Pressure to Grow

Paul C. Light

After years of downsizing, the federal government is poised for expansion again. Directly or indirectly responsible for about one-eighth of the jobs in the nation’s economy, the government is about to grow, whether the electorate chooses George W. Bush or Al Gore as the next president. Both candidates are making promises that can only be kept by adding to the true size of government, as measured not only by the size of the federal civil service but also by the number of employees working indirectly for Uncle Sam under contracts and grants. Most notably, Bush and Gore both have vowed to increase military modernization.


Any increase in federally generated employment would pose a political problem for either candidate. After all, Gore has taken credit for shrinking government to a size not seen since the 1960s, while Bush has signaled his intention to continue the trend by cutting an additional 40,000 positions from federal managerial ranks. It is quite possible that Bush or Gore would seek to mask the growth in government-related jobs by using trends in the civil service head count alone as the measure of the government’s size. That, after all, is one of the yardsticks President Clinton used in 1996, when he declared the era of big government was over.

But Clinton’s measure was imperfect. It failed to account for the huge number of “off-budget” employees doing the government’s work. A more accurate count would have to include not just the civil service, but also the uniformed military, postal workers, and the contractor, grantee, and state and local government workforce needed to carry out the federal mission. This fuller accounting puts the total number of jobs attributable to the federal government at 16.8 million in 1999.

Absent a radical downsizing of the federal government’s mission, which neither Bush nor Gore has proposed, or extraordinary productivity growth rooted in some yet-to-be discovered technology, the total federal workforce is on a growth trajectory. Since 1990, employment to fulfill the missions of the Defense and Energy departments has shrunk by 2.1 million jobs, from 8.4 million to 6.3 million in 1999. But at the same time, a surge in program activity outside Energy and Defense during the past decade produced more than 400,000 new jobs. Employment related to domestic programs grew from 5.5 million in 1990 to 5.9 million in 1999.

Like the 1990, 1993 and 1996 data published by the Brookings Institution and Government Executive magazine two years ago, the 1999 numbers combine two different measures of federal employment: (1) head counts of full-time-equivalent civilian, military and postal employees, and (2) estimates of total employment generated by federal contracts and grants. Estimated numbers of state and local government jobs, generated by mandates under federal law, have been added to these counts. The federal head counts can be found in the historical tables of the fiscal year 2001 federal budget, while the contractor and grantee job estimates were generated by Virginia-based Eagle Eye Publishers Inc., using data contained in the Federal Procurement Data System and Federal Assistance Awards Data System. Along the way, Eagle Eye also updated its 1990, 1993 and 1996 calculations to account for updated information in the Federal Procurement Data System, which had the net effect of increasing head counts for all three years.

The true size of government estimate also includes a first-ever count of jobs in state and local governments generated by mandates under federal laws. The estimate was not updated for 1999, but it surely has not gone down since 1996, when it was produced using a survey of state and local government employees.

These data (see figure 1) reveal five broad conclusions that both presidential candidates should consider as they make promises about the future of government:

  • The true size of government in 1999 was roughly 17 million employees, about eight times larger than the civil service workforce.
  • Nevertheless, the true size of government in 1999 was still down nearly 1 million jobs from 1993, when the Clinton administration took office, and down almost 1.8 million jobs from 1990.
  • The end of the Cold War, not reinventing government, supplied most of those cuts.
  • Except for huge cuts at Defense and Energy, and lesser cuts at four other federal agencies, the true size of government would have increased over the past decade.
  • With the post-Cold War downsizing coming to a close, the true size of government will almost certainly grow.

True Size of Government, 1999

The most readily measurable government-related employment is, of course, the ranks of full-time federal civil servants. Employees who owe their jobs to federal contracts and grants also can be counted, though less directly. Adding all these workers together brings a total of 12.2 million full-time-equivalent employees for 1999, comprising 900,000 postal workers, 1.4 million military personnel, 1.8 million civil servants, 2.5 million grantee employees and 5.6 million contractor employees. When state and local government employees working under federal mandates are added to the total, the true size of the federal government reached 16.8 million last year, down from 17.2 million in 1996.


While 16.8 million is an eye-catching figure given the political rhetoric suggesting a smaller government, it is merely one of several perfectly reasonable ways to measure the size of the federal government’s mission. The nation simply could not defend itself without the Defense Department’s 3.7 million contract employees, nor could it maintain its roads and highways without the Transportation Department’s 800,000 grant employees. Given the growing list of labor-intensive campaign promises, whether for more police officers and teachers or prescription drug coverage, it is hard to imagine further cuts in the total number of employees who must do the work. Congress and the president can move the bodies from one category to another, from contracts to grants, civil servants to consultants, but somebody has to do the work.

