Commentary: Monetary Policy and the
Well-Being of the Poor
Alan S.Blinder
This paper is a notable achievement that derves a place in the Jackson Hole Hall of Fame.Christie and David Romer have man-aged to accomplish two things here.First,they actually add a new thought to the academic literature on the effects of macroeconomic performance on the distribution of income—a topic on which I wrote 20years ago1and on which,I thought before reading this paper,there was not much new to say.Second,they reach a conclusion that can-not fail to warm the hearts of central bankers everywhere:“that com-passionate monetary policy is,most likely,sound monetary policy.”But I am an academic now,not a central banker,so—mindful of that great tradition—I’ll begin with a quibble.My quibble actually precedes the first word of their paper.I’d like to suggest a change in the title.Christie and David claim to have written a paper on how monetary policy affects the well-being of the poor.But when you read it,you find out that it is really about how macroeconomic per-formance affects the poor.In fact,when you dig deeper,you find out that the result they most emphasize is,basically,that lower inflation rais a society’s average income.Whether or not it holds any spe-cial benefits for the poor is less clear.
知识点It is perhaps a sign of the times that the Romers cavalierly identify “monetary policy”with“macroeconomic performance,”as if they were one and the same.But they are not.Cyclically,such things as
203
abs是什么函数204Alan S.Blinder fiscal policy,other demand shocks,and supply shocks influence the economy’s short-run performance.Secularly,no one(certainly not the Romers)would suggest that either monetary policy or inflation is the main influence on long-run economic growth.
When I think about the effects of monetary policy,as oppod,say, to the effects of fiscal policy,on the distribution of income,I think of questions that relate specifically to interest rates,such as:
–Are the poor more likely to be debtors who pay interest or
creditors who receive interest?
端午吃什么–Do the ctoral effects of interest rates(for example,on the
shares of housing,consumer durables,and net exports in
GDP) affect the distribution of income?
–Do the intertemporal allocative effects of interest rates
(for example,on saving and investment versus consumption)
affect the distribution of income?
The questions do not engage Christie and David much,and per-haps for good reasons.So my quarrel is only with the title of the paper,not its content.
The paper carries two main messages.First,that the standard cyclical finding—that the poor are hurt by unemployment but not by inflation—is less important than is commonly believed.Second,that the poor have a strong long-run interest in price stability.I will take up each of the in turn,arguing that while the Romers are basically right on both counts,they overstate their ca.The lady and gentle-man doth protest too much,methinks.
The first part of the paper is about the effects of the business cycle on the economic status of the poor in the United States.It has been known for a long time that poverty ris and the share of low-income families falls when unemployment increas,while higher inflation has little effect on either.T
his finding is confirmed again by Romer and Romer though,puzzlingly,they find no systematic effect on the
Commentary205 shape of the income distribution.But they add an interesting twist, which was not emphasized by earlier authors.
Under the conventional view of the natural rate hypothesis,the average unemployment rate must equal the natural rate over any period of time during which the inflation rate does not change.If the poverty rate depends positively on the unemployment rate(as it does),it follows that any decline in poverty achieved on the upswing must be given back on the downswing.With a linear Phillips curve—which is a pretty good assumption for the United States—the swap is exact:What the boom giveth,the slump taketh away.As far as macroeconomic policy is concerned,then,there is a“natural rate of poverty,”corresponding to the natural rate of unemployment. As an85percent believer in the natural rate hypothesis,I basically accept this conclusion.But let me take a few minutes to e if we can escape from the Romers’Iron Law of Poverty,at least around the edges. One possibility,which they discuss but dismiss,is that inflation might change permanently.Consider first a period of disinflation, which most of our countries have recently been experiencing.To push inflation down,the central bank must hold unemployment above its natural rate,on average.So the poverty rate must als
o be above its natural rate.Hence,the basic conclusion from the older lit-erature is maintained:“The poor pay a disproportionate share of the burden when high unemployment is ud to wring inflation out of the system.”2This disproportionate share stands as an extra cost of dis-inflation.Now suppo,instead,that the central bank’s job is raising the inflation rate,as in contemporary Japan.Then the alleviation of poverty becomes an additional benefit from reflation.
