The First DisinflationInflation and Its Discontents
Michael J. Boskin
Luncheon address to the American Economic Association/American Finance Association Annual Joint Meeting, January 4, 1997, New Orleans, Louisiana.
I am pleased to be here today and honored to be included among such a distinguished list of former superstar speakers--Bob Merton, Steve Ross, Joe Stiglitz, Larry Summers, and Alan Greenspan, to name but a few. I am especially honored that you have invited me back for a second speech. I last spoke at this luncheon when I was chairman of the Council of Economic Advisers, and my topic at that time was "The Benefits of Financial Innovation." My conclusions at that time--that financial market innovation, much of it made possible by analytic developments in finance, such as Black-Scholes option pricing, have been enormously beneficial to the functioning of the American and world financial markets and economies and that the proclivity of the government to regulate should be limited primarily to disclosure and transparency--remain valid today. Let me also in this spirit congratulate Secretary Rubin and the Treasury for the soon-to-be issued indexed bonds.
Today I want to talk about another subject of immense importance both to finance and to the rest of economics--inflation. As we move from 1996 to 1997, history reminds us that inflation is ignored at our peril. This is the hundredth anniversary of William Jennings Bryan's "Cross of Gold" speech in the 1896 election; the seventy-fifth anniversary of Weimar hyperinflation and the publication of Irving Fisher's classic The Making of Index Numbers; the fiftieth anniversary of the first American double-digit inflation of modern times after the removal of price controls following World War II; and the twenty-fifth anniversary of President Nixon's wage and price controls. Perhaps it is particularly important for us to be thinking about inflation now, despite its relatively low level, to avoid complacency and future problems.
I've been thinking a lot about inflation lately or, more accurately, about how to measure it and the implications of its mismeasurement. I will return to that topic in a moment. I want to emphasize today how far we have come in understanding inflation, its costs, and its role in the long-term performance of the economy. Indeed, it would be difficult to find any subject about which the consensus of economists has changed more in the time since I was in graduate school.
At that time--the late 1960s and early 1970s--inflation was just starting to accelerate from the relatively benign levels of the 1950s to late 1960s. Milton Friedman and Ned Phelps were publishing their seminal works on the natural rate of unemployment. When I first came to Stanford University as an assistant professor, President Nixon was about to freeze wages and prices and to close the gold window, ending the Bretton Woods era. As Alan Blinder demonstrated, the wage and price controls actually worsened subsequent inflation.
At that time, in all but a few places, it was standard to teach an augmented Keynesian macroeconomics--there was no Lucas critique, let alone a whole school of rational expectations, no credit/liquidity constraint neo-Keynesianism. Of course, Milton Friedman and other monetarists emphasized the importance of low inflation and of rules rather than discretion in monetary policy, based in part on the perceived empirical stability of velocity.
More important, the conventional wisdom was that the cost of inflation was low, negligible in fact. A true giant of our profession--and one of the founders of modern finance--Jim Tobin, and his colleagues at the Cowles Foundation summarized this widely held view by saying that the cost of high inflation was primarily shoe leather--the need to hoof it down to the bank to get more cash (recall that this era predates ATM machines).
Of course, at that time we had experienced high inflation only temporarily, after the removal of price controls, usually after wars. Social Security and other budgetary outlays had not yet been explicitly indexed, and the tax brackets would not be indexed for another decade.
Finally, and most important, we had not yet experienced the trauma of the high, accelerating, and widely fluctuating inflation that occurred in the 1970s. Nor had we had the gross historical experiment of the 1970s, which brought rising inflation and unemployment simultaneously, something that caused all economists to rethink the relationship between inflation and unemployment.
There were, of course, those who had warned of the dangers of high inflation. Hayek, for example, emphasized that high inflation was inevitably followed by recession. High inflation was accompanied by speculative excesses and confusion in relative price signals that would have to be adjusted at significant cost.
The neutrality of money of classical economics was increasingly coming under attack from other quarters. For example, public finance economists, myself included but especially Martin Feldstein, emphasized that when tax rules are not perfectly indexed--the definition of income is nominal rather than real with historic cost depreciation, nominal capital gains and losses, and nominal interest, all taxed or deducted or both--higher inflation will lead to tax-induced distortions in capital accumulation.
Flash forward a decade, from circa 1970 to 1980. Inflation, as officially measured, is running at more than 13 percent. The overwhelming bulk of leading economists argues that we should learn to live with this high inflation, indexing more contracts and government programs, and attempt to stabilize the inflation rate at around 10 percent. The thinking at the time was that the substantial cost of disinflation was unbearable. There were, of course, people with different views--Tom Sargent and William Fellner, for example, who emphasized the possibility of a lower cost disinflation if the Federal Reserve policy was widely viewed as credible. I was in this camp and vividly recall a debate I had with the late Walter Heller on this issue in 1978--he still believed in jawboning as an effective inflation remedy!
When the Volcker Federal Reserve put on the brakes, velocity collapsed, and we had rapid disinflation accompanied by the worst recession since World War II. The cost of that disinflation in terms of lost output was substantially less than the simple Okun's law calculation implied. My own personal opinion is that the Federal Reserve has been gaining substantial credibility--gradually--ever since and that this is enormously valuable social capital that should not be readily squandered. Let's take a closer look at the two rounds of disinflation.
