Impact of U.S. Government Shutdown to Show Up in January Jobs Data
Twenty-five days into the latest U.S. government shutdown, it is appropriate to begin estimating the impact on the domestic labor market. Based on our research, we now expect the U.S. unemployment rate to increase by 0.3 percentage points in January, which would bring the metric to 4.1 percent in the month (currently 3.856 percent). We now anticipate the first estimate of the change in total employment for January 2019 will result in a net loss of jobs. That would represent the first monthly negative number since September 2010, or 99 consecutive months. At this time, we do not consider this move sufficient to cause longer-term damage to the labor market, but it is troubling. Moreover, the combination of what will almost certainly be a damping of consumer and business sentiment due to the government shutdown and elevated anxieties around the various trade conflicts is surely imposing a growing uncertainty tax that will influence outlays on critical capital expenditures and hiring in the near term.
Since the shutdown lasted through the Jan. 12 reference period, the 380,000 furloughed government workers—.23 percent of the total workforce—will be classified as unemployed in the Bureau of Labor Statistics household survey of employment used to conduct the unemployment rate estimate. The 420,000 individuals working without pay will be considered employed in the establishment survey. Hours worked and earnings data are compiled using private sector employees, so this portion of the BLS January payroll estimate will not be affected by the government shutdown.
While roughly 800,000 government workers are now going without paychecks due the shutdown, some 420,000 are currently working without pay and about 380,000 are furloughed; that cohort is permitted to apply for unemployment insurance. Forward-looking investors and firm managers should anticipate a modest increase in initial jobless claims associated with the government shutdown to be reported Thursday morning for the week ended Jan. 12, 2019, and over the next few weeks, should the shutdown be sustained. It is important to note that the number of furloughed workers that file will not show up in the weekly initial jobless claims, but rather, will be reported separately by the U.S. Department of Labor. For example, the DOL in 2013 reported an increase of 70,000 claims linked to furloughed workers. Now, due to variations in state laws, government workers filing unemployment insurance claims may be without benefits for up to three weeks following the application for first-time filers. According to the National Employment Law Project, some furloughed workers may need to pay those claims back once they receive back pay, which will likely damp the number of first time claim filings based on individual expectations about how long the shutdown will endure. In any case, the Labor Department will provide an estimate of direct and indirect first-time filings caused by the shutdown in coming days.
In table 1, we show the number of first-time filings versus the thirteen-week moving average of claims around the number of government shutdowns since 1980. We did not notice a significant jump in claims in any of the shutdowns; thus, we do not anticipate a major increase this week or in coming weeks. However, the nature of this shutdown does appear different from past episodes, including the twenty-one-day shutdown in 1995 and 1996. Both sides are digging in and planning for a possible end of a shutdown at the end of February; it is critical to note that is when food stamps (SNAP) funding runs out and the shutdown would begin to impact the program’s roughly 40 million recipients; the stoppage of food stamps would foster conditions resulting in a shutdown with different outcome than shutdowns of the past.
However, the 1995-1996 shutdown did result in a monthly loss of 13,000 jobs in the January 1996 estimate. The current shutdown started on Dec. 21, marking 25 days at the time of this writing. It has crossed the Jan. 12 reference day that will be used in the monthly sample for the Bureau of Labor Statistics payroll estimate. Thus, it is highly likely that the monthly payroll estimate will deviate from the three-month rolling average of 254,000 and show a net loss of jobs on the month.
We have not altered our economic estimate of the shutdown: we expect economic losses of roughly $1 billion per week with an upside of $1.5 billion, should indirect effects turn out to be larger than currently estimated. Since funding for food stamps is currently secure through the end of February, we are standing by our estimate of a 0.53 percent hit to GDP if that funding is lost.










