Not So Big

s the federal government today bigger than it was a generation ago? That's the clear impression left by Government Executive's January cover story, "The True Size of Government." But a closer look suggests that the role and reach of Washington have in fact diminished since the 1960s.

To be sure, author Paul Light makes several important points, most notably that the scope of the federal government remains vast, and that the numbers employed directly in civil service and the military are but the tip of the federal iceberg.

He argues further "that the federal government is much bigger, not smaller than it was 30 years ago," implying that government is much bigger because of the growth of what he terms the "shadow of government," millions of people employed by states, localities, nonprofit organizations, universities and businesses that are funded, directed and guided by the federal government.

As we debate the role, scale and scope of government, getting the facts on the table can only help. But a few more facts are needed to put Light's analysis in context. And his conclusion-that the federal government is much bigger today than 20 or 30 years ago is suspect at best, and probably just plain wrong.

Equally important, the long debate about big government has not really been about the magnitude of government employment. Nor do I believe (and I hope that Paul Light is not suggesting) that this is where the debate should go, for the following reasons:

• Whom and how many people the federal government employs are poor surrogates-incomplete at best, misleading in some respects-for its reach, scope and effect in American society.
• The broad outlines of federal employment are consistent with our constitutional framework and 200 years of history.
• The reach of the federal government has in fact been reduced, at least in part in response to the big government debate. Arguably its size, scope and reach is smaller today than a generation ago, not bigger as Light claims.

Defense and the Mail

The majority of federal employees are in two agencies-the U.S. Postal Service and the Defense Department. It has been this way for 200 years.

Data from the late 1700s are spare, but it is clear that the military outnumbered the civil service. In 1816, three-quarters of the civil service jobs were in the Post Office, which remained the largest civilian employer throughout the 19th century.

In 1929, before the New Deal, the Post Office employed 295,000 people, just over half of all federal civilian employment. The War and Navy departments had about 100,000 civilian employees, so the Post Office and the Pentagon accounted for about two-thirds of the civil service, not far off their proportion of today's federal workforce.

There are 850,000 full- and part-time Postal Service employees. The Defense Department employed 1.5 million in the uniformed military and just under half of the 1.9 million federal civil servants in 1996. These two organizations account for three-quarters of the federal workforce. They also account for the major share of the federal government's purchases of goods and services, an important element in Light's shadow of government.

The Defense Department has been reduced in size and scope considerably over the past decade. Both civilian and military employment are down (nearly 700,000 jobs), the Defense budget is $100 billion less in real terms than its Reagan-era, Cold War peak, and DoD's purchases of goods and services have been reduced dramatically. There has been a real peace dividend.

The Postal Service is a different story. As Light notes, this side of government is growing. It accounts for most of the increase in federal civilian employment he cites. But a little context may help. When the Postal Service was created in 1970, it employed 740,000 people. Today the figure is about 850,000. In this nearly 30-year period, the postal workload-the volume and weight of mail delivered-has more than doubled and employment is up 15 percent.

The Postal Service, a public corporation, made a $400 million profit last year. Over several preceding years USPS profits ran into the billions. And service is much improved. Today, 92 percent of first-class mail sent to local addresses in any given area arrives the next day. That's up from between 50 percent and 60 percent five years ago.

Surely, this is big government. Yes and no. My colleagues, Joseph Nye, David King and Philip Zelikow, looked closely at this issue in their study "Why Americans Hate Government." They discovered that the Postal Service and the military establishment, the two biggest agencies of the federal government, also are the two elements respected and trusted to do a good job by a substantial majority of the American people. But these two organizations are the wrong targets. The attack on big government, by and large, is not aimed at federal employment.

The Historical Context

For many years, one of the clearest statements of federal policy was found in OMB circular A-109, the policy governing the purchase of goods and services. It said simply, "Government should not compete with its citizens." The OMB circular expressed an enduring theme in American government: We strongly prefer to rely on others-the private and nonprofit sectors, states and localities-to carry out our policies.

