This recession really is different.
All other downturns in the past 50 years have been led by goods-producing sectors (mining, construction, manufacturing).
This is the first services-led recession.
"Republicans thought 38 percent of Democrats were gay, lesbian or bisexual, while the actual number was about 6 percent. Democrats estimated that 44 percent of Republicans make more than $250,000 a year. The actual share was 2 percent."
The June COVID19 surge is much more widespread than the April spike. It is not concentrated in a few counties or hotspots.
More people live in counties where cases per capita are high and rising today than at any point in the pandemic.
Remote work has kept many people employed, healthy, and safe during the pandemic.
But the benefits of remote work are mostly for the fortunate. Very few people without a college degree are working from home.
Core unemployment jumped in June to 5.9% from 5.0% in May. It's rising at an accelerating rate.
Core unemployment removes temporary layoffs and adds the marginally attached.
We're now tracking state job search.
Search activity fell BELOW national trend in states cutting off federal UI imminently (June 12 and 19). That's not what you'd expect if people were upping their searches ahead of UI ending ...
4/
Let's be clear: permanent unemployment is still rising.
Temporary unemployment spiked in April and dropped in May. It's huge and is driving the headline unemployment rate. But strip that out, and the picture looks very different.
1/
The metros that swung most Democratic in 2020 vs 2016 include several in Colorado.
In general, highly educated places swung away from Trump. The partisan education divide widened.
4/
I was just the Under Secretary for Economic Affairs at the U.S. Commerce Department. Here's the economic research that was really helpful to us.
(Thank you
@mattyglesias
for the opportunity to guest-post!)
But the main point is:
Permanent unemployment climbed in March, April, and May. The number of "permanent job losers" is up 79% in the last three months.
Summary table here:
This is no time to be complacent about the labor market.
6/end
Jed Kolko is an accomplished economist with the ability to transform data into research that is both accessible & actionable.
I am excited to welcome his leadership at one of the world's leading statistical agencies,
@BEA_News
.
Iowa and New Hampshire rank 37 and 41 among states in demographic similarity to the US overall.
The top two: Illinois and New York.
‘Normal America’ Is Not A Small Town Of White People | FiveThirtyEight
The current surge in COVID19 cases is much more widespread than the April wave. It's not concentrated in a few hotspots.
38% of the US lives in counties where the rate of new cases is now high and rising -- way above the previous peak of 21% in April.
Another view of Hispanic/Latino vote:
On average, higher % Hispanic counties swung more Republican, but entirely driven by majority-Hispanic counties mostly in Texas and Miami. Very non-linear.
Suggests specific regional swings or strategies more than broad realignment?
1/
My two-part mantra for making sense of NIMBYism and other failures of housing policy.
1. 80% of voters are homeowners.
2. Homeowners hear "affordability" and think "my property value."
Unfortunately, some days it seems as simple as that.
Coronavirus and inequality, for the thousandth time.
% of workers teleworking because of pandemic:
5% of no HS degree
13% of HS degree only
22% of some college / associates
48% of bachelors degree
63% of graduate degree
1/
County vote data are nearly all reported. (Finally!)
Here's what happened in the 2020 election.
New from
@MonkovicNYT
and me this morning in
@UpshotNYT
1/
Unemployment fell in May thanks to a drop in temporary layoff. But more people permanently lost their jobs. Time for a new measure -- the core unemployment rate.
My latest in
@UpshotNYT
1/
Don’t Cheer Too Soon. Keep an Eye on the Core Jobless Rate.
I can't stop wondering whether part of the labor supply story is rising value of leisure time.
If you haven't traveled or seen loved ones for 15 months, and are burnt out, then vaccinations and reopenings make leisure time really valuable.
Not sure how to measure or assess.
Job postings on
@indeed
20% below last year's trend, as of August 28.
Teensy uptick vs. last week, after two weeks of backsliding.
Bigger picture: job postings flattened in Aug after 3 months of gradual gains. Now back up halfway from low point.
1/
We're now tracking state job search.
Search activity fell BELOW national trend in states cutting off federal UI imminently (June 12 and 19). That's not what you'd expect if people were upping their searches ahead of UI ending ...
