28 March 2018

The problem with central bankers’ inflation preoccupation

By Scott Sumner

In recent years, a growing number of prominent economists, including Michael Woodford, Christina Romer and Larry Summers, have switched their allegiance from inflation targeting to targeting nominal GDP, the non-inflation-adjusted sum of all spending in the economy. One of the best ways to understand the growing popularity of NGDP targeting, is to re-examine what went wrong during the global financial crisis and how NGDP targeting could have mitigated the damage.

When the Federal Open Market Committee (FOMC) met on September 16, 2008, the United States was already nine months into the Great Recession, the worst slump since the 1930s. Two days earlier, Lehman Brothers had gone bankrupt, triggering a major banking crisis.

Despite these worrisome signs, the Fed voted not to ease monetary policy, keeping the Fed funds target unchanged at 2 percent. In his recent memoir, Ben Bernanke acknowledged that this was a mistake, noting that the Fed was distracted at the time by banking problems and not sufficiently focused on monetary policy. But the minutes from the meeting suggest an addition problem: too much focus on inflation risks.

At the time the Fed met, the most recent inflation data showed a 4.2 per cent rate over the previous year, well above the Fed’s 2 per cent target. In contrast, NGDP growth had been running at about 2.7 per cent, which was well below the roughly 5 per cent trend rate of NGDP growth during the 1990-2008 period.

High inflation is not a good indicator of the state of the economy, because it can either reflect excessive demand or a decrease in aggregate supply. In contrast, NGDP growth shows the increase in total spending in the economy — i.e. aggregate demand — and is an excellent indicator of whether the economy is overheating and needs monetary restraint.

In 2008, the Fed focused on the wrong indicator.

Inflation was distorted by soaring oil prices, and Fed policy during 2008 ended up being far too contractionary for the needs of the economy. As a result, NGDP fell by more than 3 per cent between mid-2008 and mid-2009, the worst performance since the 1930s.

A policy aimed at 4 per cent or 5 per cent NGDP growth would have likely led to above 2 per cent inflation during 2008-09, but that sort of problem is nowhere near as costly as 10 per cent unemployment.

The Fed’s policy error in 2008 is not an isolated incident. The European Central Bank (ECB) is even more focused on inflation than the Fed (which has a dual mandate of stable prices and high employment) and made two even costlier errors during the Great Recession.

The ECB actually raised rates during July 2008, insuring that the Eurozone would have an even deeper recession than the United States, despite the subprime mortgage fiasco having occurred in America. Then in 2011, when rising oil prices and VAT increases briefly pushed Eurozone inflation above target, the ECB raised rates two more times, pushing the Eurozone into a severe double dip recession, which worsened the debt crisis.

Unstable growth in NGDP creates two problems: labor market volatility and financial market turmoil. Recall the circular flow model in economics, where output equals income. This means that NGDP is not just total output in current dollars; it is also the total gross nominal income in the economy.

Think of NGDP as the total resources available to pay wages and salaries, and also to repay nominal debts. If all wages and debt payments were indexed to NGDP, then unstable NGDP growth wouldn’t be much of a problem. In fact, however, almost all wage and debt contracts are priced in nominal terms. This means that when there is an unexpected drop in NGDP, people, businesses and governments have fewer resources to pay wages and interest.

To explain the strongly negative correlation between NGDP growth and the unemployment rate, I use the metaphor of the game of musical chairs. If firms are suddenly receiving far less revenue than they anticipated when signing nominal wage contracts, then some workers will be left sitting on the floor (i.e., unemployed) when the music stops.

The relationship between financial crises and NGDP is slightly more controversial, but it really shouldn’t be. Many pundits reverse the causality between NGDP growth and financial distress, arguing that a debt crisis causes falling NGDP. While there may be cases where this occurs, there is probably much stronger causality going in the other direction.

The misdiagnosis occurs because when NGDP falls sharply, the first debtors to get into trouble will be those who made the most foolish decisions — say subprime borrowers in America, or Greek government borrowing in Europe. Thus reckless borrowing is seen as causing a financial crisis, which then depresses NGDP.

There is some truth to this claim, but pundits overlook that after any major decline in NGDP, debt problems spread far beyond the most reckless class of borrowers. Most bank failures in America during the Great Recession were attributable to defaults on loans to businesses, not subprime mortgages.

Falling NGDP causes a debt problem of manageable size — due to reckless borrowing — to spread to other borrowers who would have remained solvent if national income had continued to grow at a steady rate.

Stable growth in NGDP is not a panacea. It does not prevent individual firms and even sovereign governments from excessive borrowing. The economy may also be hit by supply shocks.

What NGDP targeting can do, however, is prevent various real problems from bleeding into nominal spending, causing unnecessary damage to the broader economy. If there may be a price to pay for reckless borrowing, that price should be less consumption and lower real wages, not high unemployment. When a country has borrowed too much, it does no good to enact policies that cause output and employment to plunge, making it even harder to repay debts.

NGDP growth equals inflation plus real GDP growth. When there is an adverse supply shock, a 4 per cent NGDP target will result in a brief period of higher than normal inflation. But that’s actually an efficient way of making sure that real wages stay at levels consistent with full employment. In that case, it’s better to keep the workforce employed at slightly lower real wages, than to obsessively target inflation at the cost of sharply higher unemployment. Similarly, during a productivity boom, an NGDP target will lead to a period of below normal inflation. Once again, that’s a feature, not a bug. It keeps labour markets from becoming overheated during a period of rapid growth.

Economists are beginning to understand that NGDP is the variable we should actually be concerned about. Instead of worrying about what might happen to inflation under NGDP targeting, we should consider what happens to NGDP if we insist on targeting inflation.

Scott Sumner is the Ralph G. Hawtrey Chair of Monetary Policy at the Mercatus Center at George Mason University, where he is director of its Program on Monetary Policy.