This site has moved to
The posts below are backup copies from the new site.

May 21, 2007

Why Republicans are Skeptical about Global Warming

Jonathan Chait analyzes Republican opposition to the idea that global warming exists, that it is caused by humans if it does exist, and to doing anything about it:

Why the right goes nuclear over global warming, by Jonathan Chait, Commentary, LA Times: Last year, the National Journal asked a group of Republican senators and House members: "Do you think it's been proven beyond a reasonable doubt that the Earth is warming because of man-made problems?" Of the respondents, 23% said yes, 77% said no. In the year since that poll, ...[t]he U.N.'s Intergovernmental Panel on Climate Change released a study, with input from 2,000 scientists worldwide, finding that the certainty on man-made global warming had risen to 90%.

So, the magazine asked the question again last month. The results? Only 13% of Republicans agreed that global warming has been proved. As the evidence for global warming gets stronger, Republicans are actually getting more skeptical. Al Gore's recent congressional testimony on the subject, and the chilly reception he received from GOP members, suggest the discouraging conclusion that skepticism on global warming is hardening into party dogma. Like the notion that tax cuts are always good or that President Bush is a brave war leader, it's something you almost have to believe if you're an elected Republican.

How did it get this way? The easy answer is that Republicans are just tools of the energy industry. It's certainly true that many of them are. Leading global warming skeptic Rep. Joe L. Barton (R-Texas), for instance... The bottom line is that his relationship to the energy industry is as puppet relates to hand.

But the financial relationship doesn't quite explain the entirety of GOP skepticism on global warming. For one thing, the energy industry has dramatically softened its opposition to global warming over the last year, even as Republicans have stiffened theirs.

The truth is more complicated — and more depressing: A small number of hard-core ideologues (some, but not all, industry shills) have led the thinking for the whole conservative movement.

Your typical conservative has little interest in the issue. Of course, neither does the average nonconservative. But we nonconservatives tend to defer to mainstream scientific wisdom. Conservatives defer to a tiny handful of renegade scientists who reject the overwhelming professional consensus.

National Review magazine, with its popular website, is a perfect example. It has a blog dedicated to casting doubt on global warming, or solutions to global warming, or anybody who advocates a solution. Its title is "Planet Gore." The psychology at work here is pretty clear: Your average conservative may not know anything about climate science, but conservatives do know they hate Al Gore. So, hold up Gore as a hate figure and conservatives will let that dictate their thinking on the issue.

Meanwhile, Republicans who do believe in global warming get shunted aside. ...Gannett News Service recently reported that Rep. Wayne Gilchrest asked to be on the Select Committee on Energy Independence and Global Warming. House Republican leader John Boehner of Ohio refused to allow it unless Gilchrest would say that humans have not contributed to global warming. The Maryland Republican refused and was denied a seat.

Reps. Roscoe Bartlett (R-Md.) and Vernon Ehlers (R-Mich.), both research scientists, also were denied seats on the committee. Normally, relevant expertise would be considered an advantage. In this case, it was a disqualification; if the GOP allowed Republican researchers who accept the scientific consensus to sit on a global warming panel, it would kill the party's strategy of making global warming seem to be the pet obsession of Democrats and Hollywood lefties.

The phenomenon here is that a tiny number of influential conservative figures set the party line; dissenters are marginalized, and the rank and file go along with it. No doubt something like this happens on the Democratic side pretty often too. It's just rare to find the phenomenon occurring in such a blatant way.

You can tell that some conservatives who want to fight global warming understand how the psychology works and are trying to turn it in their favor. Their response is to emphasize nuclear power as an integral element of the solution. Sen. John McCain, who supports action on global warming, did this in a recent National Review interview. The technique seems to be surprisingly effective. When framed as a case for more nuclear plants, conservatives seem to let down their guard.

In reality, nuclear plants may be a small part of the answer, but you couldn't build enough to make a major dent. But the psychology is perfect. Conservatives know that lefties hate nuclear power. So, yeah, Rush Limbaugh listeners, let's fight global warming and stick it to those hippies!

The thinking may have been led by a few, but they found many willing followers. I think the influence of business in the GOP, not just the energy industry, is a factor. The fear is that any policy to address global warming will require business to implement costly changes, or, in the case of unilateral action by the U.S., reduce competitiveness causing profit to fall. Thus the policies, and even the idea the global warming exists are resisted. With Libertarians joining them based on their general opposition to any government interference, opposition has become, as Jonathan notes, part of the party's core principles.

