AUGUSTA — Unemployment in Maine was the lowest in more than 42 years, according to a March 23 release from the Maine Dept. of Labor. Both the rate and number of unemployed were the lowest recorded since the current methodology was implemented in 1976.

**Household Survey Estimates** - The preliminary seasonally adjusted unemployment rate estimate of 2.9 percent for February was the lowest recorded since at least 1976. The rate was little changed from 3.0 percent for January and down from 3.3 percent one year ago. The number of unemployed declined 3,000 over the year to 20,000. Maine's unemployment rate has been below 4.0 percent for 27 consecutive months, the release said.

The U.S. preliminary unemployment rate of 4.1 percent for February was unchanged over the last five months, but down from 4.7 percent one year ago. The New England average was 3.6 percent.

February unemployment estimates for other states in the region were 2.6 percent in New Hampshire, 2.8 percent in Vermont, 3.5 percent in Massachusetts, 4.5 percent in Rhode Island, and 4.6 percent in Connecticut.The employment to population ratio estimate of 61.5 percent remained above the 60.4 percent U.S. average.

**Payroll Survey Estimates** - The 626,100 preliminary nonfarm payroll jobs estimate for February was up 3,600 from one year ago. Job gains were primarily in the healthcare and social assistance and professional and business services sectors.

The not seasonally adjusted statewide unemployment rate estimate of 3.3 percent for February was down from 4.1 percent one year ago. Unemployment rates decreased over the year in all counties. The lowest rate was 2.4 percent in Cumberland County and the highest was 5.2 percent in Washington County.

The unemployment rate was below the statewide average in the Portland-South Portland (2.5 percent) and Lewiston-Auburn (3.0 percent) metro areas and close to the average in the Bangor metro area (3.4 percent).

March workforce estimates will be released Friday, April 20 (Data Release Schedule: http://www.maine.gov/labor/cwri/releaseDates.html ).

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

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

Nonfarm payroll jobs data is available at 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 2017 preliminary and revised unemployment rate estimates is available at www.maine.gov/labor/cwri/blogs/2017_workforce_data_revisions.pdf .

2. The 90 percent confidence interval for statewide unemployment rates in 2018 is 0.6 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 www.maine.gov/labor/cwri/blogs/imprecise_data.pdf .

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 2017 to September 2018 will be replaced with actual payroll data in March 2019. Those benchmark revisions are likely to show less volatility than preliminary estimates.