意理My point is simply that the natural rate hypothesis does not rob cyclical findings of all interest,as Romer and Romer come very clo to suggesting.But they are surely right that it does rob them of quite a lot,and this is worth pointing out.
As a cond possible escape from the Iron Law of Poverty,let me rai the issue of hysteresis.Until quite recently,most of us felt
206Alan S.Blinder comfortable in dismissing hysteresis in unemployment as theoreti-cally interesting but empirically unimportant for the United States. The natural-rate Phillips curve just worked too well.But,as is well known,the United States is an outlier in this respect:Virtually no other country has such a well-behaved Phillips curve.For many other nations,especially in Europe,hysteresis—and thus a perma-nent tradeoff between unemployment and inflation—remains a live possibility.And recent data rai the specter of hysteresis even in the United States.
Furthermore,hysteresis might be prent in the relationship between unemployment and poverty.For example,perhaps a pro-longed tight labor market enables marginal workers to break out of the“cycle of poverty.”The Romers do not investigate this possibil-ity.In any ca,if there is hysteresis in either the Phillips curve or the relationship between unemployment and poverty,then a business cycle can leave a permanent imprint on the poverty rate. Finally,I must take issue with the way the Romers state the impli-cation of their cyclical conclusion for policy.Their penultimate para-graph states that:时间安排
“...the usual emphasis on the short-run effects of monetary policy
on poverty is fundamentally misguided.It is certainly true that
expansionary policy can generate a boom and reduce poverty tempo-
口香糖
李昊东rarily.But the effect is unquestionably just that—temporary.”
They suggest that this temporary effect is hardly worth worrying about.
I beg to differ.Change the word“poverty”to“unemployment,”which is legitimate according to their regressions,and the statement becomes:
“...the usual emphasis on the short-run effects of monetary policy on
unemployment is fundamentally misguided.It is certainly true that
墨梅王冕expansionary policy can generate a boom and reduce unemploy-
ment temporarily.But the effect is unquestionably just that—tem-
porary.”
Commentary207
Where have you heard that before?
As many of you know,I have never accepted the view that central bankers should ignore the effects of monetary policy on unemploy-ment simply becau they do not last forever.In fact,we have at this conference the greatest fine-tuner in history.I do not believe that the efforts of Alan Greenspan and the Federal Open Market Committee to limit recessions and reduce unemployment have been“fundamen-tally misguided.”On the contrary,I think America has benefited enormously from their success.
Turning now to the Romers’cular findings,let me begin by emphasizing a caveat that they correctly
enter before they analyze their multination,cross-ctional data.It takes a pretty big leap of faith,probably veral leaps,to believe that,say,the effects of infla-tion on the incomes of the poor are the same in all countries.Some nations do extensive indexing;others do not.Wage-tting institu-tions differ.So do tax-transfer systems.Furthermore,the institu-tions and others that bear on the distributional effects of inflation not only can but do adapt to changes in the inflation rate.To take a clo-to-home example,there are now many fewer escalator claus in wage agreements in the United States than there were in1980.
So I tend to take cross-country regressions with many grains of salt—unless the findings are extremely strong or there is good reason to believe that the relationship is roughly invariant across space.To my mind,a t of time-ries regressions,one for each country,that tells more or less the same story would be far more convincing. That said,look at Charts6and8,which show the Romers’main cular result:that low inflation is associated with higher incomes of the poor.(The direction of causation is unclear.)I pick the two charts becau I prefer removing the outliers;to e why,take a peek back at Chart4,which includes eight high-inflation countries omitted from Chart6.The data look to me like a nearly horizontal scatter for the eight high-inflation countries joined to a nearly vertical scat-ter for the remaining58.I e little virtue in joining the two datats.
