The direct effect of the shutdown on measured realgross domestic product (GDP) in the United States occurs through the loss of pay for government workers on furlough and those working without pay, as well as the loss of hours worked for workers on furlough. (Once the shutdown ends, workers will get their back-pay, and if that happens before the end of the first quarter (Q1), then the part of the loss to Q1 real GDP owing to deferred pay will be reversed.) Estimates suggest the direct effects on real GDP from these impacts on government workers run on the order of 0.1% per 10 days or so, which means that if the shutdown lasts for the entire quarter it would directly shave nearly 1% off real GDP.
Then, there are the indirect effects (lost output and wages for private-sector government contractors, multiplier effects on other spending, etc.), which are much harder to calibrate, though they surely grow the longer the shutdown lasts. Estimates here range from a low impact on real GDP of roughly the same order as the direct one from the government workers, to as high as more than double that impact. Even using the more conservative (and we think, more realistic) estimates would suggest the shutdown could shave another 0.1% or more off real GDP growth every week or so, and these effects are likely to grow the longer it lasts. So if the shutdown lasts through the quarter, you could get total impacts of 2% or more taken off real GDP – though the uncertainty surrounding this estimate is substantial. Also, at this point, we have little visibility on when the shutdown will end. When it does, much of this lost output is expected to be restored (though perhaps not quite all). All told, at this point, we have refrained from making major changes to GDP estimates for Q1 (still expect growth of about 2%). But this figure will come down if the shutdown lasts longer, and especially if it starts to look like it will last much of the quarter (in which case there will be little time for the losses to be recouped until Q2).
Finally, the shutdown may impact other data. For starters, some data (e.g., retail sales, business inventories, housing starts, personal consumption expenditure (PCE) prices) are not being released during the shutdown, and will take time to be released once the government reopens, reducing visibility on economic trends. Second, even some data that are being released (e.g., the employment report), may be distorted by the shutdown. The government workers who are working without pay will be counted as employed in nonfarm payrolls in the report to be released on February 1, and the U.S. Bureau of Labor Statistics (BLS) just announced that those who are furloughed (about 380,000) will also be counted as employed (even if they do not receive their back-pay by then). Still, there could be some minor indirect effects on employment if private-sector government contractors and others have been laid off by January 12 (the end of the survey reference week). Like workers laid off in the private sector, furloughed government workers can apply for jobless claims, but are not included in the total headline figures released each week (they are in a separate category; also, once they get back-pay, they are required to return their jobless benefits). The unemployment rate could be boosted by as much as 0.2% if the furloughed workers are captured in the household survey as "unemployed, on temporary layoff," which technically they are. The BLS will likely provide estimates of all this when the report is released. And we can look to surveys of consumer and business confidence to see what impact the shutdown may be having on sentiment.
As for U.S. Federal Reserve (Fed) policymakers, the government shutdown, and the added uncertainty it imparts to an already cloudy outlook, reinforces their inclination to hunker down in "wait-and-see" mode.
Josh Feinman is U.S. Chief Economist at DWS.