The preliminary seasonally adjusted unemployment rate of 3.2 percent for May was up slightly from 3.0 percent in March and April and down from 3.9 percent one year ago. The number of unemployed was down 4,000 over the year to 22,600. The slight unemployment rate increase was expected; it does not indicate a deterioration in workforce conditions (see note 1 below). The unemployment rate has been below 4 percent for the last eight months and 16 of the last 19.

The U.S. preliminary unemployment rate of 4.3 percent for May was little changed from 4.4 percent for April and down from 4.7 percent one year ago. The New England unemployment rate averaged 4.1 percent.

May unemployment rate estimates for other states in the region were 2.9 percent in New Hampshire, 3.1 percent in Vermont, 4.2 percent in Massachusetts, 4.1 percent in Rhode Island, and 4.9 percent in Connecticut.

The employment to population ratio estimate of 62.0 percent remained above the U.S. average of 60.0 percent.

Payroll Survey Estimates - The preliminary nonfarm payroll jobs estimate of 619,500 for May was up 3,400 from one year ago. Nearly all growth was in the private sector with little change in government jobs. The largest job gains were in the healthcare, construction, and hospitality sectors.

**Not Seasonally Adjusted Substate Data**

The not seasonally adjusted statewide unemployment rate estimate of 3.5 percent for May was down from 3.8 percent one year ago. The lowest rate was 2.7 percent in Cumberland County and the highest was 6.2 percent in Aroostook County.

The unemployment rate was below the statewide average in the Portland-South Portland (2.7 percent) and Lewiston-Auburn (3.2 percent) metro areas and close to the average in the Bangor metro (3.6 percent).

June estimates will be released on Friday, July 18 (Data Release Schedule: http://www.maine.gov/labor/cwri/releaseDates.html ).

This release is available at http://www.maine.gov/labor/cwri/news/release.html .

Labor force and unemployment data is available at http://www.maine.gov/labor/cwri/laus1.html .

Nonfarm payroll jobs data is available at http://www.maine.gov/labor/cwri/ces1.html .

Monthly workforce estimates are cooperatively produced and released by the Maine Department of Labor, Center for Workforce Research and the U.S. Bureau of Labor Statistics.

**NOTES**

1. Preliminary labor force estimates, including rates (labor force participation, employment, and unemployment rates), and levels (labor force, employed, and unemployed) tend to move in a direction for several months and then reverse course. Those directional trends are largely driven by a smoothing procedure and may not indicate a change in underlying workforce conditions. Annual revisions (published in March each year) tend to moderate or eliminate those directional patterns. A comparison of 2016 preliminary and revised unemployment rate estimates is available at http://cwri.blogspot.com/2017/03/2016-workforce-data-revisions-stronger.html.

2. The 90 percent confidence interval for statewide unemployment rates in 2017 is 0.7 percentage points above or below the published estimate each month.

3. To assess employment growth, we recommend looking at nonfarm jobs from the payroll survey rather than resident employment from the household survey. The payroll survey is larger, has smaller margins of error, and is subject to smaller revisions. More on the differences in accuracy of the two measures is at http://cwri.blogspot.com/2016/04/reports-that-maine-is-losing-jobs-are.html .

4. Nonfarm payroll jobs estimates tend to be volatile from month to month because there is variability in the sample of reporting employers and their representativeness for the universe of all employers. Additionally, seasonal adjustment is imperfect because weather, the beginning and ending of school semesters and holidays, and other events do not always occur with the same timing, which can exacerbate monthly volatility. Users should look to the trend over multiple months rather than the change from one specific month to another. Estimates for the period from October 2016 to September 2017 will be replaced with actual payroll data in March 2018. Those benchmark revisions are likely to show less volatility than preliminary estimates.