POLICY NOTE:
WHY DOES THE FED OPPOSE CONTINUED GROWTH?
DRAFT
L. Randall Wray,
Professor of Economics, University of Missouri-Kansas City, Jerome Levy
Economics Institute Visiting Senior Scholar, and Senior Research
Associate, Center for Full Employment and Price Stability
In June, a
variety of newspaper headlines proclaimed Good News All Around: Largest
employment losses in nine years! Retail sales down again! Home starts
collapsing! Manufacturing output down for third straight month! Debt
service burdens up, delinquencies and defaults rising! Even imports are
down! Can we stand any more good news?
The Fed has
raised interest rates six times over the past year in an effort to slow
economic growth. Mounting evidence shows the economy has slowed
perceptibly in recent months. While the FOMC chose to leave interest rates
alone at its last meeting (June 28), it threatened to raise them again in
the future if robust growth resumes. I suspect that growth during the rest
of this year will be sufficiently sluggish that the Fed will not increase
rates further. However, it must be remembered that the Greenspan Fed
continued to raise rates in 1989 as the economy slowed into our last
recession.
It is becoming
increasingly obvious that the six rate hikes so far this year resulted
from a colossal misreading of the economy by the Fed. Inflation was never
around some corner or other. The US expansion was self-limiting because it
was fueled by unsustainable private sector deficit spending–mainly by
households that were spending more than their incomes by a record amount.
Early this year, households began to retrench, cutting borrowing and
spending. The Fed’s rate hikes played a role in hastening the inevitable
by increasing debt service burdens. While most analysts believe the Fed
can engineer a soft landing by gradually lowering rates as the economy
slows, I think it is much more likely that we will suffer a Japanese-style
hard landing.
THE FED’S RATIONALE
While the Fed
has long been (legitimately) concerned with the “irrational
exuberance” of stock market excesses, I do not believe that interest
rates have been raised to cool equity prices. Indeed, if anything, I
expect that concerns over effects on Wall Street have moderated Fed rate
hikes. Instead, I believe Chairman Greenspan when he says he is worried
about inflation and suspect that rate hikes would be discontinued if the
Fed became convinced that inflation risks had subsided. This is not to say
that there aren’t other forces that push the Fed to action. An active
Fed is a watched Fed. Since 1989, the Greenspan Fed has changed the fed
funds rate 45 times–or, on average, once every quarter. Frequent, small,
interventions keep the Fed in the limelight and remind the press that
Greenspan is the Second-Most-Important-Man-in-the-World.
The Fed’s model runs
something like this. High aggregate demand growth must eventually cause
inflation because growth of potential supply is limited to growth of the
labor force plus productivity growth. The current expansion has proceeded
at an unsustainably high pace, however, inflation has not yet reared its
ugly head because potential supply was, until recently, able to keep
pace–partially because we began with a large stock of unemployed and
partially because productivity growth has been well-above its post 1973
trend (the New Economy and all that). We are now living on borrowed time,
with aggregate demand outstripping supply–as evidenced by unemployment
rates that remain too low. While there is no strong evidence that
inflation is accelerating, monetary policy works with a lag so it must be
used preemptively. Interest rate hikes reduce borrowing and thus spending
along a more-or-less smooth demand for loans curve. Interest rate effects
are largely independent of the state of the economy and proportional to
the size of the change.
This
model has scant empirical support and it stands on shaky theoretical
ground. While there is probably some link between unemployment rates and
the rate of growth of wages, there is little support for the belief that
low unemployment leads to inflation. Since 1970, every case of rapid
acceleration of inflation has taken place during high unemployment
periods. In contrast, our low unemployment periods–during the 1960s and
late 1990s--have come during periods of robust growth and low inflation.
Inflation in the US has always been driven by three components of the
consumer basket: energy prices, food prices, and, most importantly,
imputed housing costs. (See Papadimitriou and Wray.) Note the
keyword–imputed. About half of US inflation is due to the housing
component of the CPI, and that does not measure what is actually happening
to real estate prices. Rather, it is constructed from a survey that asks
homeowners to make a guess about how much rent they might receive if they
were to put their homes onto the rental market. There is no reason to
suppose that this guess varies systematically with tightness in the market
for labor to build new houses–or with anything else the Fed ought to be
worried about.