Update: Brad DeLong says Jonathan Chait should ask a deeper question:

Why have the industry shills and the hard-core ideologues led the thinking for the whole conservative movement? They have led the thinking because the energy industry has funded them.

Coordinated Capitalism and Beyond

A review of Barry Eichengreen's The European Economy Since 1945: Coordinated Capitalism and Beyond:

Boom and Bust, by Sheri Berman, Book Review, NY Times: Postwar European history falls neatly into two periods. From 1945 to 1973, the countries of Western Europe recovered rapidly from the almost unimaginable devastation caused by World War II and then took off, growing faster than the United States and more than twice as fast as their own historical trends. From 1973 to the present, however, their economies have struggled with low growth and high unemployment...

As a result of this divided history, the so-called European model has both cheerleaders and naysayers. Social democrats and others on the left focus on the first period, applauding the continent’s ability to generate high living standards while cushioning individuals and societies from the ravages of unfettered markets. Right-wing critics and free marketeers focus on Europe’s contemporary problems, arguing that the continent’s generous welfare benefits and heavy regulation condemn it to continuing decline.

Both views contain some truth. But since the same conditions that led to success in one era have produced problems in the next, neither interpretation fully explains the story. In “The European Economy Since 1945,” Barry Eichengreen ... of Berkeley presents not only a comprehensive account of Europe’s postwar economic experience but also an important analysis of capitalist development more generally.

Drawing on his credentials as both an economist and a political scientist, Eichengreen argues that the key ... lies in recognizing that the recipe for growth varies, depending on one’s position in the economic race. In the years after 1945, Europe needed to recover from the war and catch up with the United States. This involved what economists call “extensive growth”... After the war, Europe developed a variety of institutions well suited to these tasks.

Large trade unions, employer organizations and corporatist arrangements, Eichengreen shows, helped labor-market partners reach and sustain long-term agreements to limit wage demands, ensure high levels of investment and plan for routine industrial restructuring. Unions agreed to hold down labor costs and in return were given either representation on corporate boards (Germany), influence over government planning and policy making (Sweden and France) or the ability to dole out government jobs and funds (Austria). Strong welfare states helped cement this bargain, providing workers with generous benefits to offset their wage restraint and the unemployment generated by industrial restructuring. ... And the government bureaucracies of nationalized industries helped mobilize and coordinate the resources necessary for the relatively clear-cut tasks associated with catch-up growth.

All this worked just as it was supposed to, generating prosperity across the continent. By the early 1970s, however, the potential for extensive growth had been largely exhausted. Europe’s businesses and infrastructure had been rebuilt, its labor force transferred from agriculture to manufacturing, the latest technology imported and adopted. At this point, Eichengreen says, “the continent had to ... switch from growth based on brute-force capital accumulation and the acquisition of known technologies to growth based on increases in efficiency and internally generated innovation” — that is, to “intensive growth.”

The problem, of course, was that Europe was now saddled with institutions ill suited to the creativity and flexibility that intensive growth demands. As Eichengreen puts it, “the continent’s very success at exploiting the opportunities for catch-up and convergence after World War II doomed it to difficulties thereafter.” The new situation called for flexible and mobile work relationships, technological novelty and the financing of risky ventures — none of which Europe’s postwar institutions were good at. ... (Eichengreen adds that a similar dynamic has played out in Eastern Europe, since Soviet institutions were not bad at extensive growth but awful at intensive growth.)

Eichengreen backs up his argument with reams of data and detailed examples... He reminds us that economic development calls for much more than simply the unleashing of market forces; it demands institutions capable of generating the resources, skills and relationships necessary to handle the particular economic challenges a country has to face at a particular time. And by demonstrating how institutions helpful in one era can become counterproductive in another, Eichengreen has important lessons about the future to teach both policy makers and publics.

“The European Economy Since 1945” should make readers wary of universal prescriptions for economic policy, since it shows how the fit between policies and circumstances is clearly ... important...

So what should the nations of Europe do now that the advantages of their “economic backwardness” have been fully exploited? Without settling the debate between the European model’s supporters and detractors, Eichengreen suggests that international competition is compelling Europe to abandon its distinctive model and become more flexible.

