How Big Government Got Its Groove Back

In 1996, President Bill Clinton proclaimed that the era of big government was over. It is now clear that the era of the end of big government is over.

The post–World War II social contract--an expanding public safety net, provision of health care and retirement benefits through a substantially unionized private sector, and robust personal savings--is under severe stress. To respond effectively to our long-term challenges, the federal government must command an increased share of gross domestic product and extend its reach in other ways as well. The public sector will be called upon to provide new forms of insurance against economic risks and volatility and to assume more responsibility for health insurance and retirement security. To the extent that markets cannot police themselves or provide reasonable returns for workers, government will have to step in. Through the public mobilization of capital and will, we must supply the public goods--investment in infrastructure, research, and post-secondary education, among others--that we have neglected at our peril. And many millions of Americans will be unable to save for the future without new forms of public encouragement and support.

As well, we will have to construct a new legal and institutional framework that counters the increasing asymmetries of bargaining power that employees in most occupations now experience. While the right to organize and bargain collectively must be aggressively enforced, the kind of union movement that dominated the field from the 1930s through the 1960s may not be adequate for the 21st century. To the extent that it is not, we will need something to supplement it, such as new legal protections for individual workers. For without effective countervailing power, employees will not be able to negotiate for a reasonable share of productivity gains, median wages and earnings will grow slowly, if at all, and the fortunate few at the top will continue to commandeer the fruits of economic growth. At the same time, the private sector will have to do its part to help finance programs for which the public sector assumes increased responsibility. And individuals will have to shoulder more responsibility, in proportion to their means, for their savings and security.

In short, we need nothing less than a new social contract that reorganizes responsibilities among government, individuals, and the private sector. It will take time, experimentation, and political contestation to hammer out its terms.

This would never have been easy, and it is especially challenging now. With large short-term and long-term deficits looming, clearing fiscal space for new initiatives will be difficult at best. And while the public is demanding change, the current administration's woeful performance since 2002 has reduced public trust and confidence in government's ability to produce change.

But however difficult it may be, we must begin the task of reconciling basic moral commitments with stubborn new realities. The alternative to a new contract is no contract--a society in which the strong take what they can and the weak endure what they must.


From June 1982 until November 1984, I served as issues director for Walter Mondale's presidential campaign. The failure of that honorable venture--the last campaign of the New Deal era--propelled me and many others into a period of rethinking that lasted through the remainder of the 1980s and into the early 1990s. We came to believe that the Democratic Party's economic program and governance philosophy reflected an industrial era that was giving way to new modes of production and new technological sources of economic growth. We believed, as well, that the party had come to be viewed as fiscally unreliable and as fixated on redistribution at the expense of broad-based economic growth. These propositions helped shape the economic outlook of the New Democratic movement and the platform on which Bill Clinton ran successfully for president in 1992. And after a slow and controversial start, they contributed to the robust, widely shared growth of the mid- and late 1990s.

At the peak of the Clinton boom a decade ago, optimistic beliefs dominated the landscape. With a regime of more open trade, the United States could compete effectively in the post–Cold War global economy. The market would not only generate wealth but also regulate itself with declining government oversight. Government would become more market-like, substituting incentives and technology-based efficiency for obsolescent command-and-control strategies.

There was evidence to support these beliefs. As productivity surged, wages increased across the board, as did labor force participation. Unemployment fell to record lows, and manufacturing jobs held their own. Inflation remained low and actually fell in the health-care sector. Although the personal-savings rate continued its decades-long decline, the steady increase in housing prices enabled families to build their net worth. Fiscal restraint turned deficits into surpluses, shrank the federal government's share of GDP, and stabilized the value of the dollar in international markets.

