Data Crunchers Crunched by Data

Discrepancies in official data on the level of unemployment in Israel has policymakers in a quandary as to how to fashion a response in a situation fraught with uncertainty.

“Nobody has a clear estimate about the number of unemployed today” as a result of the coronavirus, Yaniv Bar, a research economist at Bank Leumi, said in a phone interview with The Times of Israel.

“We are working in a fog,” Bar said. “The data collection process has been messed up. There is a lot of uncertainty around this subject.”

In recent days, Central Bureau of Statistics (CBS), the research department of the Bank of Israel and the National Employment Service have published statistics that are startlingly at variance with one another.

The CBS, for example, said that in the first half of June 2020, the unemployment rate was 5.2 percent. The Employment Service said that the jobless rate was a staggering 21 percent. The Bank of Israel, using the CBS’ own numbers, pegged the rate at 6.2 percent for people ages 25-65.

Unless they each were visiting different countries, the differences were hard to reconcile.

Definitions of unemployment no doubt account for some of the discrepancies. CBS, which issued the lowest figure, did not include those who are temporarily without a job, such as workers who were furloughed due to the coronavirus. When adding those statistics, the rate jumps to over 10 percent, according to the agency’s own statement. That narrows the gap, but only somewhat, with the Employment Service’s 21 percent.

The Bank of Israel research team has further complicated matters (striving for more clarity, of course) by introducing a concept of “wider unemployment.” This would include the unemployed, the temporarily unemployed, and those who have exited the workforce. Crunching all those categories together lifts the unemployment figure to over 25 in the second quarter of 2020, bringing it within shouting distance of CBS.

“Normally we look at the official unemployment figure” for economic projections, Leumi’s Bar said. But “these days, when the official unemployment rate does not represent what is actually happening on the ground, we must look for alternatives.”

“In practice, we know there are many more people without employment than the official number of unemployed,” he said, adding that all sort of numbers “get tossed in the air.”

Those who are furloughed — on unpaid leave, and eligible for unemployment benefits— make the forecasts more difficult, since “we don’t know that all of those who are furloughed will exit the work force and become really unemployed” because some will eventually return to their jobs.

Whether consolidating the research and publication of economic statistics under one agency, as has been proposed in the past, would help, Bar did not say in the interview.