According to the Research Department’s forecast, the recording difficulty that is unique to this period is expected to abate in the second half of the year for two reasons: first because unpaid leave as an alternative to employment or dismissal is a temporary solution, and second because a reduction of health restrictions will enable anyone who was placed on unpaid leave to more actively seek work.
The Central Bureau of Statistics definitions are based on international standards, and are founded on common practice in many countries in the world. However, during the interim period, the Employment Service figure on job seekers (according to which about 25 percent of those employed in the economy are registered as job seekers, most of them on unpaid leave) reflects a phenomenon that we must analyze, in addition to the effect it is having on the National Insurance budget. This figure also made a significant contribution to monitoring the state of the economy when the crisis began, since it was published on an almost daily basis, and without a lag. However, it is important to note its two characteristics: It does not reflect the full temporary impact on employment, since it does not include the self-employed who did not work or a large portion of salaried employees who are not eligible for unemployment benefits; and it does not reflect continuing unemployment, since a significant portion of these job seekers are expected to return to employment when full economic activity resumes.
The terminological difficulty explained herein is reflected in the forecast published recently by the International Monetary Fund, which estimated that the average unemployment rate in 2020 will be 12 percent (compared with 6 percent in the Bank of Israel forecast). The IMF based its unemployment forecast partly on Employment Service data, and included as unemployed all workers who had been placed on unpaid leave (about 800,000 people), while the Bank of Israel forecast was based on the assumption that only a few of the workers on unpaid leave—up to 200,000 people—would actually be defined in the end as unemployed. As such, the forecasts of the IMF and the Bank of Israel differ mainly in relation to the second quarter. The forecasts in relation to the second half of the year are quite similar.