Starting with 13 percent inflation in 1980,
what would have been considered successful
inflation outcomes? Since inflation had risen
at corresponding points in each postwar business
cycle, the initial goal might be to prevent
this from happening again or to keep inflation in the
post-1982 expansion no higher than in the
197579 expansion, an average of more than 7
percent. Many would have thought this remarkable,
reversing a thirty-year trend. Perhaps the
United States could match its longer-term
historical average for the postBretton Woods era,
almost 8 percent, or do better than the inflation
forecasts of the main econometric
models--DRI, Wharton, and Chase--which were
predicting 8 percent or higher in the 1980s and
almost 7 percent in the 1990s (as of 1981).
Or perhaps the United States could match the best
of the major industrialized countries' postBretton
Woods experience--Germany's at more than
5 percent--or do better than the inflation
rates implicit in long-term Treasury Bond yields for
the period of the late 1970s and early 1980s
(using a conservative assumption of long-run real
interest rates implied an average inflation
rate of 6 percent in the 1980s and 1990s).
The Federal Reserve bettered each of these
commonsense goals in times of rapid financial
innovation, technical change and globalization,
which rendered the making of monetary policy
more complex. For example, the traditional
relationships between changes in reserve positions
and the money supply and gross domestic product
became less reliable, globalization rendered
traditional measures of capacity utilization
suspect, and technological change drove down the
prices of important inputs.
The Second Round of Disinflation
Inflation was reduced to the 4.5 percent range
for most of the 1980s, much lower than the
1970s but still at a level that caused a conservative
Republican president, Richard Nixon, to
impose wage and price controls. Worse yet,
that level risked becoming the floor, with nowhere
to go but up. Hence, the Greenspan Federal
Reserve, starting in early 1988, launched a second
round of disinflation, trying to achieve the
practical equivalent of "price stability" by engineering
a so-called soft landing. A combination of
factors--some not foreseeable when the Federal
Reserve began to raise interest rates, such
as Saddam Hussein's invasion of Kuwait in 1990
and the resulting oil price shock and an immense
and accelerated defense drawdown following
the collapse of the former Soviet Union, some
more predictable, such as the impending
recession in Europe and Japan and the credit
crunch--hardened the landing, but it was far less
severe than in any other major industrial
country, each of which had a far worse recession than
that of the United States. American economic
growth has been by far the strongest among the
Group of 7 nations since 1992, and inflation
remains low.
Relative to the economic problems that the
country has confronted over the past decade and a
half, the Federal Reserve has done a remarkable
job of engineering two rounds of disinflation
with considerably less cost than many thought
possible at the time, laying a foundation for
longer-lived expansions. Indeed, growth was
at least as strong in the decade of the 1980s, when
inflation fell by 9 percentage points, as
in the 1970s, when it rose by 9 percentage points (a
fact conveniently ignored by the decryers
of the 1980s).
The Volcker and the Greenspan Federal Reserves
have been, by a substantial margin, the stars
of economic policy in recent times (unfortunately,
this followed one of the worst Federal
Reserve performances by their predecessors
in the 1970s).
Although it is always dangerous to talk about
a consensus among economists, I believe it is fair
to say that most economists today would argue
that low and stable inflation is an important
pillar of maximizing long-run real economic
growth and rising living standards. That contrasts
remarkably with the notion that the cost of
inflation is trivial or that it is often desirable to
increase inflation to reduce unemployment.
We now know that is likely to be unsuccessful, as
we are likely to accelerate inflation without
permanently affecting the level of unemployment in
relatively normal times. We also know that
the higher level of inflation will not be neutral in the
long run but will damage capital formation
and growth.
Let me briefly mention something that has occupied
a disproportionate amount of my time not
only in the last eighteen months, when I chaired
the Congressional Commission on the
Consumer Price Index (CPI), but for many years,
as I've thought about and worked to try to
improve the quality of our economic statistics.
We know that our usual measures of inflation
overstate changes in the cost of living because
of the failure of some indexes, such as the CPI,
to allow for consumer substitution among commodities,
which partially insulates consumers
from increases in relative prices of goods.
Despite continuing improvements by government
statisticians, we are not keeping up with
the pace of introduction of new products and quality
change in existing products. Our CPI commission
concluded that this was an important source
of bias (and we did not even fully include
the Hicksian reservation price/consumer surplus for
new products). We focused primarily on introducing
new products--such as VCR's, microwave
ovens, and personal computers--a decade or
more after they entered the market and after their
price had fallen 80 or 90 percent. (Cellular
telephones will enter the index in 1998.) In a
complex, dynamic market economy with hundreds
of thousands of goods, many thousands of
which are being introduced, produced, and
consumed each year and others of which are
leaving the market; the quality change in
many products; and an ever-growing share of what
the nation produces and consumes in areas
where it is much harder to measure output as
compared to the traditional tons of steel
and bushels of wheat, these issues have profound
ramifications.