From today's vantage point, however, the 1990s appear to have been the proverbial calm before the storm. Although the Bush administration's misguided fiscal and foreign policies have worsened our plight, our problems are structural and long-term, and no simple return to the status-quo ante will resolve them. Most analysts and policy-makers underestimated the impact of huge numbers of new workers in China, India, and the former Soviet Union entering the global market system. International economic forces are limiting wages for most U.S. workers, increasing income inequality, and heightening pressure on the World War II–era system of benefits provided through the private sector. In these circumstances, average families have resorted to record levels of borrowing to maintain purchasing power, driving the savings rate into negative territory for the first time on record and raising personal consumption to an unsustainable 70 percent of GDP. The Bush administration has squandered the resources it could have used to ease the reform of the large entitlement programs. And the back-loaded costs of deregulation are now clear: among them, an epidemic of corporate misconduct and crisis in credit markets, here and abroad.


It is easy to dismiss all this as the latest iteration of the gloom-and-doom narrative that every economic down-cycle generates. But there is evidence that our difficulties are more than cyclical.

Compensation. Recent work by MIT economists Frank Levy and Peter Temin shows that a wedge has been driven between productivity gains and compensation (wages plus health care and fringe benefits) for full-time workers at peak earning age. Since 1980, productivity has increased by 71 percent while median compensation rose by only 19 percent, and 82 percent of personal income gains went to the top 1 percent of the population.

In recent years, these trends have actually worsened. Since 2000, labor productivity has grown by 16 percent, but the median weekly compensation of male high school graduates has declined slightly. To be sure, education matters more than it used to: The difference in median weekly wages of college and high school graduates has just about doubled since 1980. One reason is that the supply of highly educated workers has not kept pace with technological advances. As Claudia Goldin and Lawrence Katz, two economists at Harvard, have recently shown, gains in educational attainment for workers born in 1975 are only one quarter as great as for workers born in 1945. According to David Ellwood at the Kennedy School, if current trends continue, college graduates will constitute only a slightly higher percentage of the work force in 2025 than they do today. The share of male college graduates may actually decline.

But recently, not even a college education is allowing workers to keep up with productivity. Since 2000, compensation for college graduates without further training has grown by only 3 percent. Only men with post-graduate education have managed to keep up with productivity growth. While women have done slightly better, two thirds of them have lagged behind productivity growth, too.

Between 1950 and the early 1980s, median family income rose more rapidly than did incomes at the very top. After sharply rising inequality during the Ronald Reagan and George H.W. Bush presidencies, this trend was briefly interrupted during 1995–2000, when incomes in every quintile rose at roughly the same rate. Since then, all the gains have gone to the top 1 percent of earners. This explains why 2002–2007 is likely to prove the only economic cycle on record in which median family incomes failed to reach, let alone exceed, their peak in the previous cycle.

Below the very top, all income groups have experienced a decline in median incomes during the current economic cycle: 2 percent for the top third, 3 percent for the middle, and 5 percent for the bottom. During this period, however, real median family expenditures have risen across the board. Debt has filled the gap. After rising only slightly faster than the overall economy during the 1990s, household borrowing has soared since 2000 and now totals almost $14 trillion--about 98 percent of GDP, up from 69 percent just seven years ago. Debt service has risen from 11 percent of disposable income in 1994 to more than 14 percent today.

Mortgage debt accounts for most of this increase. While housing prices were rising rapidly, homeowners withdrew hundreds of billions of dollars in home equity to finance current consumption. As housing prices have reversed course, families can no longer use their homes as piggy banks, depressing purchasing power and boosting bankruptcy and foreclosure rates.

Analysts have long debated whether these economic trends are "hollowing out" the middle class. Evidence is accumulating that they are. Between 1970 and 2006, the share of adults living in the middle-income tier declined from 40 percent to 35 percent, while both upper- and lower-income tiers increased. The shrinking of the middle class was especially pronounced among adults aged 18 to 29--from 45 percent to 37 percent--and among 30- to 44-year-olds--47 percent to 38 percent. Only retirees became more likely to live in the middle class.

