I am monitoring the US Department of Labor’s weekly data releases for the unemployment insurance claimants account, that I reported in my last commentary on the US labour market – Tip of the iceberg – the US labour market catastrophe now playing out (April 6, 2020). Their latest release (April 9, 2020) – Unemployment Insurance Weekly Claims – shows that in the prior week ending April 4, 2020, the initial claims rose by 6,606,000, but this was down on the increase the week before by 261,000. In the last three weeks, the total initial claims is 16.8 million persons. The impacts are quite stark already. For example, as you will see, in just one month (March), service sector occupations have shed 36.7 per cent of the total jobs that were added in the ‘recovery’ period between January 2010 and February 2020. And given the timing of the surveys (biased towards earlier in the month), the situation was much worse by the end of March. It is quite obvious that this crisis is impacting heavily and disproportionately on the least-advantaged workers and communities in the US. This cohort always suffers during a recession. But this time, the specific occupation biases are exacerbating the problem and inequity, given the nature of the economic shock (closures, shutdowns etc). It means the fiscal support should be heavily weighted to assisting the most impacted both in terms of people, their families and the regions they live in. The maps show that the spatial impact of the downturn to date is also very uneven. As yet, I have not seen a commensurate response from the US government. The fiscal support funds so far announced do very little for the most impacted communities and people. They certainly shore up the top-end-of-town which, while predictable, will come back to haunt the nation in the years to come.
The Department of Labor provides an archive of the weekly unemployment insurance claims data back to July 1, 1967 – HERE.
The weekly data can be found in the – UI Weekly Claims Report.
The following reality emerges.
The Department of Labor reports that:
In the week ending April 4, the advance figure for seasonally adjusted initial claims was 6,606,000, a decrease of 261,000 from the previous week’s revised level …
The largest increases in initial claims for the week ending March 28 were in California (+871,992), New York (+286,596), Michigan (+176,329), Florida (+154,171), Georgia (+121,680), Texas (+120,759), and New Jersey (+90,438), while the largest decreases were in Nevada (-20,356), Rhode Island (-8,047), and Minnesota (-6,678).
Bringing together the archived data and the most recent release (April 4, 2020), the following table tells the shocking story.
|Week ending||Initial Claims (SA)||Weekly Change||4-Week Moving Average (SA)|
|March 7, 2020||211,000||-6,000||161,500|
|March 14, 2020||282,000||+71,000||232,500|
|March 21, 2020||3,307,000||+3,025,000||1,004,250|
|March 28, 2020||6,686,700||+3,560,000||2,666,750|
|April 4, 2020||6,606,000||-261,000||4,265,500|
If we assume all those new claimants were unemployed then the unemployment level by the end of March would be 22,849 thousand instead of the BLS estimate (take up to the end of the survey week – March 14) of 7,140.
This would mean that the unemployment rate would be 14.1 per cent by the end of March rather than the official BLS March figure of 4.4 per cent.
The peak unemployment rate during the Great Depression was 24.9 per cent in 1933, before the New Deal brought it down somewhat.
The next two graph show the extraordinary nature of the current period.
The first graph shows new weekly claimants for unemployment benefits from July 1, 1967 to the week-ending March 13, 2020 (that is two weeks ago).
You can see the big cycles in the US labour market – the mid-1970s downturn, the recessions in the early 1980s, the early 1990s recession, the early 2000s and then the GFC.
The next graph adds the last three weeks of available data (week-ending March 21 to April 4, 2020) to the entire sample.
The spike at the end of the graph shows how drastic the situation is in the US.
This is a quite extraordinary graph.
The spatial patterns
In the Department of Labor report cited above, we saw there were sharp differences in initial claimants across the US states. I decided to do some analysis and produce some maps to visualise the patterns that are emerging.
The first map shows the Sum of the Initial Weekly UI Claimants over the period February 29, 2020 to April 4, 2020 for the mainland American states (Alaska was 39,800 and Hawaii was 114,408).
To scale this better, the next map shows the changes expressed as a percentage of the Working Age Population in each state (Alaska was 7.32 per cent and Hawaii was 10.49 per cent).
This graph more starkly demonstrates where the loss of jobs is impacting most significantly.
The following Table presents the same data for those who prefer numbers.
The Occupational Impacts to date – low wage workers bearing the brunt
As regular readers will know I have been tracking the sectoral and occupational employment changes in the US in relation to earnings for some time now. I am working on a very disaggregated version of that approach – more about that in the coming months.
