Here are four charts presentinginitial claims for unemployment insurance reported by Washington state’s Employment Security Department (ESD) by sector for the period from March 1 to June 6.

A note of caution about initial claims numbers: In May, Washington’s unemployment insurance system was hit by a rash of fraudulent claims (press release here). As a result, the unemployment insurance claims data released weekly by ESD have overstated the number of individuals who have truly lost their jobs. ESD has estimated that the value of fraudulent claims is in the $550,000,000 to $650,000,000 range. It is believed that most of the fraudulent claims were filed between May 1 and May 16.

The first chart shows the number of claims by sector.

The five sectors with the greatest number of claims are, in descending order, accommodation and food services; health services and social assistance; retail trade; construction; and manufacturing. Together these five sectors accounted for sixty percent of the initial claims.

The second chart shows initial claims sector as a proportion of the sector’s employment in the month of February. I have separated claims into three time periods: March 1 to May 2, May 3 to May 16 (the period believed to have the most fraudulent claims), and May 17 to June 6 (the most recent three weeks).

The total number of private sector claims from March 1 to June 6 is equal to 35.2 percent of February private sector employment. The education services sector has experienced the greatest total number of claims relative to February employment, 90.5%. However, because nearly one-half of those claims occurred between May 3 and May 16, the period with a high number of fraudulent claims, this 90.5% figure severely overstates the number of legitimate claims.

The third chart shows, by sector, the percentage of total March 1 to June 6 initial claims that were received during the problematic May 3 to May 16 interval.

For all sectors together, 19.7 percent were in the May 3 to May 16 interval. The variation across sectors is huge, from 58.5 percent for the utility sector, to 8.6 percent in the accommodations and food services sector. Generally, sectors for which the share of total March 1 to June 6 claims received during the May 3 to May 16 interval is high are sectors with low numbers of total claims relative to February employment.

The fourth chart is a scatter plot with claims as a share of February employment measured along the vertical axis and the sectors’ average weekly wages in 2018 (the most recent year for which average wages are available) is measured along the horizontal axis. Claims from the problematic May 3 to May 16 interval are excluded.

Low-wage sectors have had more initial claims relative to employment than high-wage sectors. The correlation coefficient between claims as a share of employment and weekly average wage is −.595.