This will not be easy. Eichengreen himself stresses the difficulty of institutional change... Yet thanks to political will and creative policymaking, as Eichengreen points out, some countries on the continent, particularly the Nordic ones, have managed to adapt successfully. They are keeping themselves internationally competitive even while continuing to provide social benefits in health, education and social insurance far above American standards. Others, like France and Germany, will have to follow their lead. Otherwise, they will probably face the decline the pessimists have long been predicting.

Here's the Economist's review of the book. In the debate between the Sachs and Easterly camps over economic development, this is worth remembering:

[Eichengreen] reminds us that economic development calls for much more than simply the unleashing of market forces; it demands institutions capable of generating the resources, skills and relationships necessary to handle the particular economic challenges a country has to face at a particular time.

Update: Here is the first chapter.

Frederic Mishkin: Inflation Dynamics

If you are interested in inflation dynamics and recent research suggesting that inflation dynamics have changed, this speech by Federal Reserve governor Frederic Mishkin looks at three important questions:

  1. What is the available evidence on changes in inflation persistence in recent years?
  2. What is the available evidence on changes in the slope of the Phillips curve?
  3. What role do other variables play in the inflation process?

The speech summarizes and interprets research on these questions, and discusses the impact of the research on the conduct of monetary policy. As I've explained before, I am in agreement with his conclusions about the role of policy in anchoring expectations and how this has changed estimated inflation dynamics, and what this implies for monetary policy and inflation targeting. Too bad my monetary theory and policy class ended last week - I can't make them read this [Update: Brad DeLong provides a nice summary of the sections discussing the role of policy in anchoring expectations]:

Inflation Dynamics, by Frederic S. Mishkin, Federal Reserve Governor: Under its dual mandate, the Federal Reserve seeks to promote both price stability and maximum sustainable employment.[1] For this reason, we at the Federal Reserve are acutely interested in the inflation process, both to better understand the past and--given the inherent lags with which monetary policy affects the economy--to try to forecast the future. We economists have made some important strides in our understanding of inflation dynamics in recent years. To be sure, substantial gaps in our knowledge remain, and forecasting is still a famously imprecise task, but our increased understanding offers the hope that central banks will be able to continue and perhaps even improve upon their successful performance of recent years.

Today, I will outline what I see as the key stylized facts that research has in recent years uncovered about changes in the dynamics of inflation and will present my view of how to interpret these findings. The interpretation has important implications for how we should think about the conduct of monetary policy and what we think might happen to inflation over the next couple of years. I will address these two issues in the final part of the talk.

Before I continue, however, I should stress that the views I will express here are my own and are not necessarily those of my colleagues on the Federal Open Market Committee (FOMC).

The Empirical Evidence on Changes in Inflation Dynamics
To see what research has discovered about the evolution of inflation dynamics in recent years, let’s explore the following questions: (1) What is the available evidence on changes in inflation persistence in recent years? (2) What is the available evidence on changes in the slope of the Phillips curve? (3) What role do other variables play in the inflation process?

Inflation Persistence
To answer the first question, we need to measure how long the effects of a shock to inflation will last. Specifically, we need to know whether inflation tends to revert quickly to its initial level, or whether the effects of the shock persist--that is, lead to a changed level of inflation for an extended period. The most obvious way of measuring inflation persistence is to regress inflation on several of its own lags and then calculate the sum of the coefficients on lagged inflation. If the sum of the coefficients is close to 1.0, then shocks to inflation have long-lived effects on inflation. In other words, inflation behaves like a random walk, so that when inflation goes up, it stays up. If the sum of the coefficients drops well below 1.0, then a shock to inflation has only a temporary effect on inflation, and inflation soon reverts back to its trend level.

Figure 1. Rolling Coefficient Sums From Univariate Model of
Core PCE Price Inflation
[click on figure to enlarge]

The evidence from these so-called autoregressions with U.S. data suggests that inflation may have grown less persistent over time. In particular, when autoregressions are run using rolling samples with a fixed width of, say, twelve years, the sum of the lagged coefficients often falls noticeably below unity as the sample advances to include the most recent data. Figure 1 illustrates this tendency, with the core personal consumption expenditures (PCE) price index as the measure of inflation. As you can see, the coefficient sum has declined to about 0.6 or so since the late 1990s, although the dotted lines indicate that the confidence interval for this estimate is wide and still includes 1.0. This result appears robust to the addition of other explanatory variables to the regression, although making such a modification does tend to reduce the extent of the estimated decline.[2] Finally, it’s worth noting that this is not just a U.S. phenomenon, as some studies have found similar results for a number of other countries.[3]

(I might proudly note in passing that the staff of the Federal Reserve System is responsible for much of this research. To avoid boring you, I will refrain from providing references to this research, but the version on this speech on the Board’s web site will include this information and will provide a more nuanced discussion of these results.)