A Decade of Shrinking

As big as the federal government might appear today, it is much smaller than it was just 10 years ago (see figure 2). Focusing just on civil service, military, contractor and grantee jobs (leaving aside state and local government mandated jobs), the true size of government has fallen by nearly 1.8 million jobs since 1990, from 14 million to 12.2 million. Of the total reduction, roughly 370,000 lost jobs came from the civil service, 670,000 came from the ranks of uniformed military personnel and 900,000 came from government contractors. During the same period, however, postal employment grew by 55,000 jobs and grantee employment rose by about 110,000, yielding a net decrease of 1.8 million overall. Unfortunately, it is impossible to go back in time to interview state and local employees to estimate changes in the number of jobs generated by federal mandates. Almost 1 million jobs, or 56 percent of the total reduction, were cut during the Clinton administration, while 800,000 jobs, or 44 percent, disappeared when George Bush was president.

Gore is free, therefore, to lay claim to having been part of a dramatic, even radical, downsizing in the federal workforce. He is most certainly well within his bragging rights to say that the federal workforce is smaller today than even in 1984, when the civil service, military, contractor and grantee workforce totaled roughly 14 million. Moreover, there is no doubt that the first three years of Gore’s reinventing campaign produced a modest reduction in civil service and contract employment outside Defense and Energy.

Where Gore clearly errs is claiming that his initiative is somehow responsible for rolling back the size of government to the smallest level since the 1960s. He implied as much on NBC’s “Meet the Press” in July, saying, “I’ve been in charge of the reinventing government project, and we’ve reduced the size of government to the smallest level since John Kennedy was president.”

The claim is only defensible if he uses only the civil service head count, including the Defense Department, and only if he can show that reinventing government was somehow responsible for the end of the Cold War. Subtract Defense from the 1960 figures, and the civil service grew from 761,000 employees in 1960 to roughly 1.2 million in 1999. Reinventing government has done many good things for the federal government, including a significant increase in customer satisfaction. But it simply did not pre-date the end of the Cold War.

In addition, Gore’s claim discounts the rapid expansion of the contractor and grantee workforce in the wake of the Cold War, Kennedy’s New Frontier, and Johnson’s Great Society. NASA grew from 23,000 civil servants and just 3,500 contract employees in 1960 to 18,500 civil servants and 388,000 contract employees in 1999, while the Environmental Protection Agency grew from a few thousand employees scattered across a half dozen departments and agencies in 1970 to an independent agency with more than 130,000 civil service, contract and grant employees in 1999. The story is the same across dozens of departments and agencies, including the Housing and Urban Development Department, which grew to almost 50,000 civil service, contractor and grant employees from its creation in 1965 to 1999, and the Energy Department, which crested in 1990 at 837,000 civil service, contract and grant employees.

Investing the Peace Dividend

The big reductions of the last decade came in the national security workforce as a result of the Cold War’s end. The Defense and Energy departments together accounted for nine out of every 10 jobs cut during the 1990s. Defense reduced its uniformed, civil service, grantee and contractor workforce by 1.7 million, while Energy cut its true size by 407,000 jobs. In all, the defense workforce was cut by a fifth, falling from a total workforce (including contract and grant jobs) of 7.5 million in 1990 to 5.8 million in 1999 as base closings mounted and demobilization continued, while Energy was reduced by more than half, dropping from 837,000 jobs to 430,000 as nuclear weapon manufacturing dwindled and plant cleanups advanced (see figure 3).

Although these reductions shared a common origin in the fall of the Berlin Wall, the two agencies produced a very different inventory of savings. All but a fraction of the Energy Department’s cuts came in service contracts, while the vast majority of Defense shrinkage came from manufacturing employment. From 1990 to 1999, the Defense manufacturing workforce plummeted from a little more than 2 million employees to 1.3 million, while its service workforce actually increased by roughly 20,000 jobs.

While Defense and Energy jobs were vanishing during the 1990s, Congress and the president invested the resulting peace dividend in a growing domestic mission. Big winners were the Transportation Department, whose true size increased by 222,600 jobs; the Justice Department, up by 116,500 jobs; the General Services Administration, up by 89,600 jobs, many of which were generated under government-wide contracts on behalf of other agencies; and Treasury, up by 58,500 jobs.

Transportation had no choice but to grow, in response to two gigantic highway bills passed during the decade. Congress stuffed the 1991 Intermodal Surface Transportation Efficiency Act with $6.2 billion in 528 separate projects, and the 1998 Building Efficient Surface Transportation and Equity Act with $9.3 billion more in 1,850 projects.