The
notion that low unemployment and rising wages ought to cause inflation has
a lot of intuitive appeal, but it ignores supply responses and price
constraints. Low unemployment and strong economic growth means firms
operate closer to capacity, hence, more efficiently. Firms adopt labor
saving-techniques and technology. Labor productivity rises. Cheaper
intermediate goods can be imported. Greater sales on a reduced markup over
wage costs allow firms to maintain profits and market share in spite of
higher wage costs. Cheap imports–especially in the last two
decades–prevent domestic producers from raising prices. Thus, it is not
surprising that there is no strong empirical relationship between low
unemployment and rising inflation–and, often the correlation goes the
other way because low demand causes firms to operate inefficiently,
placing pressure on prices.
THE
LIKELY IMPACTS OF THE RATE HIKES
The
six rate hikes thus far have already hurt the economy, and the Fed might
increase interest rates further later this summer. The main impact this
will have on the economy will come later–hindering recovery. If the
economy were strong, interest rate hikes alone would have little effect.
However, let me address the belief that the Fed ought to tighten in order
to raise unemployment to fight prospective inflation.
The
old “reserve army of the unemployed” argument that underlies all of
Greenspan’s musings on the necessity of maintaining high unemployment
rates in order to keep labor insecure holds appeal. But it is flawed.
Monthly movements of the official unemployment rate make good press but
provide little information. Most of those who would like to work do not
get counted. Even Greenspan has argued that there are actually 10 million
people who would like to get jobs–4 million more than the official
count. But that is still too low–the actual number is closer to 16
million. (See Pigeon and Wray.) If we are to take Greenspan at his word,
something like 10-16 million people must remain unemployed to keep
inflation down. What makes this particularly disturbing is that
unemployment is never equally distributed. It is always disproportionately
visited on the disadvantaged–low skilled, lowly educated, minorities,
last-hired-first-fired.
It would be nice if we could
convince them, and ourselves, that while it is really a shame that they
can’t find jobs, they’ve got to do their civic duty by going without
in order to keep the employed sufficiently insecure. But that is a lie
because they are not close substitutes for the highly educated, highly
skilled workers now in short supply. Yes, the US might be facing a
shortage of experienced electrical engineers, but does anyone seriously
believe that can be resolved by raising the unemployment rate of innercity
teens, or other disadvantaged groups who are enjoying the only half decent
job market in a generation and a half? I do believe that the low
unemployment rates have helped to reverse declining wages at the very
bottom of the labor market, but this is not going to spur general
inflation, nor can rising unemployment at the bottom help to relieve
bottlenecks at the top.
The costs of unemployment far
exceed the lost income that one suffers while unemployed precisely because
unemployment is visited disproportionately on those who are already
disadvantaged and because long spells of unemployment generate large
social costs in terms of higher divorce, suicide, crime, and mortality
rates. It is a principle of public policy that government should not adopt
policies that cause concentrated burdens on a small subset of the
population in order to generate benefits that are widely spread. If
unemployment were equally shared across the entire population, then it
might be acceptable to use unemployment to fight inflation, the benefits
of which would be shared by those who bear the burden. However, this is
manifestly not the case. Until unemployment reaches levels not seen in the
USA since the 1930s, it will be concentrated in certain easily
identifiable, disadvantaged, groups.
The
Fed’s policy is also based on faulty understanding of the impacts of
rising interest rates–the tool used to try to increase unemployment
among innercity teens. Interest is a cost imposed throughout all markets,
much like an OPEC-induced increase of energy costs. Thus, the initial
effect is quite similar to the effects of rising oil
prices–inflationary. Business firms normally finance day-to-day
operating expenses through short term loans; when interest rates rise,
their finance costs immediately rise, and to the extent that they are
able, they pass these along in the form of higher prices. Again, their
ability to raise prices is mitigated by competition from foreign firms
that are not impacted by rising US interest rates; further, they will also
try to cut other costs to
compensate for rising interest costs. Over the longer run, firms might
decide to reduce some kinds of spending, although so long as demand
remains high it is unlikely that spending by firms will be much affected.
It is ironic that the Fed stands ready to fight rising wage costs because
of their supposed inflationary impacts by raising interest costs, which
increases business costs in a much more direct and universal manner than
do scattered wage increases.
While
conventional wisdom presumes that a primary effect of tight monetary
policy is through the investment channel, there is virtually no empirical
evidence that higher interest rates lower investment. This isn’t
surprising, as US firms mainly use retained earnings to finance
investment. (See Fazzari.) Conventional wisdom also holds that consumers
cut back spending on big ticket items as interest rates rise. While there
may be some impacts on origination of new mortgages, even this effect is
overstated when variable rate mortgages exist as an alternative. What
matters to households is the monthly payment, not the interest rate. If
incomes are rising, or if debt loads are relatively small, rising interest
rates have little effect because households can make their monthly
payments without too much trouble. However, if rates rise enough and if
households are sufficiently indebted, there is a large “net income
effect”(income that is left after servicing debt is reduced). This means
that interest rate effects depend critically upon the amount of debt
carried.
