The Macroeconomics of Deleveraging
The Macroeconomics of Deleveraging
Brandon Adams
Charles Dickens said of debtors’ prisons, “Any one can go IN…but it is not every one who can go out.”[1] In today’s highly leveraged macroeconomic environment, Dickens’s words have found renewed resonance. Current economic circumstances increase the pressures for a period of sustained deleveraging. Deleveraging occurs when debt levels decline relative to a measure of income or assets. The macroeconomic effects of deleveraging are far more severe than most observers appreciate. This chapter supports one of the more controversial theses of this book: that successful, comprehensive deleveraging and long-term US federal government solvency are incompatible.
The United States has experienced an incredible increase in indebtedness over time. Total debt in nonfinancial sectors to GDP in the United States is at 243 percent as of 2009. GDP in 2009 was $14.256 trillion.[2] This implies a total debt of around $34.702 trillion. The rough breakdown of this is government debt of $10.168 trillion,[3] corporate debt of $10.998 trillion, and household debt of $13.536 trillion.[4] As of December 21, 2012, total debt in nonfinancial sectors to US GDP stands at about 250 percent. Recent best estimates for 2012 GDP sit at $15.88 trillion, implying a total indebtedness of about $39.7 trillion. The nominal indebtedness alone is quite profound, but what is most important is that the debt trend remains post-crisis.
If one is analyzing the world from the perspective of neoclassical economics, then one is flying blind in a world of massive indebtedness. In Stabilizing an Unstable Economy, Hyman Minsky notes, “In the neoclassical view, speculation, financing conditions, inherited financial obligations, and the fluctuating behavior of aggregate demand have nothing whatsoever to do with saving, investment, and interest rate determination.”[5] Minsky thought that modern economic theory, by abstracting away from institutional realities, missed too much of what is important. I’ll take a deeper look at his theories later in this chapter.
The macroeconomic effects of debt are a bit difficult to get one’s head around. Debt that is taken on for the purposes of either consumption or investment will tend to increase the money supply in the period it is taken on and decrease the money supply in the period it is paid back. Debt taken on for purposes of consumption will tend to increase GDP in the period of consumption by an amount somewhat greater than the value of the debt, and it will tend to decrease GDP in the period it is paid back.
The ratio of total debt to GDP has trended relentlessly upward for thirty years. This is the basis for George Soros’s superbubble hypothesis—simply put, the notion that markets reflect more than the fundamentals would suggest that cause asset prices to move to extremes. The superbubble is not confined to the United States—most Western economies have traveled a similar trajectory. According to Peter Warburton, who examined the performance of major Western economies during this superbubble, “It appears that a given percentage addition to private sector debt is associated with less and less economic growth with the passage of time.”[6] There appears to be diminishing short-term macroeconomic returns to debt, and, in the United States, at least, we are far out on the curve.
Roughly speaking, every dollar used to pay off debt in a given period represents a reduction of aggregate demand by one dollar and a reduction of GDP by one dollar. This is counterintuitive but approximately correct. However, our standard economic intuition might suggest that when one dollar is used to pay off debt, some of that dollar will then be spent by the lender, offsetting the direct decrease in aggregate demand; aggregate demand, therefore, could stay somewhat constant. But this logic is wrong. Understand that, at the end of the day, the economy consists only of individuals; government and corporations are both economic constructs that are worthless in the absence of the individuals behind them. A dollar of debt reduction in one period will reduce income, expenditure, and output by a dollar in that period, unless there is a change in individual preferences regarding consumption and saving decisions.
Suppose I have a salary of $100,000, a bank account containing $100,000, and a note receivable that has a market value of $50,000. When choosing how much to spend in the next year, the primary variables of interest are my wealth level, my income level, the real interest rate offered by potential investments, and the value I attribute to future consumption as opposed to current consumption. Whether my wealth takes the form of cash or notes receivable doesn’t make much difference. For my $50,000 note receivable to be paid off in a given time period and converted to cash in my account, someone somewhere in the economy will have to have income that exceeds expenditure by at least $50,000 in that period. On some level, I might understand that if I don’t increase my expenses by $50,000 in the period that my debt is paid off, then aggregate demand and aggregate output will be light by $50,000 in that period, but I don’t much care: my objective is to maximize my individual well-being. If debt repayment occurs en masse, then we can expect a sharp retrenchment in output and expenditure. Corporations can hardly be expected to invest heavily when they see consumption on the retreat….