Recent work by Jim Stock and Mark Watson (2007) provides an alternative and, to me, a quite intuitive way of thinking about inflation persistence. They estimate a model that decomposes inflation into two components. The first component, which can be thought of as the underlying trend, follows a random walk, so that shocks to this component persist indefinitely and thus affect the trend inflation rate going forward. The second component is a serially uncorrelated shock, meaning that such shocks are temporary and lead to only transitory fluctuations around the trend. Stock and Watson then allow the volatility of these two kinds of shocks, trend and temporary, to vary over time.

Stock and Watson find that the importance of the trend shocks relative to that of the temporary shocks started to rise at the end of the 1960s in the United States, peaked in the mid-1970s, stayed elevated over the next ten years, and then declined to a historical low (upper panel of figure 2). When the relative importance of the trend shocks became high, as it did in the 1970s and early 1980s, inflation became highly persistent. Under such conditions, if inflation went up, the trend component typically rose in tandem so that inflation stayed up. In contrast, when the temporary shocks were relatively more important, as was true before the 1970s and after the mid-1980s, a change in inflation tended to reflect a change in the temporary component, not the trend. As a result, fluctuations in inflation tended to fade away, implying that inflation persistence was much lower. In the context of the Stock-Watson model, then, when inflation rose to double-digit levels during the "Great Inflation" period of the 1970s and early 1980s, persistence became very high because the trend rate of inflation moved around a lot--that is, trend inflation became unanchored.

Figure 2. Core PCE Inflation: Estimates from the Stock-Watson UC/SV Model. Panel A. Standard deviation of innovations. Panel B. Estimated trend component and actual inflation
[click on figure to enlarge]

Although the importance of the trend component has fallen markedly since the 1970s, the estimates produced by Stock and Watson suggest that it is still large enough to be economically meaningful. The bottom panel of figure 2 illustrates this point; as you can see, the estimated trend (the solid line) has drifted up a bit since the mid-1990s. Of course, such estimates are inevitably imprecise to some degree. For example, Cecchetti et al (2007) employ a slightly modified version of the Stock-Watson procedure in a recent paper and find that inflation persistence has fallen so low that their estimated trend for gross domestic product (GDP) price inflation has been almost perfectly flat at about 2.2 percent over the past few years. In contrast, Cogley and Sargent (2005), using a different modification of the procedure, obtain results more in line with the original Stock and Watson paper. Finally, follow-up work by staff members at the Federal Reserve Board suggests that the degree of persistence found with these techniques varies across different measures of inflation. Thus, the empirical evidence on this question can vary. Nonetheless, I think it fair to say that inflation has become less persistent over the past two decades but that the underlying trend may not yet be perfectly stable.

When Cecchetti and his co-authors apply this type of analysis to other countries, the results are remarkably similar. In particular, in Canada, Italy, and the United Kingdom, the importance of the trend shocks began to rise in the late 1960s and early 1970s and then declined only in the mid-1980s. France followed a similar pattern, although there the rise began a bit earlier, in 1963. In short, all these countries experienced a similar "Great Inflation," when trend inflation became unanchored. Germany and Japan had shorter periods of high persistence of inflation, with persistence beginning to decline around 1969 in Germany and around 1979 in Japan. Cecchetti and his co-authors conclude that the rise and subsequent decline of inflation persistence has thus been a worldwide phenomenon.

Slope of the Phillips Curve
In traditional Phillips-curve equations, inflation depends on past values of inflation, an unemployment gap (the difference between the unemployment rate and an estimate of its natural rate), and variables such as the relative price of energy and import prices. When researchers estimate these equations, they typically find that the coefficient on the unemployment gap has declined (in absolute value) since the 1980s, often by a marked amount.[4] In other words, the evidence suggests that the Phillips curve has flattened.

