The Normalization Delusion

Adair Turner

LONDON – There is a psychological bias to believe that exceptional events eventually give way to a return to “normal times.” Many economic commentators now focus on prospects for “exit” from nearly a decade of ultra-loose monetary policy, with central banks reducing their balance sheets to “normal” levels and gradually raising interest rates. But we are far from a return to pre-crisis normality.

After years of falling global growth forecasts, 2017 has witnessed a significant uptick, and there is a good case for slight interest-rate increases. But the advanced economies still face too-low inflation and only moderate growth, and recovery will continue to rely on fiscal stimulus, underpinned if necessary by debt monetization.

Since 2007, per capita GDP in the eurozone, Japan, and the United States are up just 0.3%, 4.4%, and 5%, respectively. Part of the slowdown from pre-crisis norms of 1.5-2% annual growth may reflect supply-side factors; productivity growth may face structural headwinds.

But part of the problem is deficient nominal demand. Despite central banks’ massive stimulus efforts, nominal GDP from 2007-16 grew 2.8% per year in the US, 1.5% in the eurozone, and just 0.2% in Japan, making it impossible to achieve moderate growth plus annual inflation in line with 2% targets. US inflation has now undershot the Federal Reserve’s target for five years, and has trended down over the last five months.

Faced with this abnormality, some economists search for one-off factors, such as “free” minutes for US cell phones, that are temporarily depressing US inflation measures. But mobile-phone pricing in the US cannot explain why Japan’s core inflation is stuck around zero. Common long-term factors must explain this global phenomenon.

Labor-market developments are key, with wage growth remaining stubbornly low even as unemployment falls to “normal” pre-crisis levels. Japan is the most extreme case: with a shrinking labor force, minimal immigration, and a 2.8% unemployment rate, all standard models predict accelerating wage growth. But however much Prime Minister Shinzo Abe urges employers to give Japanese workers a raise, growth in compensation remains sluggish: in June, total wages grew just 0.4%. In the US, too, each new batch of monthly data indicates strong employment growth and surprisingly low wage growth.

Three factors may explain this trend. For 30 years, labor markets have become more flexible, with trade union power dramatically weakened. At the same time, globalization has exposed workers in the tradable sector to global wage competition. But, most important, information technology delivers ever-expanding opportunities to automate all economic activities. In a fully flexible market labor with, as it were, a reserve army of robots, the potential for pervasive automation can depress real wage growth even with full employment.

Nominal demand, meanwhile, is still being held back by an overhang of unresolved debt. Between 1950 and 2007, advanced economies’ private debt grew from 50% to 170% of GDP. Since 2008, debt has shifted from private to public sectors, with large fiscal deficits both an inevitable consequence of post-crisis recession and essential to maintain adequate demand. In addition, the global economy has been kept going by China’s enormous leverage increase, with the debt-to-GDP ratio up from around 140% in 2008 to 250% today. Worldwide, total public and private debt has reached a record high, up from 180% of global GDP in 2007 to 220% in March 2017. As a result, interest rates cannot return to pre-crisis levels without risking a new recession.

Facing this debt overhang, loose monetary policy alone was bound to be ineffective and, beyond some point, potentially harmful and counterproductive. Neither investment nor consumption responds strongly to ever-lower interest rates when debt burdens are high. Very low interest rates, meanwhile, generate asset-price increases, which benefit the already wealthy and reduce the income of less wealthy bank depositors, who in some circumstances might cut consumption more than deeply indebted borrowers increase it.

In this context, as Princeton University economist Christopher Sims argued in 2016, loose monetary policy cannot work through normal transmission channels, and is effective if, and only if, it facilitates fiscal expansion by keeping government borrowing costs low. Nominal GDP in the US has grown faster than in the eurozone since 2007, because the US ran deficits averaging 7.2% of GDP versus the eurozone’s 3.5%. Global growth today is crucially underpinned by China’s 3.7%-of-GDP fiscal deficit, up from 0.9% in 2014. Japan’s continued growth is assured only by large fiscal deficits stretching well into the 2020s; the Bank of Japan, which now holds government bonds equivalent to about 75% of GDP, will hold some of them forever, permanently monetizing accumulated fiscal debts.

The partial recovery this year thus reflects neither a return to pre-crisis normality nor the success of monetary policy alone. But, even if inflation rates remain below target, there is still a good case for some interest-rate increases. Because ever-looser monetary policy alone is decreasingly effective beyond some point, it can be partly reversed with little danger to nominal demand; and slightly higher interest rates would temper, even if only mildly, the inegalitarian impact of the current policy mix.

But the rate increases will and should be very small. I doubt that the US federal funds rate will exceed 2.5% in 2020, while Japanese and eurozone rates will rise only marginally, probably remaining well below 1%. Inflation is more likely to undershoot than to exceed 2% targets. Moderate growth at best will be insufficient to offset the impact of the lost decade of 2007-17.

The psychological bias to expect a return to “normality” will remain strong. But the drivers of post-crisis economic performance are so deep that no return to normality is likely any time soon.