If our CPI commission's estimate--which we
believe to be prudent, even conservative--of an
overstatement of about 1.1 percentage points
a year is correct, it implies that the federal budget
will be an extra trillion dollars in the red
over the next dozen years just as a result of this
overstatement, which by itself will be the
fourth-largest item in the budget after Social Security,
health care, and defense. It also means that
we are mismeasuring economic progress, not just
misstating the contemporaneous inflation rate.
Instead of falling in the last quarter century, real
wages have risen--obviously, more slowly than
in the previous quarter century, but risen
nonetheless. Likewise, real median family
income has gone up modestly, rather than stagnated.
And the poverty rate would be closer to 9
percent than almost 14 percent. Of course, there are
issues in the other aspects of measuring each
of these variables--fringe benefits, bonuses,
compositional shifts such as changes in family
size, and so on.
The finance types in the audience may now look
forward to an expanded Ibbotson(1) with an
adjusted inflation measure to help you get
to real rates of return, inflation risk premia, and so
on.
But getting back to my main theme, inflation--at
least as approximated by a cost-of-living
index, which must be the starting point for
measures of a consumer and social welfare notion
of why we want to keep inflation low and stable--is
now running around 2 percent a year. Our
estimate of the bias is roughly invariant
to modest changes in the rate of inflation, that is, about
the first percentage point or so of measured
inflation is really other things rather than true
inflation--for example, quality change and
introduction of new products. Thus, if the inflation
rate accelerates, as conventionally measured,
from 3 to 5 percent, it is really likely accelerating
from 2 to 4 percent. As you can infer from
my earlier remarks, except under extremely unusual
and rare circumstances, I would expect the
Federal Reserve to be as diligent against a rise in
inflation from 2 to 4 percent as against that
from 3 to 5 percent. Keeping inflation low and
steady is important and a major achievement.
It is also a major achievement of our
profession--perhaps somewhat forced on us
by historical events in addition to our own
analysis--that we now view low and stable
inflation as an input to longer expansions and
stronger growth, rather than an impediment
to some other social goal such as full employment.
Let me conclude, then, by saying that although
much progress has been made, we must not be
complacent. Increasingly we hear what I would
consider to be a complacent attitude about the
economy--accepting too modest economic growth,
wishfully believing that the business cycle is
obsolete--rather than demanding fiscal, regulatory,
legal, and tax policy changes that could
improve long-run economic performance. (I
remember a paper titled "The Business Cycle is
Obsolete" written in the late 1960s just before
the outbreak of higher inflation and a subsequent
recession.) Surely our private economy has
developed various institutions and behavior that in
many ways will mitigate some types of shocks
to the economy. And, it is hoped, our
policymakers will continue to do a better
job than their predecessors a generation or two or
three ago. We also hear that the Federal Reserve
should be probing much lower interest rates
to see if we can get the unemployment rate
down still further and that, if inflation heats up, we
can deal with it later. Although I believe
the economy can and should grow more rapidly, and
in fact is doing somewhat better than the
conventional statistics report--after all, the problems
in the consumer price index I mentioned spill
over into the national income accounts, perhaps
lowering real gross domestic product growth
about a half a percentage point from a more
accurate measure--we certainly don't want
to let the inflation genie out of the bottle. We've
tried that in the past, and it has caused
serious problems, wreaked enormous hardship, which
took more than a decade to work out of the
economy.
I also get suspicious whenever I hear some
new law of economics that seems to violate known
historical experience. Let me close with an
example. How many people here remember the
phrase "The coasts are recession proof"? That
was the phrase used to describe the economy in
the 1980s because the recession of the early
1980s, which ravaged the industrial Midwest, the
oil patch, and the farm belt, left the coasts
relatively unscathed. I said at the time, and I always
firmly believed, that such a view was simply
nonsense. Well, unfortunately, residents of
California and the Northeast discovered that
they were not recession-proof and were clobbered
at a pace much worse than the national average
in the early 1990s.
So, in conclusion, while cautiously optimistic,
let us remain vigilant. Low and stable inflation
has been hard-won at substantial cost, and
the Federal Reserve's credibility should not be
squandered. We must seek to raise the nation's
long-run growth potential--even a few tenths of
a percent a year will make an enormous difference
to long-run improvements in living
standards. But let us seek the policy changes
where they are likely to strengthen noninflationary
growth: a simpler, broader, flatter rate tax
system less inimical to saving, investment, and
entrepreneurship; slowing the growth of government
spending, reforming entitlements, and
moving to budget balance at least on average
over business cycles; less and more-flexible
regulation; expanded open rulesbased
trade; incentives in the civil justice system to reduce the
prevalence of frivolous lawsuits dramatically;
private market-style reform incentives to
strengthen and improve elementary and secondary
education and job training. If these reforms
together with the natural evolution of the
private economy can raise the nation's growth
potential substantially--and I believe they
can--I have no doubt the Federal Reserve can and
will be able to adjust monetary policy accordingly
to accommodate the more rapid
noninflationary growth.
NOTES
1. See R. Ibbotson, Stocks, Bonds, Bills and
Inflation Yearbook, 1997, which is the classic
historical reference on bond yields, stock
returns, and so forth.