The founders had a profound distrust of strong central authority. They established a system that placed explicit limits on the central government, reserving for the states and the people all powers not specifically enumerated and assigned to the federal government.

Our distrust of central authority helps explain why we didn't create national police, a national railroad, a state telegraph or telecommunication authority, a state airline, a national water board, a national electricity authority and so on.

When the federal government established policies and appropriated funds to build interstate highways, hospitals, low-income housing, mass transit, water and sewer programs, welfare, Medicare, Medicaid and much more, distrust of central control helps explain why the programs were and continue to be administered through states, cities and the varied private and nonprofit entities that make up Light's shadow.

This observation is entirely consistent with the claim that the real issues (and millions of personnel) lurk in the shadow of government, unburdened and unprotected from federal rules and regulations.

Is It Bigger?

If we compare the federal government today with Revolutionary times or the 19th century, it is clear that the government is bigger, by every measure. But the right comparison may be a bit more complicated. Do we want to know only if government has grown in some absolute sense?

Yes, the Postal Service is bigger today (in terms of employees) than in 1930. But the U.S. population in 1999 is more than twice that of 1930, and the volume of mail handled by the Postal Service is more than seven times bigger.

Shouldn't we take into account the fact that the country today has a larger population, a larger economy, and so on? A bigger population may need more schools and teachers, more roads and more mail carriers. In 1962, the United States had a population of 180 million and 5 million people employed in the federal government. Today, the country has a population of 270 million people (50 percent more) and 4.3 million federal employees (about 15 percent fewer).

We need to examine the relationship between measures of the size of government and the size of the country, or indicators of the scale of government's tasks and how these have changed over time. Rather than making a comparison between lemons and grapefruit, we ought to ask whether the federal government is taking more or fewer resources (jobs, taxes, etc.) relative to the size of the country. Paul Light is correct that the employment figures don't tell the whole story.

How can we assess whether the federal government is bigger in relative terms today than 20 or 30 years ago? We can follow the money. Today, as has been the case for the past half century, roughly one dollar in five in the American economy passes through federal coffers. The world's largest commercial firm is not one-tenth this size.

What and how much is the federal government spending? In 1962, all federal expenditures accounted for 18.8 percent of the gross domestic product (GDP). In 1998, the figure was 20.1 percent. Spending appears to be up.

But look more closely. In the 1960s, 1970s and 1980s, federal government spending climbed steadily, rising to a peak of 23.6 percent of GDP in 1983. Since then, over the past 15 years, the federal claim on the nation's economic resources has proportionally declined. At 20.1 percent of GDP, federal spending today is down 3.5 percent of GDP from the mid-1980s peak. In today's $8.5 trillion economy, that decline amounts to about $300 billion per year.

What else has happened to spending? Federal spending can be usefully divided into three categories:

• Discretionary spending, the policies and programs that generate the payrolls, grants and contracts that make up most of what the government actually does. It encompasses everything the Defense Department does, along with the Coast Guard, the Federal Aviation Administration, the Environmental Protection Agency, the IRS and the FBI, national parks and forests, the National Institutes of Health, housing assistance, grants to states for environmental programs, education, mass transit and so on. It is these programs and the money they spend that generate the direct and indirect shadow employment Light cites.
• Entitlement payments, primarily to individuals, that include military and civil service pensions, Social Security, welfare, Medicare and Medicaid, and unemployment insurance. These are transfers of money through the federal government.
• Interest payments on the national debt, paid out to bondholders. Like most entitlements, these are payments to individuals and are outside the shadow of government.

In 1962, discretionary federal spending amounted to 12.7 percent of GDP. The largest share of discretionary spending was in Defense, which alone accounted for 9.3 percent of GDP. About 6 percent of GDP went to entitlements. In 1997, discretionary federal spending was down to 6.9 percent of GDP. Defense remained the largest share of discretionary spending, at 3.4 percent of GDP. About 11 percent of GDP went to entitlements.

