A victory, and a test

By R.A. | WASHINGTON

YESTERDAY, Tyler Cowen has declared, was Scott Sumner day. That strikes me as a fair honour to bestow. For it does appear that the Federal Reserve's statement yesterday represented a subtle but perceptible shift in monetary policy strategy—one that probably would not have occurred without Mr Sumner's efforts.

To appreciate the shift, take a step back to consider the evolution of the discussion over the past few years. Prior to the crisis, the prevailing economic consensus was that central banks could and should stabilise the macroeconomy by maintaining a low and stable rate of inflation, and that interest-rate policy was an appropriate and effective way to do that. In the summer of 2007, the American economy was showing clear signs of a loss of momentum, and the worsening housing crisis began to have its first serious impacts on the global financial system. In response, the Fed slowly began reducing its benchmark interest rate. Further worrying signs from December of 2007 (the official beginning of the recession) prompted a faster pace of rate cuts. The Fed then paused over the summer while oil prices spiked, then sprang into action when the financial crisis suddenly intensified in September (four years ago this week).

The Fed's response consisted of several different components. First, it slashed its benchmark interest rate by 200 basis points between October and December, leaving it at a target range between 0% and 0.25% by the end of the year. Second, it played a substantial role as lender of last resort in order to prevent a collapse of the financial system. In some cases this involved direct intervention and emergency lending to banks (including, after the collapse of Lehman Brothers, to broker-dealers). In others it meant involvement in critical markets, like that for commercial paper, that were on the verge of breaking down. Third, it initiated a policy of "credit easing". By purchasing mortgage-backed securities and debt issued by Fannie Mae and Freddie Mac it hoped to unclog paralysed mortgage markets in order to pry open a key transmission channel for monetary policy. The organising principle for these moves was clear: the Fed wished to prevent a catastrophic fall in the money supply that would usher in deflation and lead to a replay of the Great Contraction of the 1930s.

By early 2009, it was clear that short-term inflation expectations were in danger of turning negative. Unable to cut rates further, the Fed added a new policy to its attack: quantitative easing. Like credit easing, quantitative easing involved the purchase of assets in the open market. But unlike credit easing, the goal was not to pry open specific credit channels. Rather, the goal was specifically to expand the Fed's balance sheet. And so in addition to the mortgage securities that it continued to buy, the Fed began purchasing long-term Treasuries—$300 billion worth per the announcement in March of 2009.

QE was thought to raise demand in several different ways (you can read about this in more detail here). First, the act of purchasing Treasuries from private investors should lead them to rebalance their portfolio: to take the cash they'd received from the Fed and use it to replace the Treasuries with other securities. That "portfolio rebalancing" would work its way through the financial system, raising asset prices in the process. Higher asset prices should encourage more consumption and investment. If the rebalancing led to purchases of foreign assets, it would lead to a dollar depreciation that could also have a stimulative impact on net exports. Second, QE could have a fiscal effect, by reducing the expected debt costs of the government and thereby reducing expected future tax burdens. And third, QE could have a signalling effect. Since higher interest rates would lead to Fed losses on its securities holdings, a larger balance sheet could conceivably signal a greater tolerance for inflation on the central bank's part. However, the Fed probably stepped on this effect to some extent with its decision to pay interest on excess reserves held at the Fed, which gave it an alternative means to rein in inflation.

Empirically, QE had many of the anticipated effects. As Ben Bernanke discussed in his recent speech at Jackson Hole, studies of recent QE in America and elsewhere find that it reduced interest rates on Treasuries and private bonds, and that it seemed to raise growth and reduce unemployment relative to but-for levels (so far as this can be determined). March did prove to be the pivot point for equity prices and inflation expectations. Shortly thereafter, American output began growing again and job losses decelerated sharply.

Yet even as this was taking place, the conventional wisdom across the economics wires was that monetary policy had largely done all it could do, or perhaps all it should do. The most straightforward argument on this score was that with interest rates at zero, the Fed was powerless to create more demand. QE could prop up banks, many suggested, but it could not influence the real economy. What was necessary instead was fiscal expansion, which could bypass a limping banking system and plow money directly into the economy.

A more sophisticated critique emerged from people like Paul Krugman, who diagnosed the economy has having sunk into a liquidity trap. With interest rates near zero, he argued, balance-sheet policy—swapping one zero-yield asset for another—was unlikely to prove effective. The only way the central bank could further stimulate the economy would be to slash the expected real return on bonds below zero by promising high inflation in the future. But this was hopeless; to do this, the Fed would have to credibly promise to "be irresponsible": to tolerate inflation in some future in which the economy was no longer at the zero bound. Since no one would believe that the Fed would do such a thing, no one would expect high future inflation. The only solution to the problem is fiscal stimulus.

The Fed itself appeared to chart a different course. The key lesson of the Depression, Mr Bernanke appeared to signal, was that deflation must be avoided at all costs. Having successfully pulled the American economy back from the brink of deflation—having in fact restored medium-term inflation expectations to near 2%—the Fed had done its job. If real growth remained disappointing, he seemed to suggest, then it fell to the government, via housing policy solutions, banking reforms, and appropriate fiscal steps, to nurse the economy back to full health. This has been the Fed's guiding principle through the recovery. It has stepped in on multiple occasions since early 2009 to announce new asset purchases or tweak its language in stimulative fashion. But the goal has always been the same: to halt a disinflationary slide in an effort to prevent any lapse into deflation.

As these debates were playing out, a dissenting voice emerged in a tiny corner of the economics blogosphere. On February 2nd, 2009, Scott Sumner launched his blog with a flurry of posts assailing the conventional wisdom on monetary policy. But almost no one knew who Mr Sumner was, and so no one read those posts. Then on February 25th, Tyler Cowen linked to his blog, and readers and comments began trickling in. Mr Cowen's status within the economics blogosphere signaled to the rest of us that these were ideas worth engaging.