Private-sector benefits. The private sector is struggling with two safety-net issues--pensions and health care. After World War II, defined-benefit plans ("pensions"), under which firms assumed the responsibility and risk of providing guaranteed retirement incomes to workers, spread rapidly. They reached their peak in the late 1970s, when 62 percent of workers were covered solely by such plans. Today, that figure is down to only 10 percent. In place of pensions, most workers now have defined-contribution plans in which they, rather than their firms directly, face the risks of future economic performance. Many workers withdraw funds from these plans when they change jobs, jeopardizing their retirement security.

Private-sector employers are retreating from employee health insurance as well. Of companies with more than 200 workers, 66 percent had retiree health-care plans in 1988, versus only 33 percent today. In 2000, 69 percent of employers offered coverage to their workers; in 2003, 65 percent; by 2006, that figure had fallen to only 61 percent.

Annual health-care premiums have increased by 87 percent during this decade, more than four times as fast as workers' earnings. The share of earnings consumed by premiums has risen accordingly, pricing many workers out of the market. As a result, although the economy has 4 million more jobs today than in 2000, the number of workers with company-sponsored plans has not increased, and the share covered under such plans has fallen from 64.2 percent to 59.7 percent. Although health coverage under public low-income programs rose from 10.6 percent to 12.9 percent of the population between 2000 and 2006, this was not enough to counterbalance reductions in the private sector. As a result, the number of uninsured rose from 38.4 million to 47 million--from 13.7 percent to 15.8 percent of the total. There is no precedent for such trends during periods of economic recovery and growth.

In recent years, health insurance has been recognized as an issue of economic competitiveness as well as social policy. In 2005, for example, Toyota chose to locate its newest assembly plant in Ontario, Canada, rather than in Alabama. The firm's management cited two reasons: the lower quality of the U.S. work force, and the competitive disadvantage of raising the price of every car to cover employee health care. In areas of our economy subject to international competition, our World War II–era model of employer-provided health insurance--a global outlier--is unlikely to remain viable.

Manufacturing. The past decade has been one of the worst on record for U.S. manufacturing. After three decades of relative stability, manufacturing employment began a slow decline in 1998 that accelerated into a rout in 2001. Nearly 4 million jobs have disappeared, and manufacturing's share of total employment has declined by three percentage points.

It is customary, and not entirely wrong, to point to productivity gains as the principal culprit. Manufacturing investment in research and development accounts for more than half of all U.S. research and development, and it has paid off. Over the past three decades, manufacturing productivity has risen more than 3 percent per year, compared to only 2 percent per year for the rest of the economy. This helps explain how the sector can produce more than 12 percent of the economy's total output with only 10 percent of the work force.

And this trend is accelerating. After growing by 2.7 percent per year from 1977 to 1992, manufacturing productivity grew by 4.7 percent yearly during the past 15 years. In 2002 alone, it rose by 10.6 percent, the largest year-over-year increase ever recorded, followed by two years of 6 percent growth.

The other half of the story is a slowdown in manufacturing output. During the 1991–2000 economic cycle, output just about kept pace with productivity, rising at 4.2 percent per year. In the most recent 2001–2007 cycle, however, output rose only a third as fast, 1.4 percent annually, far behind productivity gains. Massive job loss was the inevitable result. A Congressional Budget Office analysis concludes that for manufacturing, the period since 2001 is more like the severe double-dip recession of the late 1970s and early 1980s than the relatively mild recession of the early 1990s.

U.S. manufacturing is so productive that its jobs offer higher pay and better benefits. An analysis by Robert Scott at the Economic Policy Institute shows that manufacturing workers without a college degree enjoy a 9 percent wage premium over similar workers in other sectors. All other things equal, then, the recent massive job loss in manufacturing has exerted downward pressure on wages. On average, noncollege manufacturing workers who lose jobs paying $16.49 per hour will exchange them for jobs paying only $15.10, and fewer of the replacement jobs will offer affordable health insurance.

