5
of real interest rates indicates that even as rates have partly rebounded from their post-crisis lows
they are unlikely to return to where they were expected to be prior to the crisis (CEA 2015;
Holston, Laubach, and Williams 2016).
The stronger form of secular stagnation argues that with low inflation, real interest rates cannot
fall low enough to restore aggregate demand as a result of the effective lower bound, leading to a
self-reinforcing spiral of weak economic performance and low interest rates. While I do not
believe the stronger form of the secular stagnation is a correct description of the United States or
Europe, the weaker form—that conventional monetary policy will be constrained more often in
the future—is certainly a source of concern (Furman 2014).
In 2000, David Reifschneider and John Williams estimated that the zero lower bound would be
constraining about 5 percent of the time in the United States, with a mean duration of four
quarters when rates hit the zero lower bound. However, the experience across the advanced
economies since the Great Recession suggests that, if anything, this estimate was overoptimistic.
As the authors clearly stated at the time, a key assumption in this result was that the equilibrium
real federal funds rate was 2.5 percent, the consensus view at the time. This is well above the
most recent projections from the members of the Federal Open Market Committee, which range
from 0.5 to 1.8 percent for the long-run real federal funds rate. Consequently, it is reasonable to
assume that the zero lower bound or effective lower bound will constrain conventional monetary
policy more than 5 percent of the time in the future (Dordal-i-Carreras et al. 2016).
3
And while
unconventional monetary policy can still operate, there is substantial controversy on its efficacy
and side effects—making other, complementary efforts to achieve the same goals desirable.
Principle 2: Discretionary Fiscal Stimulus Can Be Very Effective in Practice
For decades after World War II, the ability of fiscal policy to affect the economy was broadly
accepted (see, for example, Blinder and Solow 1973). In fact, the principal objection to the use of
fiscal policy was not that it did not affect the economy. It was, in fact believed, to do just that—
just that policymakers would do a bad job timing its impact, so that in practice it would add to
instability rather than reducing it (Friedman 1953).
A decade ago, however, even the basic premise underlying the earlier debate about fiscal policy
was increasingly under assault. On one side was the Ricardian view that rational, forward-
looking agents could effectively undo fiscal stimulus. In this view, what matters is a country’s
consolidated balance sheet, and if the government takes on more debt, this action would flow
through to private agents, who would in turn take on less (Barro 1974). On the other side was an
increasing focus on the side effects of fiscal stimulus in terms of higher interest rates and reduced
private investment (Ball and Mankiw 1995). In fact, one argument was that the 1990 and 1993
fiscal consolidations in the United States were actually expansionary (Blinder and Yellen 2001),
an argument that was subsequently generalized (Alesina and Ardagna 2010).
3
Changes to monetary policy rules could affect the frequency with which the effective lower bound is binding
(Goodfriend 2016; Williams 2016). But my argument applies to the degree that these policy rules have not changed;
to the degree that changing them is costly, so more active fiscal policy could obviate the need to incur those costs; or
to the degree that, even with the new rules, monetary policy still has limitations or side effects.