The finding that inflation is less responsive to the unemployment gap, if taken at face value, suggests that fluctuations in resource utilization will have smaller implications for inflation than used to be the case. From the point of view of policymakers, this development is a two-edged sword: On the plus side, it implies that an overheating economy will tend to generate a smaller increase in inflation. On the negative side, however, a flatter Phillips curve also implies that a given increase in inflation will be more costly to wring out of the system. We can quantify this latter consideration using the so-called sacrifice ratio--the number of years that unemployment has to be 1.0 percentage point greater than its natural rate to reduce the inflation rate 1.0 percentage point. Averaging estimates obtained from a comprehensive battery of equation specifications suggests that the sacrifice ratio may be 40 percent larger--that is, it may be 40 percent more costly to reduce inflation than it was two decades ago. Is this really bad news? I will return to this question later.

Role of Other Variables in the Inflation Process
Empirical evidence suggests that inflation has also become less responsive to other shocks. The two oil price shocks in the 1970s were associated with large jumps in core inflation, whereas recent surges in energy prices have not had a similar effect. This reduced sensitivity manifests itself in traditional Phillips-curve models as a substantial decline in the estimated coefficient on the energy price term in these equations. Because this term equals the change in relative energy prices multiplied by the share of energy in aggregate output, energy price effects on inflation appear to have fallen by more than can be accounted for by the greater energy efficiency of the economy.[5]

In contrast, unpublished empirical work by the staff at the Federal Reserve Board suggests that, once we take the rising share of imports into account, the influence of import prices on core inflation in the United States has not changed much in the context of reduced-form forecasting models.[6] At the same time, the influence of exchange rate movements on import prices--the so-called pass-through effect--may have fallen substantially, at least according to some studies.[7] If so, then the influence of exchange rate fluctuations on domestic inflation may now be less than it once was, when one controls for changes in the volume of our foreign trade.

What Interpretation Can We Give to Changes in Inflation Dynamics?
In interpreting these stylized facts about changes in inflation dynamics, we must first recognize that all of them are based on reduced-form relationships, and thus they are about correlations and not necessarily about true structural relationships. Because the explanatory variables in inflation regressions are themselves influenced by changes in economic conditions, changes in the underlying monetary policy regime are likely to be a source of changes in reduced-form inflation dynamics. This problem is especially acute for structural relationships involving expectations or other factors that are not directly observable and so cannot be included in reduced-form regressions. In such cases, we cannot use the reduced-form equations to disentangle the effects of such unobserved factors--which themselves may be driven by changes in monetary policy--from that of other influences.

The most important development in monetary economics that I have witnessed over my now-long career has been the recognition that expectations are central to our understanding of the behavior of the aggregate economy. Theory tells us that inflation expectations must be a key driving force behind inflation. This dependence has long been implicit in traditional Phillips-curve analysis, but expectations--now in explicit form--are also a central feature of the increasingly popular New Keynesian Phillips curves, in which current-period inflation is a function of expectations of next period’s inflation and resource utilization. Therefore, a natural first place to look for explanations of changing inflation dynamics is a possible change in the expectations-formation process.

The first stylized fact discussed earlier is that inflation persistence, which rose in the 1970s during the Great Inflation, has since declined to a much lower level. An intuitive way of thinking about this rise and fall in inflation persistence is that it resulted from an un-anchoring of trend inflation during the period of the Great Inflation, and a re-anchoring in recent years, as the work of Stock and Watson suggests. When we think about what drives trend inflation, inflation expectations--particularly long-run expectations--come to mind. A de-anchoring of inflation expectations would surely lead to trend inflation becoming unanchored, whereas an anchoring of inflation expectations at a particular level would necessarily lead to a stabilization of trend inflation and hence a decline in inflation persistence.

Figure 3. Trend and Expected Inflation
[click on figure to enlarge]

Do indicators of inflation expectations support this story? They certainly do. Consider the measure of expected inflation for the coming twelve months reported by the Livingston survey. As illustrated by the dashed line in figure 3 , this measure of expected short-run inflation--adjusted by a constant factor to convert it from a CPI basis to a PCE basis--soared during the 1970s and then fell back markedly during the 1980s and early 1990s. Over the past few years, the short-run series has fluctuated around 2 percent, with its month-to-month movements correlated with swings in energy prices, as one might expect. An even more striking story is told by the survey-based measure of expected long-run inflation used in the FRB/US model, the dotted blue line in figure 3.[8] When we look at this long-run series (which should be more closely related to expectations about policy objectives than the short-run measure), we see that the public’s expectations stood at a high level of 7-3/4 percent at the start of the 1980s, when this information first began to be collected. From that point on, however, expectations fell steadily over the 1980s and most of the 1990s as the process of re-anchoring continued. By 1998, this process was apparently completed, and since that time the public’s expectations--at least according to this particular measure--have been steady as a rock. Estimates of inflation compensation derived from indexed Treasury yields have also been remarkably stable in recent years. Figure 3 also reproduces the estimate of trend inflation produced by the Stock-Watson procedure (the solid line). As you can see, this series more or less tracks the survey-based measures of long-run expectations, suggesting that the anchoring of long-run inflation expectations in recent years may explain the finding of a marked decline in inflation persistence.