While these facts explain the tenor of the 2008 presidential campaign throughout the deindustrializing Midwest, they do not prove that outsourcing and trade are to blame for our manufacturing woes. In fact, the bulk of U.S. overseas investment is in other high-income countries, all of which have more robust regimes of worker protection than we do. And recent studies suggest that this investment generates more--not less--employment in the United States. Productivity increases are driven by technological advances and increased global competition from both high-income countries and from nations, such as China and India, whose low-wage structure exerts downward pressure on the wages of lower-skilled U.S. workers.

While it is tempting to try to insulate ourselves from these forces, that course would prove counterproductive and futile. Instead, we should use public policy to spread the gains of economic growth, create equal opportunity for all, and insure workers against wage and income losses against which they cannot protect themselves.


Rebuilding the social contract is our key domestic challenge during the next generation. While some of its elements will require trial and error, others are tolerably clear.

First, as employer-provided benefits continue to shrink, we will need to create means-tested, market-based alternatives. Some version of the Massachusetts model represents the next step in health insurance; some experts believe that it will function as a way-station on the road to a more unified system of public finance. A system matching individual contributions to savings accounts with means-tested employer and public contributions would help defined-contribution retirement plans fill the void created by the disappearance of traditional pensions. Automatically enrolling employees in such plans unless they affirmatively opt out may yield near-universal participation. If not, a mandatory savings program may well be necessary. Workers below a low-income threshold would receive employer and public matches without having to make automatic contributions from wages.

Second, whatever the relation between globalization and wage stagnation may be, it is unrealistic and unfair to expect workers to tolerate a situation in which they bear the costs of economic change while receiving none of the benefits of economic growth. At the very least, we can index the minimum wage to inflation and expand the Earned Income Tax Credit to cover single workers. And we should create an actuarially sound insurance system--financed by a modest increase in the payroll tax--that would compensate workers for abrupt downward lurches in wages.

Third, we cannot rebuild U.S. manufacturing on the basis of 20th-century industries. Steel mills and automobile plants will never again provide expanding employment opportunities. This does not mean that employment in the sector is fated to continue its decline. One piece of the answer is to focus on the bottlenecks that are harming our economy and society. This means investing much more in 21st-century infrastructure--not only roads and bridges but also airports and light rail. In addition, we should encourage the development of technologies in areas, such as alternative energy and environmental protection, for which there is growing demand, at home and abroad.

And fourth, despite clear market signals that post-secondary education and training are essential, too many young adults are not getting the message. Lower-income students from families without a history of college education are unwilling to incur debt, in part because they do not yet believe in the promised returns on education, and in part because they doubt their capacity to repay loans. This is a market failure that hurts the country as well as the individuals who underinvest in their future. The public sector should follow the lead of a handful of elite private universities, which have begun to make higher education free for working- and middle-class students. This would require means-tested higher education grants that would allow students below a certain income threshold to attend public universities and community colleges and emerge debt-free.

There is no way to build a 21st-century social contract without increasing the size of the federal government well beyond the roughly one-fifth share of GDP that it has averaged in recent decades. At the same time, we cannot expand government indefinitely without reducing long-term economic growth. If we make no policy changes, government will command more than 26 percent of GDP by 2030. If we simply put the programs I have recommended on top of the existing baseline, that figure would increase still more.

This leads to a controversial point, with which I conclude. To be sustainable and pro-growth, we will need a new approach toward the large entitlement programs--especially Medicare and Medicaid--that drive so much of the long-term increase in the federal budget. While universal health insurance would give us a fighting chance to restrain the rate of growth in medical costs, we will probably have to go further. One possibility would be to change our budget procedures. We could establish long-term budgets for public health-care programs and require Congress to close the gap between projections and actual results by either raising revenues or reducing benefits. But whatever we do in this area, we will have to rethink the comfortable assumption that a 21st-century social contract can simply add a new wing to the existing edifice.