Andy Haldane asks in the Financial Times whether Keynesianism is dead. His argument is familiar. Fiscal policy worked when government debt was low, but after the enormous interventions following the financial crisis, COVID and the Ukraine war, public debt has become so large that fiscal stimulus may now be ineffective or even contractionary. Households anticipate future taxes, bond markets raise borrowing costs, and central banks tighten in response. What has become known as a Ricardian argument, and the older and traditional Treasury view. Haldane even resurrects the old argument about expansionary fiscal contractions, citing Ireland, Denmark and the eurozone periphery. There is a rather obvious problem with this story. The most recent large scale Keynesian experiment worked exceptionally well.
Compare the recovery from the Pandemic recession with the three previous US recoveries. The recoveries following the 1990–91, 2001 and especially 2007–09 recessions were notoriously jobless. After the Great Recession, output recovered painfully slowly and employment even more slowly. A decade after the crisis, I noted on this blog that the economy had taken roughly as long to recover as it had after the Great Depression, despite the much smaller initial decline. Fiscal stimulus had prevented another Depression, but it had been far too small and, crucially, was withdrawn too soon.
This slow recovery eventually generated an elaborate discussion of secular stagnation. Perhaps, it was suggested, mature capitalist economies simply suffered from a chronic shortage of profitable investment opportunities, an excessively high propensity to save, or a natural rate of interest so low that monetary policy could no longer produce full employment. I argued at the time that this put the causality backwards. There was no mysterious secular stagnation mechanism condemning the United States to slow growth. There were stagnation policies related to insufficient fiscal expansion and premature austerity.
The Pandemic provided something close to a natural experiment. This time the federal government responded on a much larger scale, first under Trump and then, importantly, with the Biden fiscal packages. And the recovery was dramatically faster. Real GDP returned to trend and, unlike the previous recoveries, employment came back rapidly too. As I noted in 2024, this was the first recovery in a long time that was not jobless, and the United States considerably outperformed most other advanced economies. The obvious difference was not some sudden disappearance of the structural forces supposedly producing secular stagnation. It was fiscal policy.
This is particularly awkward for Haldane because the large fiscal expansion occurred when the US public debt ratio was already very high. According to his argument, that is precisely when the fiscal multiplier should have become small or even negative. Instead, the largest fiscal intervention since the war produced the fastest recovery in decades. If high debt causes households to respond to government spending by saving in anticipation of future taxes, the Ricardian households evidently failed to show up when they were most needed by the theory.
Nor does the subsequent inflation rescue the argument. The Pandemic recovery certainly coincided with inflation, but that does not establish that excessive aggregate demand was its cause. The acceleration began amid extraordinary supply-chain disruptions and was reinforced by energy and commodity shocks. More importantly, inflation subsequently fell sharply without the large increase in unemployment that the excess-demand story implied would be necessary. I argued at the time that the Pandemic inflation was being used to revive the old New Consensus precisely when the experience of the recovery had cast doubt on it.
There is also something peculiar about Haldane's treatment of interest rates. He tells us that high government debt raises bond yields and thereby offsets fiscal expansion, the conventional crowding out argument, but then acknowledges that central banks have themselves raised short-term rates dramatically and engaged in quantitative tightening. Bond yields are not prices determined independently of monetary authorities by anonymous bond vigilantes (see my Jacobin piece). Central banks have considerable influence over the whole yield curve. To raise rates deliberately and then point to higher rates as evidence that fiscal policy has become unsustainable is rather circular.
None of this means that fiscal policy faces no limits. But those limits need to be correctly identified. For a government borrowing in its own currency, the fundamental domestic constraint is the availability of productive resources and the possibility of inflation, not an arbitrary debt-to-GDP ratio. Countries that do not issue the international currency face an additional external constraint because they cannot create the foreign exchange needed to purchase imported goods. That is quite different from treating a government's domestic-currency debt as if it were analogous to household debt. Keynes himself eventually came to distinguish sharply between what could be financed domestically and the genuine constraint created by the need for foreign resources. In the case of most advanced economies, and in all truth most but a few peripheral countries (those that didn't accumulate foreign reserves and have large payments in foreign currency, like Argentina), there is no significant limit to fiscal expansion.
Perhaps the greatest irony is that the Pandemic should have settled at least part of this debate. After years in which economists tried to explain weak growth through demographics, technology, a savings glut, the natural rate of interest and secular stagnation, a sufficiently large fiscal expansion produced rapid growth and an extraordinarily rapid labor market recovery. When fiscal support now begins to disappear, growth unsurprisingly slows again. As I noted recently, declining government expenditure is already exerting a negative influence on the US economy.
Haldane’s final invocation of the paradox of thrift is particularly odd. For Keynes, the paradox was that an attempt to save more by spending less reduces income through the multiplier, so that aggregate saving would fall, not rise. If governments respond to high debt by cutting expenditure, that is not the cure for the paradox of thrift. It is a textbook way of reproducing it.
The evidence favors Keynes' argument on the paradox of thrift. The sluggish recovery after 2008 followed an inadequate stimulus and a premature turn toward austerity, and the remarkably rapid recovery after 2020 followed an exceptionally large fiscal expansion. Keynesianism is not dead. If anything, the Pandemic recovery suggests that what had been mistaken for secular stagnation was largely the consequence of insufficient effective demand. What refuses to die is the belief that prosperity can be restored by governments spending less. The zombie is austerity!
No comments:
Post a Comment