The costs of the shutdown

by: Peter de Bruin

US government partially shuts down

In today’s daily, we assess the likely implications of the government’s first shutdown since 1996. As of midnight, with the start of the new fiscal year, the current continuing resolution expired. With no budget in place, all non-essential government services will be forced to stop operating. For instance, all museums and parks will close. It no longer will be possible to apply for a visa or passport, while the Education Department will close, and most likely this Friday’s labour market report will not be published. In contrast, essential services, for instance related to national security and medical care will continue to operate. As a result of the shutdown, hundreds of thousands of employees would be put on furlough. Although there is no guarantee that these furloughed employees will continue to receive their salary, federal employees who had been furloughed during past shutdowns generally were paid, albeit retroactively. Still, with large parts of the government having grinded to a halt, there will be a more permanent loss in economic output, as furloughed employees no longer fulfil their duties.

The longer the shutdown, the more significant the damage…

The good news is that a shutdown of a couple of days will have negligible effects on the economy. That said, if the shutdown lasts longer, its effect tend to become larger. For instance, during the shutdown in 1995-1996 when non-essential government services closed for 14 working days in the fourth quarter of 1995, federal government spending declined by 11.6% QoQ saar, and deducted 0.9 percentage points off quarterly GDP growth. Indeed, based on the 1995-1996 shutdown, we estimate that every week non-essential government services are not allowed to operate would cost the US economy around $7.5 billion. According to our estimates, this would mean that a two week shut-down would shave around 0.5 percentage points off quarterly GDP growth .

…though there is some room for payback

This more or less accounting approach to estimate the likely cost of a government shutdown does not take into consideration that demand for government services might be larger than normally is the case when non-essential government services reopen for business. Indeed, in contrast to the shutdown in 1995 (when the government closed its non-essential services predominantly during the final weeks of the quarter), the current shutdown occurs in the beginning of the quarter. The ultimate effects on the economy are therefore likely to be less, as there might be a catch-up effect. Indeed, although federal spending figures in the national accounts tend to be a bit erratic from time to time, we think that it is no coincidence that in the first and second quarter of 1996 we saw firm rebounds in federal consumption and investment. All this suggests that, while disruptive in the short-run, the longer term macroeconomic effects of the shutdown tend to be less significant.

Softer indicators also quickly rebound

A similar story holds with respect to survey indicators. Granted, both consumer confidence and the ISM manufacturing index declined in the period around the previous government shutdown, both indicators quickly rebounded when the government resumed full operations.

Short shutdown still most likely

Although there is a risk that it may take longer to solve the current situation in Washington, our base case scenario continues to expect that the shutdown will be relatively brief. This is because we think that the event of the shutdown will put pressure on both Democrats and Republicans to come up with a compromise (most likely by kicking the can down the road). However, given that the second deadline – in the shape of the debt ceiling – is also approaching, we could well see more financial market unrest in coming days.