But for now I just updated my research into the cyclical shifts in employment at the main occupational levels for the US today.
And the results are very interesting.
I am exploring now is how is the US labour market collapse impacting on workers across the wage distribution.
The burning question is whether low-paid workers bearing the brunt of the downturn?
In the past, I have demonstrated that the proportion of jobs in the total employment in sectors that pay below-average pay has increased.
But at that level of aggregation, we are unable to say whether these jobs in question were high-pay or low-pay.
The next table helps to expand on that understanding.
It shows the net job losses (in the downturn) and net job gains (in the recovery to date) for the major occupations in the BLS classification.
I have sorted the occupations relative to median weekly earnings as at the December-quarter 2019.
Low-pay is 75 per cent of the median.
1. In the downturn, 90.6 per cent of the jobs lost were in occupations that paid below median weekly earnings (1.3 per cent of those were in low-paid occupations). Very few jobs (relatively) were lost in the higher paying occupations.
2. Given 86.1 per cent of the total jobs lost in the downturn were in sectors paying above average pay. The inference is that the jobs lost were predominantly the lower paying jobs in those sectors (although we cannot strictly compare mean and median in a wage distribution given the skewness).
GFC recovery to February 2020
1. In the upturn to February 2020, the net jobs added had not yet replaced those lost in the occupations with below median weekly earnings. 67 per cent of the net jobs added were in occupations with above median weekly earnings. That proportion has risen in the last three months to February 2020.
2. While only a small number and proportion of jobs were lost in the low-pay occupations in the downturn, the recovery saw a much larger number of those jobs added. Of the 33 per cent share of below median earning jobs added in the recovery, the proportion that were in low-pay occupations was currently 13.2 per cent.
3. This tells us that there is a polarisation going on in the occupational employment structure with a bias towards low-pay jobs in the below median weekly earnings occupations and towards jobs in the above median weekly earnings. This supports the evidence provided by the Job Quality Index data.
4. That is, over the recovery period a hollowing out around the overall median pay levels was going on.
The Covid downturn
The current crisis is very different to the GFC recession, given that a host of activities were forced to stop altogether ad authorities (slowly) enforced shutdowns and isolation.
The areas that are most affected include hospitality, sports, entertainment, restaurants and cafes, travel, and similar, which are major employers of low-paid, precarious workers.
So I expect the patterns of job loss to be somewhat different as a result.
The summary results are that:
1. Low-paid work has slumped by 3.41 per cent in March 2020.
2. In just one month, service sector occupations have shed 36.7 per cent of the total jobs that were added in the period between January 2010 and February 2020. Think about that!
2. Below-median pay jobs slumped by 2.88 per cent.
3. Above-median pay jobs fell by 0.48 per cent.
4. In terms of total numbers of jobs lost:
(a) 12 per cent (341 thousand) have been in above-median pay occupations.
(b) 88 per cent (2,507 thousand) in below-median pay occupations.
(c) 32.8 per cent of the total (933 thousand) have been in low-pay occupations (in the GFC downturn only 1.3 per cent of the jobs lost were low-paid).
The initial analysis, based on one-month’s data (March 2020) for the occupational sub-gropus is presented in the following Table.
|Occupational Group||Jobs lost/gained (000s)||Proportion of total jobs lost (%)|
|Management, business, and financial operations occupations||+92.0||n/a|
|Professional and related||-301.0||10.6|
|Installation, maintenance, and repair||-132.0||3.2|
|Below-median pay but not low pay||-1,574||55.0|
|Sales and related||-490.0||17.2|
|Office and administrative support||-2.0||0.1|
|Construction and extraction||-120.0||4.2|
|Transportation and material moving||-587.0||20.6|
|Farming, fishing, and forestry||-36.0||1.3|
It is quite obvious that this crisis is impacting heavily and disproportionately on the least-advantaged workers and communities in the US.
This cohort always suffers during a recession.
But this time, the specific occupation biases are exacerbating the problem and inequity, given the nature of the economic shock (closures, shutdowns etc).
It means the fiscal support should be heavily weighted to assisting the most impacted both in terms of people, their families and the regions they live in.
The maps show that the spatial impact of the downturn to date is also very uneven.
As yet, I have not seen a commensurate response from the US government.
The fiscal support funds so far announced do very little for the most impacted communities and people. They certainly shore up the top-end-of-town which, while predictable, will come back to haunt the nation in the years to come.
That is enough for today!
(c) Copyright 2020 William Mitchell. All Rights Reserved.