The result was a long and detailed conversation on Mr Sumner's key ideas, the heart of which was a call to refocus monetary policy on nominal GDP rather than inflation. As Mr Sumner is quick to admit, this is not an idea original to him. Rather, he argues (compellingly, in my view) it is one that follows directly from Milton Friedman's monetarist revolution but which lost out to inflation targeting during the early years of the Great Moderation. The insight is that the central bank's provenance isn't the money supply or interest rates or inflation but simply demand, best captured in nominal output or income: the total amount of money spent each year. If one accepts, as most macroeconomists do, that nominal shocks can have real consequences, and if one then accepts, as is common, that central banks ought to try to smooth out nominal shocks, then it makes the most sense to just stabilise the biggest nominal aggregate. Alternatively, the best way to insure against a damaging fall in demand is to make sure that demand doesn't fall, or at least to promise to quickly rectify any demand drop that does occur, by maintaining a stable path for nominal output growth.

In the context of the most recent crisis, this philosophy emerged as the argument that the economy's weakness was both captured by and caused by the sharp decline and subsequent slow growth of nominal output. Furthermore, an effort to raise nominal output should, on its own, be enough to raise real output—if in fact the weak economy is correctly diagnosed as suffering primarily from a demand shortfall. And finally, if the economy is instead suffering primarily from supply-side factors, stabilising nominal output and moving NGDP back to trend is nonetheless a good way to make sure that monetary policy doesn't frustrate the process of real adjustment, even if that corresponds to inflation rates above current targets. In other words, if the economy has been hit by a supply shock, holding back NGDP growth to prevent inflation from rising above target makes things worse, not better.

The notion that the central bank should focus on raising NGDP and that inflation is largely a sideshow has taken a while to catch on. But caught on it has. That is down partly to the effectiveness of the idea itself and the argument developed by Mr Sumner. And it is down partly to the fact that Mr Sumner's framework has done a good job explaining the ups and downs of recovery. Blogs helped his idea find an audience. As the audience grew, Mr Sumner was able to find print outlets for his views. And through blogs, economists advancing the idea were able to communicate and deepen their arguments, eventually forming what has been called the first school of economics to emerge online: the market monetarists. This school now has a book out, edited by David Beckworth.

As the blogosphere cottoned on to the idea, it seemed to trickle up. This newspaper covered the idea in a column in August of last year. In October of 2011 Goldman Sachs' Jan Hatzius endorsed the idea in a research note, and Christina Romer, a former head of Barack Obama's Council of Economic Advisers, signed on in a column at the New York Times. In November of last year, Mr Bernanke was asked about NGDP targeting at a post-meeting press conference. And at this year's Fed convocation in Jackson Hole, renowned monetary economist Michael Woodford presented a paper to an audience of central bankers which came down in support of the concept.

In his response to the NGDP question last November, Mr Bernanke was fairly dismissive. Yet not long after, Chicago Fed President Charles Evans began to speak in favour of a strategy that would in practice detach Fed policy from strict inflation targeting and move it toward an NGDP goal. Mr Evans recommended that the Fed tolerate inflation up to 3% until unemployment falls below 7%. It is a policy that seems to acknowledge that strict inflation targeting is insufficient to generate a full demand recovery. More inflation, as part of more nominal output, may be necessary to raise real output back toward potential.

The Fed didn't adopt Mr Evans' policy yesterday; it remains uncomfortable with explicit numerical targets and very uncomfortable with explicit targets that involve inflation above its official medium-run goal. Yet as my colleague and I noted yesterday, this line:

If the outlook for the labor market does not improve substantially, the Committee will continue its purchases of agency mortgage-backed securities, undertake additional asset purchases, and employ its other policy tools as appropriate until such improvement is achieved in a context of price stability...

Looks like an embrace of the Evans framework, sans the explicit thresholds. Markets clearly read the statement as a qualitative shift in policy. Equities soared, reflecting an increased outlook for nominal growth. Market expectations of inflation also rose sharply (relative to the recovery norm, at any rate; 5-year expectations remain at a subdued 2.34%).

The intellectual building blocks for yesterday's Fed move existed before Mr Sumner began blogging. It is hard to imagine them being assembled into actual policy so quickly without his efforts and their rapid spread through the economics blogosphere.

As the market monetarist community is now pointing out, the Fed's new policy is a step in the right direction, but it is a long way from what they would actually recommend implementing. And they're right. Fairly or not, however, the policy will be judged as a test of market monetarist ideas. Yesterday's market moves suggest that nominal output growth should accelerate in coming quarters. How much acceleration is likely to occur will depend on how much room the Fed is willing to give the economy to run. If the rise in inflation expectations leads the Fed to begin walking back its new language a month from now, the gain will be small.

But the expectation should be that there will be higher nominal growth. And given the broad view that demand-side weakness is the primary constraint on recovery, that nominal growth should be accompanied by faster real growth and faster employment growth. Based on my own observations of the co-movement of inflation expectations and payrolls, I'm expecting nonfarm employment to be growing at more than 200,000 jobs a month by the end of the year, with more possible depending on the Fed's touchiness.

I recognise that if nothing like that occurs, it will represent evidence that I've gotten one or several things wrong. Yesterday's move was exciting no just because it represented an intellectual victory or because it could lead to real economic improvement, but also because it holds out the possibility of new knowledge; one way or another we should learn something important. It was a victory and a test for market monetarism.

It is unfortunate for the school's adherents that both the victory and the test are so incomplete. But they should be glad and proud all the same. And for the sake of the unemployed, let's all hope they're right.