America’s state will expand whoever wins

Image of Lawrence Summers

With the selection of Paul Ryan as the Republican vice-presidential candidate, it is clear both political parties agree that the central issue in the presidential election will be the scale and scope of government involvement in the US economy. There will be disagreement over what constituted “normal” levels of spending in the past and indeed over what constitutes “spending”. But there is a widespread view in both parties that it is feasible and desirable that in the future the federal government will be no larger as a share of the overall economy than it has been historically.

Unfortunately, this aspiration is unlikely to be achieved. Even preserving the amount of government functions the US had before the financial crisis will require substantial increases in the share of the economy devoted to the public sector. This is the case for several structural reasons.

First, demographic change will greatly expand federal outlays unless politicians decide to degrade the level of protection traditionally provided to the elderly. Between Social Security, Medicare and Medicaid and other smaller programmes, about 32 per cent of the US federal budget, or about 7.7 per cent of gross domestic product, is devoted to supporting those aged over 65. The ratio of this age group to those of working age will increase from 1:4.6 to 1:2.7 over the next generation, implying a rise in federal spending of 5.6 percentage points of GDP, if no other adjustments are made. True, as Americans’ health and life expectancy improve, it may be appropriate to revise upward the assumed retirement age. However, it will be unlikely to counteract the expected 34 per cent increase over the next generation in the share of the population who will be within 15 years of estimated life expectancy.

Second, the accumulation of more debt and a return to normal interest rates will raise the share of federal spending devoted to interest payments. In 2007, before the financial crisis, federal debt held by the public was equivalent to 36.3 per cent of GDP. On a very optimistic view, where recommendations such as those of the National Commission on Fiscal Responsibility and Reform (the Bowles-Simpson commission) are implemented, net debt held by the public will nearly double to 65 per cent of GDP by 2020. This implies that the federal government’s outlays to service its debt will rise from 1.7 per cent of GDP in 2007 to 3.2 per cent of GDP in 2020.

Third, increases in the price of what the federal government buys relative to what the private sector buys will inevitably increase the cost of state involvement in the economy. Since the early 1980s the price of hospital care and higher education has risen fivefold relative to the price of cars and clothing and more than 100-fold relative to the price of televisions. Similarly, the complexity and hence the cost of everything from cutting-edge scientific research to regulating banks rises faster than overall inflation. These trends reflect long-running trends in globalisation and technology. They imply that if government is to continue providing the same level of these services, government spending as a share of the economy has to rise, by at least 3 per cent of GDP.

Fourth, several methods that have been used to repress the deficit will soon be found to be unsustainable. Federal pension liabilities and the deferred maintenance of federal infrastructure are two examples.

Meanwhile, there is a steady decline in the fraction of tax returns that are audited and there is evidence of growing tax non-compliance. Both are a reflection of unsustainable cuts in spending. And on almost any reasonable view of the state’s responsibility, large increases in inequality such as those we have observed in recent years should call forth increased government activity. All of these factors suggest the likelihood of increased pressure on federal budgets over the years ahead.

There are ways in which federal spending can be reduced. Defence spending, which now represents 4.7 per cent of GDP (its average level over the past 40 years) could be reduced significantly. On the other hand, the fact that in a dangerous world our military is badly stretched by sustained deployments that are far smaller than even the first Iraq war suggests there is little ground for confidence that the Pentagon budget will be cut dramatically.

In some areas technology could greatly reduce government costs but it is important to recognise that by far the largest parts of the federal budget involve cash or in-kind transfers. These parts are far less susceptible to productivity-enhancing technologies than areas that involve the production of goods or services. There is scope for the elimination of outdated or duplicate programmes but efforts to identify waste, fraud and abuse invariably come up with only negligible savings.

For the next three months the US will debate the merits of growing versus shrinking government. But for the next three decades it will confront the reality that major structural changes in the economy will compel an increase in the public sector’s fraction of the total economy unless there is a substantial scaling down in the functions that the federal government has long performed. How government can best prepare for the pressures that will come, and how greater revenues can be mobilised without damaging the economy, are the great economic questions for the next generation.

The writer is Charles W. Eliot university professor at Harvard

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.