The August Employment Report is out and it came amazingly close to expectations finding 1.37 million jobs versus the 1.35 million median estimate. Despite the positive beat, the scope of the damage still to be repaired is daunting. 10.5 million jobs have been created in the last four months versus 22.2 million jobs lost in March and April. Thus, more than 11 million workers remain unemployed from the pandemic. The surprising result this month was the dip in the unemployment rate, dropping from 10.2% to 8.4%, easily beating the pre-release expectation of 9.8%. The issue here is the much smaller Household Survey—which is used to derive the unemployment rate—found 3.76 million new jobs compared to the much larger Establishment Survey’s 1.37 million jobs. That’s the largest discrepancy in the history of the jobs report. We’ll lean towards the larger Establishment Survey here and expect some adjustment in the Household Survey in coming months, but it’s still an impressive result just perhaps not as impressive as it looks at first blush. Also, for the sixth straight month, the BLS noted obvious errors in survey responses, but they are improving. Instead of adding 3% to the official rate if corrected as in prior months, the August adjustment is estimated at just 0.7%, similar to the July error ate. We look at some more of the details in the report below.
The unemployment rate fell from 10.2% to 8.4%, easily beating the 9.8% expectation, but for the sixth month in a row the BLS noted errors in survey responses, but for the second straight month they were way down. If the responses were adjusted they would have added perhaps 0.7% to the official rate. The Household Survey—which is used to generate the various employment ratios— found 3.76 million new jobs and compared to the much larger Establishment Survey’s 1.37 million jobs the discrepancy is the largest in the history of the report. Because of the larger size of the Establishment Survey we’ll lean in that direction and expect an adjustment in the Household Survey in the near future. It also reported that 2.8 million persons left the unemployment rolls (13.5 million versus 16.3 million) while 968 thousand persons joined the labor force, which as the denominator also helped drop the rate.
For the month, 1.37 million jobs were added to payrolls (1.027 million private sector plus 344 thousand government jobs, mostly for census workers). That total slightly beat the 1.35 million expectation, but with a wide dispersion of estimates coming in, anything close to the median is a win. While the leisure/hospitality field led job gains since the March/April plunge, it was the retail sector that led in August, gaining 249 thousand jobs while leisure/hospitality followed at 174 thousand. The services sector as a whole added 984 thousand jobs versus 1.4 million in July and 4.2 million in June. Health care and social assistance, as you would expect, added jobs as well with 90 thousand versus 196 thousand in July. Goods-producing jobs continue to slow adding only 43 thousand versus 61 thousand in July and 485 thousand in June. Construction added 16 thousand and manufacturing 29 thousand jobs. 2 thousand mining jobs were lost. Those numbers seem at odds with the solid run of ISM Manufacturing reports this summer.
The underemployment rate (unemployed plus part-time workers wanting a full-time job and those wanting a job but not currently looking) fell from 16.5% to 14.2%. In addition to the drop in unemployed the rate was helped by a 871 thousand decline in part-time workers. The Fed will want to see this rate back in the mid to high single-digits before declaring victory.
Drop in Initial Jobless Claims Has More to do with Seasonal Adjustments
On the face of it, initial jobless claims falling below a 1.0 million for the first time since the pandemic struck would be a good sign but like so much else about this virus it’s more to do with funky seasonal adjustments than honest to goodness improvement. Applications for regular unemployment benefits fell last week, reflecting a change to seasonal adjustments that are normally designed to smooth regular swings in the data but ended up distorting figures because of the pandemic. Initial jobless claims totaled 881,000 last week. That follows 1.01 million in the prior week, a figure that’s not directly comparable because of a change to seasonal adjustment methodology. On a non-seasonally adjusted basis claims rose by 7,591 to 833,352 last week, led by a 39,958 increase in California. Applications under the separate federal program which targets the self-employed and gig workers typically not covered by traditional unemployment insurance jumped by 152,000 to 759,000. So a mixed message to be sure.
The Dollar and TIPs Breakeven Rates
There’s been a lot made of the increasing TIPs breakeven rates since the lockdown-induced panic of March and April. And a lot of the that upward movement can be attributed to the weakening US dollar (white line), while the breakeven rates (blue line) head the other way. The falling dollar can be attributed to a number of things: uber-dovish Fed, a fading fiscal stimulus, and less-than-ideal response and taming of the virus, other countries getting back quicker than the US. It’s interesting to note, however, while TIPs breakeven rates have risen they remain below the Fed’s 2% target and generally in the range that prevailed prior to the pandemic panic. We think the move is more to do with a deflationary scenario being reversed rather than a real fear of gathering inflation.
|Treasury Curve||Today||Chg Last Wk.||LIBOR Rates||Today||Chg Last Wk.||FF/Prime||Rate||Swap Rates||Rate|
|3 Month||0.10%||0.01%||1 Mo LIBOR||0.15%||Unch||FF Target Rate||0.00%-0.25%||3 Year||0.237%|
|6 Month||0.11%||Unch||3 Mo LIBOR||0.25%||+0.01%||Prime Rate||3.25%||5 Year||0.3349%|
|2 Year||0.14%||+0.03%||6 Mo LIBOR||0.29%||-0.02%||IOER||0.10%||10 Year||0.683%|
|10 Year||0.67%||+0.13%||12 Mo LIBOR||0.43%||-0.01%||SOFR||0.10%|