WHEN DID THE U.S. DOWNTURN START AND WHAT SHOULD BE DONE ABOUT IT?

It appears that the United States and other countries have entered recessions. Focusing on the U.S., the U.S. Bureau of Economic Analysis (https://www.bea.gov/news/2020/gross-domestic-product-1st-quarter-2020-advance-estimate) reported that seasonally adjusted, annualized real GDP fell by about 4.8 per cent in the first quarter of 2020. Given the COVID-19 coronavirus shutdown, many if not most or all expect a larger decrease in U.S. real GDP in the current quarter (2020q2). Two or more consecutive quarters of decreases in real GDP are often considered as a recession according to undergraduate textbooks such as Macroeconomics (second edition) by Krugman and Wells (2009, p. 160, Worth Publishers), although Krugman and Wells point out that a precise definition of a recession does not exist. The National Bureau of Economic Research (NBER) will determine based on factors like quarterly real GDP growth or lack thereof and employment whether or not the U.S. entered a recession, and if so announce the month that it began.

According to the FRED website of the St. Louis Federal Reserve Bank 

(https://fred.stlouisfed.org/data/CE16OV.txt) with seasonally adjusted data from the U.S. Bureau of Labor Statistics (BLS), employment based on the household survey decreased four out of the last six months (November 2019 through April 2020) based on my calculations.  The only exceptions were December 2019 and February 2020, but my calculations indicate that the increase for February was only about 45,000 jobs.  Although the estimated decrease between October and November 2019 was only about 8,000 jobs and is quite small compared with the estimated decreases of nearly 2.99 million jobs between February and March of 2020 and about 22.37 million jobs between March and April of 2020, it is still a decrease. 

Using the BLS establishment survey instead, my calculations find seasonally adjusted U.S. nonfarm payroll employment increased each month from November 2019 through April 2020 except for March and April 2020, when the decreases were about 881,000 jobs and roughly 20.5 million jobs, respectively, both less than the job losses from the household survey. (https://fred.stlouisfed.org/data/PAYEMS.txt)  If my calculations are correct, the monthly change in U.S. nonfarm payroll employment based on the establishment survey has been either more positive or less negative in five of the last six months.  Is this important?

Will the NBER consider retail sales data in its determination of whether or not the U.S. economy is in recession, and if so, when it began? According to data from the U.S. Census Bureau (https://www.census.gov/retail/marts/www/marts_current.pdf), advance total retail sales (seasonally adjusted) for April 2020 plummeted by approximately 16.4 per cent, following a substantial decrease in March 2020 by about 8.3 per cent.  However, census data accessed from the St. Louis Federal Reserve web page (https://fred.stlouisfed.org/data/RSAFS.txt) show that this series also decreased in February 2020, albeit by only about 0.44 per cent, according to my calculation.

Regardless of when the downturn started, I believe that more should be done to help those who have lost jobs and/or income. Additional tax cuts and increases in government spending benefiting those who need additional disposable income the most and would be most likely to spend a large percentage of their newfound income would be appropriate, given that monetary policy may be either at or near its limit in terms of what it can achieve in increasing economic growth. Interest rates in the U.S. are already quite low and money velocity has been falling, raising the question of how frequently U.S. dollars newly added to the economy are being spent in real GDP transactions. Fiscal policy may be much more effective in ensuring that spending in the economy increases and in reducing economic inequality.

(Note that the time series variables mentioned above may be revised. Such revisions could change the analysis in this blog entry. Also, as time passes, data releases that are 'current' as of May 15, 2020 will no longer be current. )