KEY TAKEAWAYS
  • During the most recent federal government shutdown from December 2018 to January 2019, 800,000 federal employees were furloughed or worked without pay
  • Federal workers constitute 24.6% of total employment in Washington, DC; more than 4.5% in Maryland, Virginia, Alaska, and Hawaii; and at least 2.5% in eight other states
  • Shutdowns affect private industry and workers compensation by disrupting work under government contracts, interrupting access to government services, and creating uncertainty about more shutdowns in the future
  • The December–January shutdown reduced annualized US gross domestic product (GDP) growth in the first quarter of 2019 by an estimated 0.2%−0.4%
  • Past federal shutdowns have had only temporary impacts on wage and employment growth, but if more frequent and longer shutdowns come to be expected, the impact on investment and employment growth is likely to become more severe

What happens when the federal government shuts down? In this issue of the Quarterly Economics Briefing (QEB), we discuss the size and scope of effects on the economy that result from a shutdown and how they impact private workers compensation. Our primary focus is the most recent shutdown lasting from December 2018 to January 2019, but we also draw insights from previous shutdowns, especially the one in October 2013.

Shutdowns affect economic activity in several ways. Most directly, federal workers and contractors are not paid. Consequently, people and businesses that interact with affected agencies cannot conduct business as usual. And businesses in areas that employ a relatively large percentage of affected workers face reduced demand as furloughed or unpaid workers dial back their spending. More difficult to measure is how uncertainty regarding government functions can affect future planning and investment decisions, slowing economic growth.

There have been 21 federal shutdowns since 1976. Shutdowns occur when Congress fails to pass appropriations bills to cover spending for federal departments and agencies. Employees in unfunded departments are furloughed or may work temporarily without pay if deemed essential. The December−January shutdown was the first since 2013 and only the third since 1979 that lasted more than a week.

How Many People Are Affected by Shutdowns?

There are about two million civilian federal employees in the United States (excluding the Postal Service). In the most recent shutdown, about 40% of federal civilian employees, some 800,000 workers, were affected. This is approximately the same number of workers who were affected during the 16-day 2013 government shutdown. Slightly over half of these affected workers spent at least part of the shutdown working without pay, while others were furloughed.

The concentration of federal workers varies widely by geography. The federal government employs over 8% of civilian workers in the Washington metro area and nearly a quarter in the District of Columbia itself. As shown in Figure 1, the highest concentrations of federal workers at the state level are in Maryland, Virginia, Hawaii, and Alaska. The next-highest concentrations are found in resource-heavy states with relatively low population density: New Mexico, West Virginia, Oklahoma, Montana, and Wyoming. States on both coasts and in the Great Lakes region usually have relatively low concentrations of federal workers.

Shutdowns do not affect states equally, even after accounting for the proportions of federal employees. The recent shutdown disproportionately impacted Washington, DC, not only because of its high concentration of federal employees, but because a higher proportion of its federal employees were affected. Over 70% of DC federal employees work in agencies that were without appropriations.

California has a larger total number of federal workers than any state or the District of Columbia, due primarily to its size, but only 29% of its federal workers were in unfunded agencies. A higher share of federal workers in Western states including Idaho, Montana, Wyoming, and Alaska were affected. This is because the Departments of Agriculture and the Interior, both affected by the recent shutdown, comprise above-average shares of federal employment in those states.

Shutdowns also impact federal contractors. One widely cited estimate from Paul Light, a professor at New York University, states that there are about four million government contract workers. The income impact of a shutdown on contractors can be more important than on federal employees because federal workers are required to receive back pay, but contractors are not. Anecdotal reports suggest that contractors are aware of this possibility and do not expect to receive back pay when furloughed. As with federal employees, some contractors worked through the shutdown, and these workers’ employers may eventually be able to bill for their time. There are also some Congressional efforts to provide back pay for low-wage contractors, although this is uncertain. Contractors are a particularly important group to consider when determining the effect on workers compensation, because unlike federal employees, many of these workers are covered by private workers compensation policies.