4/
Every once in a while, mundane data-checking turns into a thing of beauty and delight.
To double-check the geocoding of search locations, I made a simple scatterplot:
twoway scatter ipLat ipLong if state_fips_code=="48"
That big and persistent drop in 55+ labor force participation from the pandemic? It's almost entirely population aging. Age-adjusted 55+ LFPR is basically back to pre-pandemic peak.
A short jobs-report-week thread.
Mothers' employment fell in August, even though employment rose for other groups.
With the Delta surge, school and daycare reopenings have been shaky. It's a big burden on mothers and other caregivers.
(New CPS microdata for August.)
Most common age in U.S.:
White: 61
Asian: 32
Black or African-American: 30
American Indian & Alaska Native: 29
Hispanic: 14
as of June 2021.
Census data and racial/ethnic definitions.
These are modes. Medians are in next tweet.
Latest: US job postings on
@indeed
are 6.7% above pre-pandemic baseline, as of last Friday 3/5.
Up full point vs week earlier.
Led by sectors that make & move stuff. Also pharmacy, nursing, & medical jobs.
Good steady improvement!
(end; no blogpost this week)
Core unemployment rose in August to 5.8%, just shy of its June high.
Core unemployment focuses on the permanently unemployment and marginally attached, removing temporary unemployment.
1/
Did the pandemic upend migration patterns? Not so much! The places that gained and lost people in 2020 were pretty much the same places as in 2019.
New! in
@UpshotNYT
with
@emilymbadger
@qdbui
Metros with the biggest shift away from Trump in 2020 vs 2016:
Colorado Springs, CO
Fort Collins, CO
Norwich-New London, CT
San Luis Obispo-Paso Robles, CA
Portland-South Portland, ME
1/
In the Texas counties where a majority of the population is of Mexican origin (Census definition), Trump did 6.6 points better in 2020 than in 2016. But outside of Texas, majority-Mexican-origin counties (mostly California) actually swung about half a point away from Trump.
2/
Texas border metros and Miami swung most toward Trump.
The correlation between county % Hispanic and swing toward Trump was 0.41. Even excluding Miami and Texas, the correlation was still 0.30.
(Note: Utah strongly affected by 3rd party votes.)
5/
The 2020 vote was very similar to 2016 -- and slightly LESS polarized.
Correlation of county vote margin between 2016 and 2020 was 0.99 (not a typo). And Trump-voting counties in 2016 swung more to Biden than Clinton-voting counties in 2016 did.
2/
Most common age in U.S.:
White: 59
Asian: 30
Black or African-American: 28
American Indian & Alaska Native: 27
Hispanic: 12
as of June 2019.
(Census data and racial/ethnic definitions.)
New COVID19 case rates have been lowest when average daily temperatures are in the 60s. New case rates are higher in the heat, and highest in the cold.
The coldest months are yet to come.
(Pooled across all counties and all months, Mar to Nov.)
Americans today move at half the rate they did in the 1950s and early 1960s. At new low in 2019. Still one of the most important and not-definitely-explained trends out there.
Core unemployment has remained stubbornly high even though the headline rate plunged. More unemployment is shifting from temporary to permanent.
The number of permanently unemployed jumped 19% between July and August.
2/
New
#COVID19
cases per capita now more than twice as high in small metros & rural areas than in urban counties.
Dramatic shift from the spring, when
#COVID19
was concentrated in urban counties.
1/
Remote work has kept many people employed, healthy, and safe during the pandemic.
But the benefits of remote work are mostly for the fortunate. Very few people without a college degree are working from home.
Drumroll ...
A full classification of neighborhoods as urban, suburban, or rural -- for all Census tracts!
New research and data product from
@ShawnBucholtz
@Molfe
, based on the American Housing Survey.
1/
Spending on services is rising as businesses reopen, but remains extremely depressed. Spending on goods has stalled out but at an elevated level compared to before the pandemic.
Huge surge in spending in goods. Not so much services. Suggests people went out and spent their stimulus checks on goods (esp. cars). I'd expect services to rebound in Q2, when the vaccination rollout really took hold.
“Planners have very little influence on city size distribution and city growth rates, unless they take active, targeted measures to destroy the urban economies of the cities that have grown “too large.’”
