Those who believe more government spending is the antidote to the current economic malaise generally rely on one or more assumptions to defend their nonsensical position: (1) the future costs of servicing debt issued to finance larger deficits don’t matter or are trivially small; (2) the central bank can control future interest rates, so we need not worry about a spike in future borrowing costs; and (3) the only channel through which government borrowing can negatively impact economic activity is through higher interest rates at full employment. Let us take each in turn.
First, it seems worth pointing out the obvious fact that the debt-to-GDP ratio cannot grow indefinitely because eventually the cost of servicing public debt would end up exhausting all of national income. So even if one believes an incremental dollar of debt-funded stimulus would accelerate economic activity, there is likely to be some point at which cumulative deficits have consumed too much of future income. For example, without a single added dollar of national debt, the cumulative deficits of the past four years have already imposed future costs larger than the annual spending reductions agreed to in the 2011 Budget Control Act (including the sequester).
Assuming interest rates of 5% – above but close to the historic average – the additional 35% of GDP in debt the federal government has accumulated since 2008 would cost future generations 1.75% of their gross income in perpetuity. This may seem small but 1.75% is equal to $280 billion of 2013 GDP and the ten-year costs of this interest expense would be $3.5 trillion at a 5% nominal GDP growth rate. This is nearly three-times as large as the $1.2 trillion in 10-year cuts from sequestration. If this spending is so important to current households, why would we rob future governments of the financial capacity to provide these services to future generations?
To be clear: interest rates today are far from 5%. The Treasury can finance itself at less than 1% weighted average rate. But the proponents of larger deficits argue that we need not ever pay the debt back. Instead, we’re assumed to “grow our way” out of debt. Since the debt principal balance is assumed to be with us forever, the appropriate way to view the long-run cost of debt is using the long-run interest rate the Treasury is likely to pay on it in perpetuity. By this standard, the cost of deficit spending to date has already shaved the current year equivalent of $280 billion off of annual national income. Asking for further reductions in future income seems self-indulgent, or, at the very least, out of place for those so concerned about the impact of spending cuts one-third as large through sequestration. It is also important to recognize that with a weighted average maturity of just 63 months and $7.9 trillion of debt coming due within 5 years, the risk of a significant re-pricing driving interest expense upward is very real.
Having established that interest expense is a real cost that cannot be ignored, it is worth turning to the second argument put forward by spending enthusiasts: there is no risk of an exponential rise in interest rates because the Fed controls interest rates. Proponents of bloated budgets often use the straw man of an impending fiscal crisis to detract attention from the cost of debt service. But the risk of eventual fiscal crisis is very real. The problem is that it’s impossible to know precisely when the debt-ratio and policymakers’ indifference to its increase combine to create a crisis of confidence.
Instead of acknowledging that a practical debt limit exists, spending enthusiasts simply assume it away. A government that borrows in a (fiat) currency that its central bank prints can never run out of money because the central bank can print whatever amount of currency is necessary to repay principal when it comes due. If the economy has resource slack, the central bank can hit whatever interest rate it wants on most points on the yield curve with minimal risk of accelerating inflation. So if the central bank can engineer negative real interest rates what is there to worry about?
First, the 2003-2008 experience teaches that “price stability” and “financial stability” are two very different things. A monetary policy stance consistent with price stability could allow for the build-up of extreme imbalances in financial markets and asset prices that lead ultimately to extreme outcomes. Second, there is some critical debt level above which an increase in interest rates would leave the government insolvent simply because the financing costs are too large relative to the income tax base.
At this point, the central bank essentially becomes a hostage of the fiscal authority and “monetary policy” becomes synonymous with public debt management. Under these circumstances, the Fed would be forced to choose to either (1) continue to suppress interest rates and accept the resulting inflation or financial instability; or (2) allow rates to find their natural level, which could leave the sovereign insolvent and result in some kind of sovereign restructuring akin to the Greek experience. Either way, debt accumulation would eventually leave the central bank with nothing but bad options.
Finally, the key zinger thrown out by Paul Krugman and others seeking to accelerate America’s national bankruptcy is that those concerned about mounting debt levels have no credibility because they argued that inflation and interest rates would have spiked by now. But by focusing on these specific arguments, Krugman ignores the much more coherent critique that larger deficits do not stimulate demand because they generate more private savings. A household that correctly sees an increase in deficit spending as an increase in its future tax burden would spend less today in response to the expectation of having to pay more in taxes tomorrow. This is a simple application of conditional expectations of tax rates to Friedman’s nearly 60-year old “permanent income hypothesis.” The result is more savings, low interest rates, no acceleration in growth, and more debt that will have to be serviced.
If this sounds unrealistic, consider that despite $140 billion in annualized tax increases, household spending actually made a much larger contribution to GDP growth in Q1-2013 (2.24%) than it did in Q4-2012 (1.28%) and Q3-2012 (1.12%). How could it be that spending increased even as disposable income fell due to the tax increases, especially the 2% increase in the payroll tax?
The answer is that, as Milton Friedman found in 1957, current period income does not determine current period spending. Instead, households make savings, spending, and borrowing decisions based on expectations of the future. Spending increased in 2013 because household savings rates fell and credit utilization increased. By contrast, spending in 2012 was less than stimulus proponents expected because households saved a large chunk of the tax cut in anticipation of the tax increases likely to come the following year. Moving tax liabilities in time does not eliminate them. More debt does not push up interest rates because households recognize this fact, particularly when debt levels are high, as they are today.