The downstream effects of the shutdown depend on two things. First, how much spending is delayed or permanently decreased as a result of workers’ temporary unemployment? And second, how much economic activity is hindered by closing federal agencies?

How Much Do Shutdowns Affect Spending and GDP?

When it comes to shutdowns, personal spending, and GDP, the impact on employees is easier to measure. The first important factor is workers’ ability to smooth out their consumption. Federal employees were guaranteed back pay from the shutdown. For those with sufficient savings, a shutdown may not be too disruptive. However, most American workers, public or private, depend on their regular paychecks to pay the bills. Many of those affected sharply cut spending during the shutdown.

This behavior was the subject of much anecdotal reporting during the shutdown, and it is also a focus of researchers. One study published in 2017 showed that affected workers substantially and rapidly decreased their spending during the 2013 shutdown, even though it lasted barely over two weeks. This result extended across many spending categories, but it was especially strong for discretionary spending such as restaurant meals. Another study of the 2013 shutdown noted that many people delayed debt repayments on mortgages and credit card balances.

Interestingly, the authors of the first study found “the largest drop is during the third week of October, during which the end of the shutdown was announced but workers had not received back pay.” This result demonstrates how fragile many workers’ savings are. It makes sense for workers to cut back when a shutdown begins because they don’t know how long it may last or what the final resolution will look like. But once a shutdown has ended, that uncertainty is resolved. As long as they have enough money to cover expenses, workers should then be willing to resume their normal spending routines. Only if their savings had already started running out would they be expected to cut back even more. A survey of workers (which included contractors) during the recent shutdown found that 62% had depleted most or all of their savings.

How much do all those little spending choices add up to? A summary of estimates produced for the Congressional Research Service found a consensus that the 2013 shutdown lowered GDP growth by around half a percentage point in 2013 Q4, with some of that made up for by increases in early 2014. As a result of the December−January shutdown, the Congressional Budget Office (CBO) estimated that GDP was $3 billion below what it would have been in 2018 Q4 and $8 billion below in 2019 Q1. This translates to 0.2% and 0.4% annualized GDP growth decreases in 2018 Q4 and 2019 Q1, respectively. The CBO further estimated that about $8 billion of these losses will be recouped in future quarters, while $3 billion of this loss is permanent.

We also looked at GDP growth estimates produced by Moody’s Analytics before and after the December−January shutdown. As shown in Table 1, their projections for growth in 2018 Q4 did not change much, but their projections for 2019 Q1 dropped by four-tenths of a percent from the December 10 to January 8 releases. The number released in February dropped another quarter of a percent. For 2019 as a whole, the projection dropped from near 2.9% before the shutdown to 2.7% during and after. The overall 2019 projection did not change between January and February, suggesting Moody’s expected the larger drop estimated for the first quarter to be offset in later quarters.

Of course, it is not easy to disentangle the effects of the shutdown from changes in expectations for any other reason. Moody’s estimated each week of the shutdown dropped quarterly growth by 0.04% through lower government output, meaning that direct effects of the five-week shutdown lowered quarterly GDP by 0.2%. This is smaller than the decline estimated by the CBO and explains less than one-third of their projected decrease from 3.1% to 2.4% for 2019 Q1 between December and February. The projection may also have changed because of business investment, consumer sentiment, private spending, and a variety of other reasons, which themselves may have changed in part because of the shutdown. In the next section, we discuss these indirect effects in more detail.

What Are the Indirect Economic Effects of Shutdowns?

The more difficult-to-measure impacts of the shutdown are the costs imposed by making it more difficult for businesses to conduct operations and by increasing uncertainty. The prior estimates are all of effects that follow directly from reduced spending, but there are a variety of functions government agencies perform that affect private businesses.