What
all this means is that there is no nice smooth credit demand schedule. It
is more appropriate to think of the process as one in which interest rate
hikes eventually push households to a debt service threshold, which varies
depending on debt loads. Further, the crunch, when it comes, will be much
more severe when the debt load is high because default risk is increased.
As credit quality erodes, banks begin to retrench even as their own
balance sheets erode–making recovery much more difficult. In recent
decades the heaviest debt service load was reached in the late 1980s; it
is not surprising that the Fed’s attempt to engineer a soft landing in
1989 led to a deep and prolonged recession accompanied by a credit crunch
that hampered recovery. Although debt service ratios today are slightly
lower than those reached at the 1980s peak, this is only because interest
rates are lower today (debt ratios are much higher–indeed, they are
easily at an all time high). As interest rates rise in coming months, debt
service ratios will rise until households are no longer able to service
their debt. The coming credit crunch could be more severe than the one
experienced in the early 1990s.
IS
INFLATION REALLY AROUND THE CORNER?
The
most recent evidence on inflation, from the PPI release of June 9, found
that the index for finished goods prices in May was unchanged. It is true
that this result was derived from large declines in the prices of foods
and energy. If we exclude those, prices rose by 0.2%, or something like a
2.5% annual rate. That is exactly where measured inflation has been stuck
over almost the whole Clinton expansion.
As
the BLS said in the PPI release, measured inflation for May included
“price increases for passenger cars, sanitary papers and health
products, book publishing, light motor trucks, cosmetics, and women’s
apparel”, which “outweighed declining prices for prescription drugs,
alcoholic beverages, household appliances, passenger car radial tires, and
for girls’, children’s, and infants’ apparel.” So sanitary paper
prices rose while prescription drug prices fell. Is that a clear case that
we risk runaway inflation if the Fed doesn’t act now? If prices went up
significantly across the board, we could probably all agree that there was
inflation. But that is almost never the case. In reality, some prices are
rising and some are falling, and the PPI or CPI weights these in
particular ways based on best guesses made at particular points in time
that rapidly become increasingly wrong. For example, the CPI uses relative
weights based on a typical consumer basket that is updated once every
decade or so, and as we move further from the update, the relative weights
used increasingly misrepresent actual consumption baskets.
It must always be remembered that the CPI and other measures of inflation
are calculated from constructed indexes. They are highly imperfect
measures, easily influenced by transitory events and arbitrary decisions
of statisticians. Particularly during periods of rapid changes in product
markets, measurement errors are large. Changes to the CPI that are below
something like 4 or 5% should probably be ignored–maybe not even
reported–because the noise to signal ratio is too high to provide for
any reasonable interpretation. The lower the measured inflation rate, the
greater the uncertainty regarding its meaning. A zero inflation target
should be dismissed out of hand merely for those reasons.
However,
it should also be dismissed because there is no theoretical or empirical
justification for the belief that moderate inflation is sufficiently
costly to warrant concern. Certainly, there is no strong evidence that CPI
measured inflation of 5% or even 10% has real costs in terms of aggregates
like GDP growth, investment, or consumption. Some careful cross-country
analysis has shown that inflation rates of even 40% have no discernible
impact on growth; a very recent study–by economists at the BOG, no
less–found that moderate inflation in the US is actually good for
investment and output. (See Ahmed and Rogers as well as Bruno and
Easterly.) Leaving aside the long-term historical evidence, it simply is
not plausible to argue that
the US faces accelerating inflation in the current conditions that include
massive excessive industrial capacity all over the globe–especially in
low wage competitor nations–and a strong dollar.
Finally,
as discussed above, even if inflation is costly, it is not reasonable to
fight it by raising unemployment. This is because the benefits of low
inflation–if they exist–are presumably shared across society, while
the costs of fighting inflation by raising unemployment are borne by
specific groups. If inflation is to be fought, this must be done in a
manner such that costs are shared across society–ideally by all those
who share in the benefits of low inflation. I’m not sure that economists
are anywhere near to formulating such policy. Given our current state of
knowledge, we should move carefully. Empirical evidence for the US
certainly does not support the case for raising unemployment to fight
inflation, for, as discussed above, we have never experienced accelerating
inflation when unemployment was low. It is particularly ironic that most
analysts support frequent, disruptive, intervention by the Fed to change
interest rates a quarter century after “fine-tuning” by government
policy was rejected. Would anyone today support quarterly changes to
federal tax rates in order to attempt to fine-tune the economy?
