Highly Regionalized Unemployment Rates: Bureau of Labor Statistics March Data

[StateBoiler]

After a highly unusual several-month lag, the US Bureau of Labor Statistics has finally updated its state and local labor market statistics, so we will dedicate today’s Data section to these numbers.

The central message here is that US unemployment trends are very location-specific, something the national data obviously fails to capture. By looking at state- and city-level data we can more accurately assess both how far the domestic labor market has recovered and what factors will drive unemployment/labor force participation trends from here.

Two points on this topic:

#1: The current difference in unemployment rates across US states are very large (March 2021 data here).

12 states are already back to essentially full employment (2 – 4 percent unemployment): Nebraska, South Dakota, Utah, Vermont, New Hampshire, Idaho, Iowa, Kansas, Alabama, Montana, Wisconsin and Indiana.

19 states have unemployment between 4 – 6 percent (note: March national joblessness was 6.0 percent): Minnesota, Missouri, Oklahoma, Arkansas, North Dakota, Georgia, Florida, Ohio, Maine, Kentucky, Tennessee, Michigan, South Carolina, Virginia, North Carolina, Wyoming, Washington, West Virginia and Oregon.

8 states have unemployment rates between 6.1 and 7.0 percent, or close to the national level: Maryland, Mississippi, Colorado, Delaware, Alaska, Arizona, Massachusetts and Texas.

That leaves 11 states with noticeably higher unemployment rates (over 7 percent), and several are in the most populous areas of the country:

California (12 pct of US population): 8.3 percent unemployment
New York (6 pct): 8.5 percent unemployment
Illinois (4 pct): 7.1 percent unemployment
Pennsylvania (4 pct): 7.3 percent unemployment
New Jersey (3 pct): 7.7 percent unemployment

Smaller states with +7 percent unemployment: Rhode Island, Louisiana, Nevada, Connecticut, New Mexico, and Hawaii.

Takeaway: US unemployment is clustered in a handful of states (CA, NY, IL, PA, and NJ), with much of the rest of the country already approaching, near, or at full employment.

#2: American big city joblessness is a large part of the country’s overall unemployment challenge, as these 2 points show:

First, consider February/March’s national unemployment rates as compared to America’s 10 largest cities in the same months:

March 2021 national unemployment: 6.0 percent
February 2021 national unemployment: 6.2 percent

Versus:

New York City unemployment Rate: 11.2 pct (March)
Los Angeles: 10.9 pct (March)
Chicago: 7.7 pct (March)
Houston: 8.4 pct (Feb data, latest available)
Phoenix: 6.7 pct (Feb data)
Philadelphia: 11.2 pct (Feb data)
San Antonio: 6.8 pct (Feb data)
San Diego: 6.9 pct (March)
Dallas: 6.9 pct (Feb data)
Austin: 5.6 pct (Feb data)

What this shows: average top-10 big American city unemployment (Feb/March 2021) is 8.2 percent, much higher than the national 6.0 – 6.2 percent rate.

Second, let’s zoom in on New York and Los Angeles and see how their unemployment situations both differ from and inform the US experience as a whole:

First, keep in mind that the US labor force (people employed or looking for work) shrank from February 2020 to March 2021 by 3.8 million people (down 2.3 percent). This highlights the problem of declining labor force participation that Fed Chair Powell regularly mentions.

NYC has not seen the same drop in workforce size. The labor force here has declined by only 9,186 people (-0.2 percent).
Los Angeles is much more in line with the national decline in labor force size, down 115,124 from February 2020 (-2.2 percent)

Add up unemployment in New York City (462,131) and Los Angeles (551,124) and you get 1,013,255 jobless workers, or 10 percent of the US total. These 2 cities only make up 3.7 percent of the US population.

[Okay, maybe Mike Johnson is a competent parliamentarian.]

Interesting data. Thanks for noticing these patterns!

There are definite policy failures in the Woke Big City Way of doing things. I'm not educated enough on how big cities work (only having lived in one for two years of my life) to hazard a guess at what those policy failures are exactly, but at this point it's pretty unmistakable that something is rotten in the states of New York and California.

[支持核绿派 (Greens4Nuclear)]

Interesting data. Thanks for noticing these patterns!

There are definite policy failures in the Woke Big City Way of doing things. I'm not educated enough on how big cities work (only having lived in one for two years of my life) to hazard a guess at what those policy failures are exactly, but at this point it's pretty unmistakable that something is rotten in the states of New York and California.

NYC could be people who would've "dropped out of the labor force" elsewhere continuing to file for unemployment?

I know Asians (which both LA and NYC have a lot of) have been hit particularly hard by unemployment, but I doubt that alone explains what we're seeing here.

[Benjamin Frank]

Interesting data. Thanks for noticing these patterns!

There are definite policy failures in the Woke Big City Way of doing things. I'm not educated enough on how big cities work (only having lived in one for two years of my life) to hazard a guess at what those policy failures are exactly, but at this point it's pretty unmistakable that something is rotten in the states of New York and California.

NYC could be people who would've "dropped out of the labor force" elsewhere continuing to file for unemployment?

I know Asians (which both LA and NYC have a lot of) have been hit particularly hard by unemployment, but I doubt that alone explains what we're seeing here.

The Bureau of Labor Statistics does not use those who file for unemployment in their data.  Their statistics are compiled by a monthly survey they conduct.

I also suggest hesitating to read too much into these unemployment statistics without a similar regional breakdown of the labor force participation rate.

[🐒Gods of Prosperity🔱🐲💸]

my guess is that working class jobs in high COVID restriction states would be most impacted.  Does that look like the case?

[TheDeadFlagBlues]

I wouldn't put too much stock into Local Area Unemployment Statistics (LAUS). First, even state-level labor force statistics have been very strange as of late, as the Current Population Survey has featured some pandemic-related sampling problems. Second, unemployment rates are very misleading, as many people have been misclassified or have dropped out of the labor force. Third, the LAUS statistics are partially based on unemployment insurance claims and claims activity has been off. While the overall picture seems accurate to me, it would be a mistake to treat these data as revelations from BLS - think of them as a point estimate with massive confidence intervals.

At some point over the next year or two, we'll get a really clear picture when the QCEW county-level private payroll employment estimates are released. This is based on a very stable methodology of a large sample of employers and will give a clear picture of the effect of lockdowns on employment. Looking at changes in employment will be the best way of understanding displacement in the labor market when survey measures on the household side are messed up, as they are right now!