America: are times a-changing?

A HUGE PACKAGE and massive growth in money supply, along with near-zero interest rates, might do the trick. And it may raise inflation indeed. Although inflation is far from being a concern globally, it is potentially important for the following two reasons: First, monetary history tells us relaxing price stability targets and/or constraints results in lower growth rates and larger output gaps. Second, nominal interest rate smoothing, i.e. minimizing the variance of short rates, combined with minimizing the output gap target leads to inflation going adrift. Within the context of a simple model that admits Calvo pricing, real business cycle-driven growth shocks, a Fisher equation and rational expectations, Goodfriend & King have demonstrated the presence of a stochastic inflation trend, which furthermore seems to match closely with the U.S. historical time series data (Marvin Goodfriend & Robert M. King, The Great Inflation Drift). We should add to this an IMF Report’s findings, and those of Reinhart & Rogoff, which claim, moreover, the recovery can be ill-fated at times. The real value of government debt tends to steam up, up-trending by a hefty 86% in the major post-WWII episode. Besides, and this point awkwardly goes against the most recent Turkish evidence, the main cause ordinarily does not lie in rising expenditures or bail-out costs and capital injections that more often than not come under the limelights for ideological reasons, but rather in the collapse of tax revenues in the wake of deep and prolonged economic contractions. Rising budget deficits soar further because the revenue side collapses first. But then this is perhaps true for economies where incomes are directly taxed, and indirect taxes don’t account for such a high share of tax revenues.


Historically, estimated multipliers on government spending lie in the range of 1.5 to 2, while multipliers for tax cuts can be much smaller, say 0.5 to 1. However, these estimates are from periods when households could – and did – use tax cuts to extend a down payment on a car or in order to cover the closing costs on a mortgage refinance. For example, in 2001, the American economy was in recession but households took advantage of zero rate financing promotions – as well as ready access to home equity withdrawals from mortgage refinancing – to lever up their tax cut checks to buy cars and boost overall consumption. With the credit markets impaired, tax cuts, as well as income earned from government spending on goods and services, will not be leveraged by the financial system to nearly the same extent, resulting in (much) smaller multipliers. The true fiscal multiplier, as well as the tax multiplier, is of extreme importance since it would give us the magnitude of the public spending effect on GDP and on unemployment, and its time path. The debate took off a decade ago when, on January 10, the Obama team’s “official” plan was released. It has come to be known as the Romer-Bernstein estimates (Christina Romer & Jared Bernstein, The Job Impact of the American Recovery and Reinvestment Plan).


A famous post-Lehman estimate, i.e. the Romer-Bernstein estimate of a 1.6x spending multiplier was based on an interest peg, i.e. that the Fed would keep a zero-rate policy (ZIRP) for at least four years, or even more. The spending increase was also assumed to be permanent, contradicting the Sargent-Wallace proposition to the effect that a permanent government spending increase, along with a fixed interest rate, would lead to instability and non-uniqueness of equilibrium in a rational expectations framework with forward-looking agents. Inflation expectations go out of hand and the price level would eventually spiral out and explode. Hence, in a framework, combining a fixed interest rate rule with a permanent spending increase would lead to hyperinflation. Other modelling choices exist, and they allow the rate of interest to respond to either bond issuance or money printing. At most, they allow for two years of interest peg.

Of course, economic actors might expect the government spending increase to continue permanently – as in the Romer-Bernstein brief – and that the spending increase is initially financed through debt issuance. Then the state-of-the- art models that derive from the classic Smets-Wouters model assume that any increase in debt used to finance the increase government spending is paid-off with higher interests by raising taxes in the future. Hence, interest rates rise as a response to new bond issuance and taxes also rise in tandem to pay off the newly-burgeoning government debt. In the interim, the NNS multipliers range between about one to about 0.4, almost a third of what the Old Keynesian Romer-Fitzgerald multiplier is. Thus, first the NNS multiplier is about three times smaller than the Old Keynesian multiplier, and second it turns negative as time goes by and as government purchases fall.

Now, modelling the fiscal stimulus as a permanent shock could be a good idea in order to set a benchmark, but the USD 787m stimulus signed into law on February 17 of 2009 was nothing of that sort. Today’s packages add up to USD 5tr, more than 6x the Obama package. In fact, The American Recovery and Reinvestment Act of 2009, as the full name goes, stipulated transfer payments and temporary tax cuts, and that added up to half of the package. Hence, at least half of the aforementioned stimulus was temporary rather than permanent. What now?


Biden has clearly understood that not only the economy is crumbling after the outbreak, but the social tissue is also worn out: America is falling apart. Well, it doesn’t stop there and it didn’t start with Biden. It may mean lots of things, ranging from political economy – old 19th century style – to electoral politics to disappointing U.S. post-recovery growth that draws a parallel with the 1930s to the fears the outbreak has unleashed. We better just stick to old fashioned economics that tell us only fiscal and monetary policies matter. Here are a few sets of graphics that might cipher the idea without theory and without politics.


The symmetry between GDP changes and M2 supply changes after 1929 strikes the eye, and it is not only a graphical illusion. Here is a second chart that strengthens the graphical vision; U.S. GDP as index and M2 supply as volume also moved in tandem in the 1930s, and not only percentage changes. We should only add to this picture late-coming public spending in the mid-1930s, which perhaps was a bit heavy handed even by the standards of Keynes if only because monetary policy had the wrong sign until 1933. Now we may add a grain of salt to this wishy-washy bit of visual evidence.

EXHIBIT 2: US GDP AND M2, 1929-1939

John Taylor – right after Lehman – provided evidence of the fact that the Fed’s policy rate was kept too low for too long a period prior to the crisis compared to what the Taylor Rule had implied, and that as an indicator of cheap and ample USD liquidity provision this was one of the causes of the crisis. He picked up from there and claimed the fiscal stimulus was too discretionary, sporadic and short-lived during the first phase after Lehman. Taylor claimed that, perhaps in a watered down new Keynesian fashion, fiscal stimulus intending to jumpstart personal consumption – checks sent to households and tax rebates – did not and couldn’t by themselves achieve that goal. Rather, disposable income rose, but did not spread to personal consumption effectively, vindicating life-cycle permanent income consumption models reminiscent of Robert Hall (1978). This shows in another and more oblique vein why the Fed has created about 20% of all existing dollars in 2020 alone. There has been a huge money supply increase in addition to rate cuts and enormous fiscal stimuli. Both the Fed and the American Administration have learned their lessons from 2008, and didn’t hesitate to use all that is in their power this time. This is why almost every observer now thinks a huge U.S. economic recovery is coming up, i.e. up to 6% GDP growth in 2021.

Leave a Reply

Your email address will not be published.