Via the excellent George Tharenou at UBS:
Unemployment rate 7.1% (after 6.4%); & underutilisation a record high 20.2%
The unemployment rate also rose more than (consensus) expected, from 5.2% in March, to an upwardly revised 6.4% in April (was 6.2%), and to 7.1% in May (UBS: 7.0%, mkt: 6.7%), now the highest since 2001. A more internationally comparable measure was 9.5% (from 5.4% in March); albeit still well below 13.7% in the US and 13.3% in Canada. The participation rate surprisingly declined much further to 62.9%, the lowest since 2000 (after 63.6% in April, and a near record high 65.9% in March). Indeed, the broader underutilisation rate – which includes underemployment, and is a more accurate measure of labour market slack, especially in the current environment – has spiked much more significantly, ticking up further to a record high 20.2%, (after 20.1% in April, and 14.0% in March), to well above the 1992 peak of 18.2%.
2.3mn affected by job loss or reduced/no hours; implies ‘real’ UR is 13.7%
Given the huge drop in participation, the LFS measure of unemployed increased ‘only’ 86k m/m and 211k across the last two months. However, a total stock of ~2.3mn people in May were affected by either job loss (0.7mn), or reduced or no hours (1.55mn) for ‘economic reasons’, of which 360k worked zero hours. This was actually ‘less bad’ than April when the total was 2.7mn (and 750k worked zero hours). Nonetheless, if we adjust the unemployment rate to include those ‘not working at all’, and also assume the participation was unchanged since March, then the ‘real unemployment rate’ would now be 13.7% (very far above the reported 7.1%).
Jobs track our long held -7%, but UR formally revised down to 8% (was 10%)
Negatively, May’s LFS was worse than expected. Positively however, the large reduction in people not working in May vs April, combined with the recovery of payrolls in late May, & ongoing easing of restrictions, overall indicate the labour market is close to a trough. The cumulative decline of employment is now tracking towards of long-held forecast of 7% (0.9mn). However, the further drop in participation means we (formally) revise down our peak in unemployment to ~8% (was ~10%), assuming some bounce in the participation rate ahead as JobKeeper ends, albeit the UR is likely to show less improvement than expected ahead, amid the ~$100bn q/q policy cliff in Q4.
Given the difficulties posed to the sharing economy by the virus, I would no be surprised to see the UE rate stay much higher for longer this cycle as what were “underemployed” this time are “unemployed”.