Again, this is not to suggest that government is somehow too big in an absolute sense. Congress and the president have every constitutional right–indeed they have a responsibility–to purchase the labor needed to faithfully execute the laws. Rather, the purpose of this paper is to identify the problem created by using the civil service head count alone to claim that government was somehow shrinking when most agencies were actually growing. Despite deep cuts at Defense and Energy, most federal agencies did remarkably well during the 1990s, including six whose reductions in civil service employment were more than offset by gains in contract employment–the Agriculture, Housing and Urban Development, and Labor departments, as well as the Agency for International Development, General Services Administration and Office of Personnel Management (OPM).

OPM, for example, cut its civil service workforce by 3,105 employees, or 45 percent, during the Clinton years. But most of those workers were immediately hired back under a five-year, sole-source contract with U.S. Investigations Services Inc., the employee-owned firm created by the same civil servants whose unit had just been privatized. Although the same people were doing the same mission for the same agency, they were gone from the civil service head count nonetheless, making OPM look smaller than it actually is.

Governor Bush cannot criticize Gore for increasing the domestic workforce, however, without criticizing his father, too. Almost three-quarters of the job growth on the domestic side actually occurred between 1990 and 1993 as President Bush cut 1 million jobs from Defense and Energy, but added 240,000 new jobs elsewhere in the federal government. From 1993 to 1999, the Clinton administration added another 90,000 jobs to the domestic side of government.

Harbingers of Growth

The post-Cold War downsizing of the nation’s defense capabilities is almost certainly at an end. Indeed, Defense budgets have been rising slowly since fiscal 1998, after a long decline that began after the Reagan-era buildup ended in the mid-1980s. President Clinton’s proposed Defense budget for fiscal 2001, which begins in October, represented an inflation-adjusted increase of 5 percent over the 1998 level. The proposed real increase over the year 2000 level was smaller, of course–just 1.3 percent–and no further increases were projected later in the decade. Congress this summer increased the Clinton Defense budget by a little more than 1 percent. And more increases are on the horizon, signaled both by Gore and by Bush during their campaigns.

Although a new round of base closings may well begin next year, these adjustments in military infrastructure will almost certainly not be accompanied by further reductions in force structure. And defense industry employment is sure to grow, pushed up by an increase in modernization spending from $48 billion two years ago to $60 billion this year. (The defense industry absorbed a job cut of 718,000 during the reduction in government’s real size that lasted most of the 1990s.)

Manufacturing employment will not be the only area of growth. Indeed, the trend in many agencies to contract for services is likely to continue its upward trajectory. Defense services contracting did not shrink in tandem with manufacturing contracting, and on a government-wide basis, contracts for services remained essentially unchanged during the 1990s, falling by a mere 3 percent over the decade. Many of the contracted services are essential to federal mission delivery. Indeed, many agencies could not even let contracts for goods or services they need without contractors to help them. In March, for example, the Defense inspector general noted that the department’s acquisition workforce had been halved since 1990, giving rise to increasing dependence on private contractors for issuing contracts, including the preparation of acquisition plans, source-selection plans and other pre-proposal source-selection documentation, contract administration, and quality assurance. The inspector general noted that before downsizing, civil servants had performed many of these tasks.

The story is the same across government. With labor markets extremely tight, federal recruitment systems failing and talented Americans turning away from government as a career choice, most agencies have little choice but to ask private contractors to supply the workers they need. Acquisition workers are just one example. Computer programmers, policy analysts, accountants, trainers, speech writers, security investigators, strategic planners, and a host of other employees with special skills are also in such great demand that the government is having trouble competing. The federal government simply cannot succeed without its shadow workforce.

Questions for Debate

There is still time to ask Bush and Gore hard questions about the multi-sectored, highly competitive public service workforce revealed in these data. Are taxpayers getting fair value for the amount of money being spent on contractors and grantees? Should the federal government compensate state and local governments for workers needed to carry out federal mandates? How can the federal government protect its institutional memory as contractors and grantees become more aggressive in the war for talent? Does the term “inherently governmental” have any meaning left in an era when every imaginable service is commercially available at the right price?

These questions cannot be answered by pretending that civil servants are the only federal employees that count. The federal government has a 17-million member workforce precisely because it has a 17-million employee mission.

Ultimately, therefore, the most important question of all is not how the federal government got to be so big, but whether it has the human capital to provide the high performance that Bush and Gore both promise. It could be that a 17-million member workforce is too small for the federal mission, or that a civil service of 1.8 million employees is still too big, or even that the federal government needs a vast expansion of its acquisition workforce to assure accountability and value. The only way to know is to look at the entire workforce, not just the portion that makes government look smaller than it needs to be to accomplish a mission that most Americans want done.