A key reason to be concerned about high debt levels is very simple– you’re going to be stuck with the bill for the interest payments for the rest of your life.
According to the Congressional Budget Office, net federal debt held by the public (which leaves out the sums owed to Social Security and other trust funds) was $11.3 trillion as of the end of 2012, or 73% of GDP. Federal net interest expense for 2012 came to $220 B, for an average interest rate paid on outstanding debt of 220/11300 = 1.9%. The graph below repeats that same calculation for each of the last 40 years. Federal interest expense as a percent of debt owed is currently lower than it’s been at any time during this period. For example, the average implied rate over 2000-2009 was 4.5% and the average over 1990-1999 was 6.6%.
That implied rate calculation tracks the actual interest rate on a 10-year Treasury bond fairly closely.
And that makes it pretty simple to calculate what would happen to the government’s total interest expenses if interest rates were to rise. For example, if today the government had to pay the same average rate that was seen over 2000-2009, interest expense would come to (0.045)(11,300) = $508 B every year, even if the level of debt stays exactly where it was as of the end of last year ($11,300 B). With an average interest rate like we saw in the 1990s, the interest cost would be (0.066)(11,300) = $746 B. For comparison, total federal discretionary spending on all categories other than defense came to $615 B in 2012, and the entire defense budget was $670 B.
The other key parameter for assessing the interest burden is the economic growth rate. The red line in the graph below shows the average annual growth rate of U.S. nominal GDP for the 5 years prior to each indicated quarter. The nominal yield on 10-year Treasuries is shown in blue for comparison. Historically the two have been fairly close, though most of the time since 1980, the 10-year yield has been higher than the nominal GDP growth rate. For example, the average 10-year yield from 1990-2013 was 5.1% and the average annual nominal GDP growth rate was 4.5%.
The comparison between the interest rate and the growth rate matters for questions like the following. Suppose tax revenues were just sufficient to cover all items in the federal budget other than the interest expense. Then the outstanding debt would still be there at the end of the year, plus we’d owe one more year’s worth of interest. On the other hand, if the economy also grew during the year, that could help make the debt look less big relative to total GDP. If the interest rate is higher than the growth rate, the first effect is going to outweigh the second. In that case, unless our taxes are enough to cover all of the government’s non-interest expenditures plus at least some of the interest expense as well, then our debt will grow as a percent of GDP, and we’ll be in an even deeper hole at the end of the year than when we started.
Figure 4 above plots the primary surplus for the United States– government revenues minus spending on all items other than interest expense– as a percent of GDP. Typically this would need to be positive in order to keep debt from growing relative to GDP. One can see this interaction by looking at the debt/GDP ratio in Figure 5 below. The large primary surpluses of the 1990s brought debt down as a percent of GDP, but most of the rest of this period saw primary deficits and a growing debt-to-GDP ratio.
Let’s use the average 0.6% difference between the interest rate and the growth rate observed over 1990-2013 for some sample calculations. If net debt is 50% of GDP, it means that the primary surplus would need to exceed 0.3% of GDP every year in order to keep debt from growing relative to GDP. If net debt is 100% of GDP, a permanent primary surplus of 0.6% of GDP would be needed to keep debt from growing relative to GDP.
But there’s another important detail to factor in. The experience of most countries has been that when the debt load becomes higher, the interest rate goes up. For example, Baldacci and Kumar (2010) found in a study of 31 advanced and emerging economies over 1980-2007 that a one-percentage-point increase in government debt/GDP was associated with a 4-basis-point increase in the 5-year-10-year forward interest rate. Ichiue and Shimizu (2013) found for 10 advanced countries over 1990-2010 (for which Germany was the only representative of the eurozone) that a one-percentage-point increase in government debt/GDP plus a 1% increase in external debt/GDP raised the 5-year-10-year forward rate by 3 basis points. Laubach (2009) inferred from changes in U.S. CBO projections over 1976-200 that a one-percentage-point increase in debt/GDP was associated with a 3-to-4-basis-point increase in the 5-year-10-year forward rate. Vincent Reinhart and Brian Sack (2000) analyzed the G7 countries over 1981-2000 (prior to formation of the euro) and found that a 1% decrease in projected surplus relative to GDP was associated with a 12-basis-point increase in the 10-year-3-month interest rate spread. And Greenlaw, Hamilton, Hooper and Mishkin’s (2013) analysis of 20 advanced economies over the last decade found that a one-percentage-point increase in debt/GDP was on average associated with a 4.5-basis-point increase in the 10-year yield.
U.S. net debt averaged 45% of GDP over 1990-2012, the base period used in the above sample calculations. We’ll be entering the next decade with a debt-to-GDP ratio 30 percentage points higher than that. The empirical studies just mentioned suggest that could easily raise the 10-year rate by more than 90 basis points relative to where it would have been if we’d held debt to 45% of GDP. In other words, given current U.S. debt loads, we might expect to see a nominal interest rate over the next decade that is 1.5% higher than the GDP growth rate instead of the 0.6% differential observed on average over 1990-2013.
There are those who argue that the interest rate is below the GDP growth rate at the moment, so why worry about it? The problem is that these debt levels are not going to go away. We’re going to be stuck paying the interest on the debt we’ve already accumulated well into the foreseeable future, through good times and bad. That’s why I think it’s important to consider the longer-run historical experience of our country and others, and not just the situation holding at the moment, to get a clear understanding of exactly what we’ve gotten ourselves into.
Consider for example the historical episode over the last two generations for which the U.S. was most successful in bringing its debt load down, namely the 8 years when Clinton was president (1993-2000). Over this period, the primary surplus averaged 2.1% of GDP. Given Clinton’s starting debt load of 49%, those primary surpluses were big enough to bring debt down to 35% by 2000. But if Clinton had started out with debt at 100%, and had exactly the same success with raising tax revenues and reducing non-interest spending relative to GDP, the numbers just discussed could mean that those same policies would have accomplished nothing in terms of reducing the debt burden. Moreover, if we let debt get to 100% of GDP, we’d have to repeat Clinton’s success decade after decade forever just to hold debt constant at 100% of GDP.
The mathematical principle behind this is very simple. If we start with debt at 100% of GDP instead of 50%, we have to run that much faster just to stay in the same place.
And next time we’re going to be running the race with a bigger fraction of the population in retirement and with much higher medical costs than under Clinton.
Here’s my advice: try not to start the race owing 100%.