Thursday, July 24, 2014

Long-Term Budget Deficits

Much of the sound and fury about U.S. budget deficits involves what should happen in the next year or so. Of course, short-run decisions about red ink do matter, and have a way of bleeding over into long-term decisions. But this focus on the short-term also risks missing the longer-term context of how U.S. government deficits and debt have changed since the Great Recession started in 2007, and where they are headed in the next couple of decades. For this purpose, my go-to starting point is "The 2014
Long-Term Budget Outlook" published  this month by the Congressional Budget Office.

Here's an overview of the basic CBO budget forecast, which shows some of the costs of the current path, but as I will explain, is probably to optimistic. This "extended baseline" forecast is based on existing law, including any ways in which the law requires future changes: for example, if current law requires a tax to be changed a few years in the future, that future change is included in this forecast. This figure shows government spending and revenues as a share of GDP. The gap between them doesn't look large, but remember that with the US GDP now in the range of $17 trillion, a gap of 1 percentage point is a deficit of $170 billion.

I'd draw two lessons from this figure. First, you can see the large jump in accumulated federal debt during the Great Recession, from less than 40% of GDP in 2008 to 74% of GDP in 2014. Federal debt relative to GDP is now at the second-highest level in US history, trailing only the explosion of debt used to finance the fighting of World War II. Second, federal debt in the longer term is projected to rise more slowly in the next few years, but to keep rising to 106% of GDP by 2039--which would be equal to the debt/GDP level in 1946.

Here's a figure showing the government debt/GDP ratio throughout U.S. history. You can see the previous debt/GDP peaks at the Revolutionary War, the Civil War, World War I, the Great Depression, World War II, and the 1980s and early 1990s. According to the CBO, the U.S. government is currently on autopilot to set an unwelcome new record.

The discussion of the effects of large budget deficits often seems to be an argument over the possibility of catastrophic debt overload and the risk of Greek-style or Argentinian-style debt defaults. But arguing over catastrophe misses the real and near-term costs of the higher budget deficits. The CBO lists three of them.

First, the current high level of government debt, and the projections for the next 25 years, mean that the U.S. government lacks fiscal flexibility. Before previous debt spikes, we had started with a relatively low debt/GDP ratio, and so we had room to borrow as needed. But with the debt/GDP ratio already at 74% and rising, we now lack that flexibility. I sometimes say that the U.S. could afford to fight the Great Recession with large budget deficits. But having done so, it wouldn't be nearly as easy to enact a similar deficit boost in the future if economic or foreign policy considerations might seem to warrant it.

Second, the current spending patterns of the U.S. government are starting to crowd out everything except health care, Social Security, and interest payments. The bottom three lines on this graph show rising government spending on major health care programs (like Medicare and Medicaid), on Social Security, and on interest payments on past borrowing. The top dark green line shows spending on everything else the government does, falling steadily as a share of GDP over time. Again, this is the projection based on current law.

Third, large government borrowing means less funding is available for private investment. CBO writes: "Large federal budget deficits over the long term would reduce investment, resulting in lower national income and higher interest rates than would otherwise occur. Increased government borrowing would cause a larger share of the savings potentially available for investment to be used for purchasing government securities, such as Treasury bonds. Those purchases would crowd out investment in capital goods—factories and computers, for example—which makes workers more productive. Because wages are determined mainly by workers’ productivity, the reduction in investment would reduce
wages as well, lessening people’s incentive to work."

These long-run dangers don't mean that an abrupt large reduction in budget deficits should happen immediately, when the economy is still struggling to generate a respectable recovery. But it does mean that we should be thinking seriously about small changes for the near future that will phase into larger effects over the next couple of decades.

I said earlier that this scenario is optimistic. I don't mean that the CBO has biased its baseline estimates: indeed, the report goes through in some detail the underlying projections behind these numbers about productivity and economic growth, health care costs and life expectancy, and interest rates (which affect the costs of financing government borrowing). But the baseline estimate, by law and custom, focuses on what is specifically in the law. This seemingly sensible rule offers a temptation to politicians, who can enact spending cuts or tax increases that aren't scheduled to start for five or 10 years. These proposals make the projected deficits look better over the next five or ten years, and then the policies can be changed later.

Thus, CBO calculates an "alternative fiscal scenario," in which it sets aside some of these spending and tax changes that are scheduled to take effect in five years or ten years or never. Remember that the extended baseline scenario projected that the debt/GDP ratio would be 106% by 2039. In the alternative fiscal scenario, the debt-GDP ratio is projected to reach 183% of GDP by 2039. As the report notes: "CBO’s extended alternative fiscal scenario is based on the assumptions that certain policies that are now in place but are scheduled to change under current law will be continued and that some provisions of law that might be difficult to sustain for a long period will be modified. The scenario, therefore, captures what some analysts might consider to be current policies, as opposed to
current laws."

What changes are assumed in the alternative fiscal scenario? As one example, the category of "Other Non-Interest Spending" in the chart above does not plummet: "Federal noninterest spending apart from that for Social Security, the major health care programs (net of offsetting receipts), and certain refundable tax credits would rise after 2024 to its average as a percentage of GDP during the past two decades—rather than fall significantly below that level, as it does in the extended baseline."

On the tax side, the usual political trick is to have various tax deductions or credits scheduled too expire in the future, which makes the projected deficit appear lower, except that when the time comes for expiration the tax provisions are renewed again. Thus, the baseline revenue estimates rise from 17.6% of GDP in 2014 to 18.3% of GDP by 2024 and 19.4% of GDP by 2034 (and keep rising after that). In the alternative fiscal scenario," tax revenue rises to 18.1% of GDP, which is "slightly higher than the average of 17.4 percent over the past 40 years," notes the CBO--but then tax revenues don't continue to rise above that level.

My own judgement is that the path of future budget deficits in the next decade or so is likely to lean toward the alternative fiscal scenario. But long before we reach a debt/GDP ratio of 183%, something is going to give. I don't know what will change. But as an old-school economist named Herb Stein used to say, "If something can't go on, it won't."