Raising
interest rates to fight inflation is actually a pretty bizarre policy, if
you think about it. It is like a tax on borrowing that hits firms that use
short term loans to pay wage bills and other day-to-day operating costs,
as well as indebted households. It increases government spending on
interest (reducing the current budget surplus). It increases income of
those who hold financial assets–who might be expected to increase
spending. Thus spending by rentiers and by the federal government actually
rise, at least partially offsetting any reduction of spending by indebted
firms and households. This is why interest rate effects must work mostly
by causing insolvency and bankruptcy. The impacts, as discussed above,
depend on the distribution and level of the debt burden, thus, are
inherently difficult to ascertain.
CONCLUSIONS
For
quite some time, it has been clear that the dangers facing the US economy
are mostly of the downside variety–and they are mostly neither the fault
of the Fed nor can the Fed do much about them. The stock market boom ended
long ago: The DOW and S&P have been essentially flat since last
November, while Nasdaq is down substantially since March. This is not
necessarily a bad thing because, as Greenspan apparently noticed, the US
equity market irrational exuberance probably represents the biggest
speculative bubble in all of human history. However, more importantly, it
was the stock market that was supposed to be driving consumption through a
“wealth effect” as richer households borrowed against capital gains to
fuel high living. What no one seems to recognize is that if the market has
gone horizontal, there are no capital gains to borrow against, hence, the
consumption boom should be over. And, indeed, recent data do seem to
indicate that spending trends are down. Recent projections have GDP growth
at just 4.5% for the entire year, far below the forecasts of only a month
ago. In May, manufacturing output dropped for the third straight month.
Home sales fell significantly in April. The unemployment rate rose to 4.1%
in May, from 3.9% in April–yes, this has something to do with
competition by the Census Bureau, but the private sector’s largest
employment losses in 9 years ought to give one pause.
Many, of course, welcome all
this bad news. Wall Street temporarily rebounds on every hint that the
economy is faltering. The Financial Times heralded the various
releases of the first week of June with the headline “Encouraging Signs:
The figures suggest that the Federal Reserve may finally have raised
interest rates enough to slow the US economy.” Earlier, the FT had
reported in another headline that “OECD looks to Fed for sharp rise in
interest rates: Big increase is needed by August to avoid global downturn,
report claims.” Not to be left behind, the Wall Street Journal
had suggested in a headline on June 5 that the “Fed’s Rate increases
Seem to have Vented The Economy’s Steam”, with a highlight indicating
that steel industry prices are “softening, while inventories are
starting to pile up”. The common thread, of course is the recognition
that May fears of overheating were in error. More relevantly for our
purposes is the common assignment of credit for the slowdown to the
Fed’s six rate hikes.
In
reality, the downturn was inevitable, with or without rate hikes. The
basic problem is not monetary policy, but tight fiscal policy in the
context of a large and growing trade deficit. In other words, the main
contributing factor to the coming recession is the very tight fiscal
policy which has meant that expansion could proceed only on the basis of
unprecedented private sector borrowing. To some extent, this may have been
fueled by the “wealth effect” induced by the unsustainable stock
market boom. I suspect that stagnant wages over most of the past 30 years
has also contributed–households really had no choice but to borrow in
order to attain rising living standards. When unemployment rates dropped
and remained relatively low, consumer confidence rose sufficiently that
record borrowing could be achieved. However, there was little danger that
this would overheat the economy–given excess capacity in the US and
abroad and given the weak position from which labor can bargain both wages
and prices have been severely constrained.
Further,
the boom was self-limiting because of the stance of fiscal policy–which
has become so biased toward surpluses that low unemployment increases tax
revenues, sucking away private sector income. The great danger is that
even if the economy turns down, the budget will remain in surplus, making
recovery impossible. This is exactly what happened in Japan, which
maintained a government surplus for six years–and continued to run
surpluses even during Japan’s horrible recession. Note also that even
though Japan’s interest rate has been kept essentially at zero for more
than 4 years, this has not contributed in any significant way to an
expansion. This ought to give Greenspan’s supporters pause when they
argue that the Fed can engineer a soft landing. We can only hope that our
landing will not be as “soft” as Japan’s.
For Further Reading
Inflation and the
great ratios: Long term evidence from the U.S., Journal of Monetary
Economics, 45, 2000, pp. 33-35, Shaghil Ahmed and John H. Rogers.
Inflation crises
and long-run growth, Journal of Monetary Economics, 41, 1998, pp. 3-26, M.
Bruno and W. Easterly.
The
Investment‑Finance Link, Investment and U.S. Fiscal Policy in the
1990s, PPB No. 9,1993, Steven M. Fazzari