In Table 2, we show US federal employment by Cabinet-level agencies, excluding the Department of Defense. The table also shows, whether the department was at least partially unfunded in the December 2018−January 2019 shutdown, selected large employers within each department and whose operations were suspended or affected, and an example of a service affected by the shutdown. The table excludes agencies that are not organized within a Cabinet-level department but may have been affected by the shutdown, such as the Small Business Administration, the EPA, and NASA. The purpose of the table is to give a sense of the workers and services affected, not to make an exhaustive list.

Table 2

N/A = subagencies are not defined in the reporting
Source: United States Office of Personnel Management, September 2018

One commonly discussed service during the December–January shutdown was air travel. If the shutdown decreased business travel, then airlines faced direct pressure on revenues. But their bigger concerns came from how the shutdown affected their core services. Higher absence rates among TSA workers and air traffic controllers led to delays, which imposed time and inconvenience costs on travelers. And there was likely an additional chilling effect from potential travelers who opted not to fly at all because of worries about delays or safety.

Other businesses suffered similar effects when the federal government was not open as usual. For example, businesses did not receive federal loans that they intended to use for capital investments. They could not receive required permits and certifications. At best, this economic activity was delayed. At worst, some businesses in limbo permanently canceled plans or lost opportunities. Even after regular operations are restored, government agencies have to deal with the backlogs created during the shutdown.

From the first moments of the shutdown, firms faced considerable uncertainty about its length. The shutdown was nearly avoided entirely after the Senate passed a continuing resolution on December 19, 2018, but the situation changed the next day when the proposal was not ratified by the House of Representatives. In early January, President Trump said that he hoped the shutdown would last only a few more days, but that it could go on for months or even years. Later in January, executive branch officials asked agencies to list programs that would be affected if the shutdown continued into March or April. While this did not occur, the December–January shutdown lasted for 35 days—twice as long as the 2013 shutdown.

A monthlong delay can be a setback, but worse for workers and businesses is not knowing whether the delay is going last a day or two, a week, several months, or not happen at all. They continue to face uncertainty about whether there will be another shutdown in the future, which is especially relevant to large-scale contractors. Most estimates, including the CBO report cited in the previous section, acknowledge these are important concerns but do not attempt to put a dollar figure on the indirect losses from these costs.

How Significant Are Shutdowns’ Overall Effects on the Economy and Workers Compensation?

In the Q1 2018 QEB, we discussed the economic impacts of hurricanes and implications for workers compensation. While they are obviously different types of events, hurricanes are in some ways a fitting comparison for government shutdowns. In both cases, economic activity is suspended in a way that affects certain people, locations, and industries more than others.

Like natural disasters, there is an immediate human price to shutdowns. Many unpaid workers burn through their savings and increase borrowing. However, on a macro level, a federal shutdown causes only minor and short-term downward pressure on employment, wages, and output. NCCI’s hurricane study found that their effects were contained within one year and were smaller for less severe hurricanes.

The impacts of federal shutdowns show similar patterns. At first, there are negative effects. The expert projections of slowing GDP growth due to the December 2018–January 2019 shutdown match observations from the field. Survey responses collected in the Federal Reserve beige book attributed some of the slowdown in first quarter growth to the shutdown. But within a year or less, spending and GDP growth returns to normal. The Moody’s estimate for 2019 GDP growth dipped two-tenths of a percentage point from December to February, and the CBO’s estimate was even smaller. GDP growth has fluctuated more than one or two tenths of a percentage point in recent years, and meanwhile, employment and wage growth has been relatively steady, as we document in the Review of Current Conditions section each quarter.

Future shutdowns are only likely to have significant impacts on wage and employment growth that affect workers compensation if they last longer or become annual events. The risk is worth monitoring. There are already rising tensions over the federal budgeting process for the next fiscal year. Even if each shutdown’s effects remain small, reducing GDP growth by 0.1% or 0.2% every year will have a much bigger economic effect than the impact of a one-time event. More dangerously, longer or repeated shutdowns could slow down hiring and investment in companies and industries reliant on government services.