With the number of COVID-19 cases in the US skyrocketing and a rising share of the population under lockdown, we need to revise our macro forecasts quite significantly: We now expect second-quarter real GDP to fall at a 40% annualised pace, with employment declining by 14 million and the unemployment rate spiking to 12%.
- With the number of COVID-19 cases in the US skyrocketing and a rising share of the population under lockdown, we need to revise our macro forecasts quite significantly: We now expect second-quarter real GDP to fall at a 40% annualised pace, with employment declining by 14 million and the unemployment rate spiking to 12%.
- That 40% annualised decline in GDP is significantly larger than our previous forecast of a 10% fall. But the latter was based on a view that the number of cases nationally would peak in the tens of thousands. As it stands now, if the number of cases continues to double every three days, New York alone will have more than 100,000 cases in less than a week’s time. The virus will inevitably spread aggressively to other states requiring widespread economic closures to try and contain it and limit the number of deaths. Admittedly, our forecast is slightly more pessimistic than other US forecasters (at least today), but it is broadly in line with the impact we expect in other countries. We’re not trying to attract attention with it, this is just where a dispassionate analysis takes us.
- Normally when forecasting GDP we would begin with the disaggregated expenditure breakdown (consumption, investment, etc), but in this case it makes more sense to start with the industrial breakdown, as we did in an earlier note here. Table 1 shows the percentage declines we have assumed for each sector, with non-food retail, air transportation, accommodation, food services and arts & entertainment the hardest hit. But with manufacturers now reporting widespread shutdowns, the factory sector will be hit harder than we originally assumed, even if supply chain disruptions linked to the earlier closure of China’s economy are partly reversed over the next few months. It is also unlikely that other parts of the services sector will escape completely unscathed. The collapse in oil prices will devastate the shale oil industry.
- Aggregating those sectors, we estimate a 12% q/q decline in second-quarter real GDP, which equates to 40% annualised.
- We then assume that real GDP rebounds by 4% in the third quarter (17% annualised) and by 5% in the fourth (22% annualised). (See Chart 1.) Although the second quarter may see the biggest hit, in a country the size of the US, it makes sense that the epicentres of the outbreak will shift gradually from state to state; meaning that the third-quarter rebound could be relatively disappointing. Overall, we now expect real GDP to decline by 5.5% in 2020 before rebounding by 6.5% in 2021. Nevertheless, that would still leave the level of GDP at end-2021 slightly lower than it otherwise would have been without the coronavirus impact. (See Chart 2.) That reflects our belief that the international border restrictions will be maintained even after the domestic economy opens again. It could take a decade for the tourism industry to recover.
- We have assumed that within each sector the percentage decline in employment is half the size of the decline in GDP. (i.e. if GDP falls by 80% in the food services industry, then employment falls by 40%. Nevertheless, the impact on employment is still very substantial because the worst affected sectors – retail, food services – account for much bigger shares of total employment than output. (See column 3 of Table 1.) Column 4 shows that the total decline in employment is likely to be around 14 million, with retail, food services and accommodation accounting for 10 million of that total. That would nearly triple the number of unemployed and push the unemployment rate up to 12%, slightly higher than its post-GFC peak.
- With states already reporting an unprecedented wave of jobless claims, we may be under-estimating the impact. In past downturns the percentage decline in employment was at least as big as the fall in real GDP, in which case the unemployment rate could hit 20%. (See Chart 3.) But, in this case, we think some employers will hold on to workers if they think the shutdown will only last a month or two. That implies a bigger-than-normal adjustment in hours worked and productivity and a smaller reduction in employment.
- Finally, the slump in energy prices means that we now expect headline CPI inflation to briefly fall below zero this year but, with the Phillips curve still flat, we don’t expect the surge in the unemployment rate to generate anything more than a modest decline in core inflation. (See Chart 4.)
Chart 1: Real GDP (%q/q Annualised)
Chart 2: Real GDP ($bn)
Chart 3: GDP, Employment & Hours Worked (%y/y)
Chart 4: CPI Inflation (%)
Sources: Refinitiv, CE
Table 1: Impact of COVID-19 Epidemic
% of Total
% Decline in Q2
% of Total
Mn Decline in Q2
Other Transportation & Warehousing
Finance & insurance
Arts & entertainment
Change in GDP
Mn Change in Employment
Sources: BEA, BLS, CE
Paul Ashworth, Chief US Economist, email@example.com