From my Christmas reading: Order Without Design.
Nearly all of the jobs rebound has been in low work-from-home industries, which shut and are re-opening.
But jobs in high WFH industries down 4.5%, w little rebound. In Great Recession, jobs in those same industries fell 3.9% peak-to-trough. We're in for some long-term damage.
Let's be clear:
The number of people who are "permanently" employment (i.e. not on temporary furlough) was
167% HIGHER
in August than in February.
And yes 167% higher means 2.7x.
And climbing.
Core unemployment rose in August to 5.8%, just shy of its June high.
Core unemployment focuses on the permanently unemployment and marginally attached, removing temporary unemployment.
1/
Median age in U.S.:
White: 43
Asian: 38
Black or African-American: 35
American Indian & Alaska Native: 35
Hispanic: 30
Bi- or multi-racial: 21
as of June 2021.
Census data and racial/ethnic definitions.
Not "soft skills." Instead, deep skills.
Communication, social intelligence, and emotional intelligence are deep skills that increasingly valuable, under appreciated, and often strongest among those traditionally disadvantaged in the labor market. Plus, robots can't do them.
The summer surge in COVID19 cases is everywhere: urban, suburban, smaller-town, and rural areas.
Huge change from April, when case rates were much higher in urban counties.
1/
United is asking people to bid to volunteer to get bumped off my flight tomorrow from San Francisco to Atlanta. I assume that's a plane full of economists going to the
#ASSA2019
meetings. Time to test those auction models!
Update on local patterns in COVID19 deaths.
Albany GA has highest metro death rate. New Orleans almost 2x New York.
Based on New York Times counts through March 31. Raw data here:
1/
New!
Job postings on
@indeed
continue to climb:
16.4% above pre-pandemic baseline as of last Friday 4/2, another big 3-point weekly gain
broad-based increases across sectors and regions
(no blogpost this week)
Another week, another jump.
Job postings on
@indeed
up almost 2 points over previous week.
24.2% above pre-pandemic baseline, as of 4/30.
By next week, Honolulu might be only big metro where job postings are still below baseline. Stay tuned!
(No blogpost this week.)
Job postings on
@indeed
5% above the pre-pandemic baseline in US, as of 2/19. Another week of solid gains.
One bright spot: job postings up in arts & entertainment, which was hit hard in the pandemic. Now just 6% below baseline.
(No blogpost this week.)
Urban counties were slow to report, but now it looks like urban counties swung MOST toward Trump -- a full point. Rural counties moved little, and suburbs swung toward Biden.
That means the urban-suburban AND urban-rural divides narrowed.
6/
Personal jobs report!
Tomorrow is my last day at
@CommerceGov
.
I am so grateful to
@POTUS
and
@SecRaimondo
for the chance to serve as Under Secretary for Economic Affairs.
My colleagues at Team OUSEA,
@uscensusbureau
, and
@BEA_News
inspired me every day. Thank you!
"the shift to remote work explains over one half of the 23.8 percent national house price increase" since late 2019
"the evolution of remote work is likely to have large effects on the future path of house prices and inflation"
intriguing new research:
Urban population growth has essentially stopped.
This morning Census released its annual county population estimates for 2019. (Obviously all pre-COVID19.)
Short thread.
TL:DR from today's 2018 Census local population estimates:
Suburbanization marches on.
Suburbs and mid-size metros grew fastest in 2018. Small rural uptick. Biggest metros shrank. Let's dig deeper. (thread)
Whoa --
Goldman Sachs estimates July payroll growth of +1150k, well above consensus and way above ADP estimate.
(I don't play this predictions game but sometimes it's fun to watch.)
New: US job postings on
@indeed
36% above pre-pandemic baseline, as of July 9.
After slower growth in May, postings climbed more rapidly in June & early July.
As more employers want to hire, HR job postings have swelled -- now 64% above baseline.
(No blogpost this week.)
Ecological fallacy, incomplete data, and all that -- but still the data so far suggest that the Hispanic/Latino swing toward Trump was highly localized.
Lots more to dig into. So, for now, this is just a reminder not to overgeneralize.
5/end