Spending Less, Doing Less

The conclusion is clear. Discretionary federal spending-the programs that generate direct and indirect federal employment-is a much smaller share of the nation's economic activity than it was 20 or 30 years ago. Since the government has proportionally less to spend, it isn't surprising that Light's figures show that, in real terms, the government's purchases of goods and services have declined by more than 10 percent over the past decade. One may rightly assume that associated employment is down as well. The shadow is fading.

One other area deserves examination. State and local government employment is growing today, and it has been growing steadily. Today, state and local government employment is about 16.6 million, up from 13.3 million in 1980. Despite the increase, when the changes at all levels of government are added together, total public employment is a smaller share of the total U.S. workforce than a generation ago.

In 1980, federal grants to states and localities were $91 billion. By 1997, the figure had jumped to $225 billion, an increase of $134 billion over 17 years. More money is arriving from the federal government, which some believe is fueling a large share of the growth. They are mistaken.

Look more closely. First, much of the growth in state and local employment is in areas such as education (which accounts for more than half of all state and local employment) and public safety (police, fire and corrections), where federal contributions are quite limited. Second, most of the increase in federal grants from 1980 to 1997 is accounted for by the rapid growth of Medicaid, an entitlement (welfare) program administered by the states and jointly funded by the federal and state governments. These funds underwrite health insurance for the poor.

Indeed, apart from Medicaid, the real (constant dollar) value of federal grants flowing to states and localities today is less than in 1980. As one looks more closely at what has happened to intergovernmental grants since 1980, there have been substantial reductions in categories such as environmental programs, employment training, mass transit, social service block grants and revenue sharing. In real terms, states and localities are receiving less federal money today than a generation ago.

So what are the facts?

• Direct federal employment is down.
• Federal spending is a lower share of the nation's resources than it has been in more than a generation.
• Within that lower federal share, there is far less discretionary money for indirect employment through grants and contracts.
• The real value of federal grants and contracts is down.
• Total public employment (at all levels of government) is a smaller share of the workforce than a generation ago.
• The conclusion is pretty clear: The federal government-measured by direct and indirect employment, by the dollars it spends and how it spends them-is doing less today than a generation ago.

The comparison does say something about the durability of the basic American approach to government: We disdain central bureaucracy and are unwilling, even when we decide to spend national resources on important problems, to create large federal bureaucracies to administer policies and programs.

What's the Beef?

The first premise of Light's article is that the focus on employment is largely misplaced, not irrelevant. Indeed, it is helpful to get the facts out, and the fact is that the federal government, by most measures, is smaller than it has been in a generation.

The second premise is that the size of the federal government isn't what has drawn the ire of government's critics. Arguably, policy initiatives and actions taken by federal agencies are the source of frustration voiced by many critics. Consider:

• In the late 1960s and 1970s, the environmental movement generated support for a much expanded federal role in environmental regulation, and massive, often unwelcome, requirements for change in business practices.
• In the 1970s, activists fueled aggressive federal intervention in almost every workplace in areas such as employee safety and health and in commercial practices such as advertising. Indeed, the stodgy Washington Post was stirred by one proposed federal initiative to label the Federal Trade Commission "a national nanny."
• In the American West, the demands of fast growing populations ran headlong into federal ownership and control of rangeland, timber, water, mineral and recreational resources; the federal government was caught between the needs of development and environmental activism.

It was these initiatives and others, sometimes abetted by arrogant or insensitive bureaucratic behavior, that led to the election of Jimmy Carter, Ronald Reagan and Bill Clinton, who all made a strong rhetorical point of the need to reorganize, rein in, reform and reinvent the federal government.

The election of these candidates and hundreds of other politicians at national, state and local levels over the past 20 years has had a profound effect on the role, scope and approach of American government. The policies they have put in place have limited the size and reach of the federal government, and increased government's sensitivity and responsiveness to citizen concerns. But it is yet to be seen whether these changes have been sufficient to re-establish trust in the federal government, or whether we have reached an end to government bashing and rhetorical attacks on big government.

Peter Zimmerman is senior associate dean of the John F. Kennedy School of Government at Harvard University.