Anchoring of inflation expectations is not a deus ex machina. It must come from somewhere, and since Milton Friedman’s adage that “[i]nflation is always and everywhere a monetary phenomenon” is still true, monetary policy must be the source of the change in the evolution of long-run inflation expectations. During the 1960s and 1970s, the Federal Reserve and a number of other central banks maintained a policy stance that, inadvertently or not, was too easy on average and that allowed both actual and expected inflation to drift up markedly over many years. Since the late 1970s, however, the Federal Reserve and many other central banks have increased their commitment to price stability, in both words and actions. The Fed pursued preemptive strikes against rises in inflation in 1994-95, 1999-00, and in 2004-06, as well as preemptive strikes against potentially deflationary forces in the fall of 1998 and in 2001-04. The result has been not only low and stable inflation but also, as we have seen, a strong anchoring of long-run inflation expectations.

The pursuit of more-aggressive monetary policy to control inflation and the achievement of anchored inflation expectations can also help explain the other stylized facts about inflation dynamics. With expectations of inflation anchored, any given shock to inflation--whether it is from aggregate demand, energy prices, or the foreign exchange rate--will have a smaller effect on expected inflation and hence on trend inflation. These shocks will then have a much less persistent effect on actual inflation. The recent experience with surging oil prices seems consistent with this story. The price of West Texas intermediate crude oil rose from about $30 per barrel in late 2003 to a peak of almost $75 per barrel in July of last year. During this episode, inflation appears to have been boosted only temporarily and by a strikingly small amount, in contrast to the 1970s when oil price shocks led to large, persistent increases in inflation.

Interestingly, monetary policy could have worked to flatten the estimated Phillips curve even without these favorable expectational effects, simply by moving more aggressively to stabilize both inflation and output. A theorem from the literature on optimal control states that implementing a policy of adjusting some instrument to stabilize a particular variable has the effect, in the limit, of driving the correlation between the instrument and the targeted variable to zero. Thus, a monetary policy that more successfully stabilizes inflation and resource utilization could well lead to a smaller estimated coefficient on unemployment gaps in a traditional Phillips-curve equation.

Research carried out with quite different models of the macroeconomy by my former colleagues at Columbia, Jean Boivin and Marc Giannoni (2006), and by John Roberts (2006) at the Federal Reserve Board, supports these conclusions. In particular, their analyses suggest that changes in the way the Federal Reserve conducts monetary policy--including changes in both the parameters of monetary policy reaction functions and the volatility of shocks to those functions--may account for most of the reduction in the coefficients on resource utilization in traditional Phillips curves.[9]

What is particularly attractive about highlighting a better anchoring of inflation expectations as probably the primary factor driving the changes in inflation dynamics is that this one explanation covers so many of the stylized facts--an application of Occam’s razor. Indeed, I have always become more confident in a theory if it can explain a number of very different facts.[10] This is why I am so attracted to the view that inflation expectations are a key driving factor in the inflation process.

Of course, many other factors also influence inflation, and some of these provide other possible explanations for the recent changes in inflation dynamics. For example, smaller coefficients on unemployment gaps and energy prices in traditional Phillips curve equations may reflect the influence of lower and less-variable inflation on the frequency with which firms change their prices. In the context of a low-inflation environment, firms may have concluded that they can leave their prices fixed for longer periods at little cost, causing inflation to be less responsive to shocks, particularly if they are transitory. In this way, monetary policy may have affected the slope of the Phillips curve without having affected the manner in which expectations are formed. Also, from the mid-1980s through the first years of this decade, energy price movements were smaller and more transitory. We see this in futures markets, in which oil prices in far-dated futures contracts moved much less than spot prices over this period, suggesting that people expected a quick reversal in any rise in oil prices. Such transitory shocks to energy price would presumably have a smaller effect on inflation than the more-persistent oil price shocks of the 1970s and early 1980s. Increased globalization and other sources of increased competition may also have lowered the sensitivity of domestic inflation to aggregate demand, thereby flattening the Phillips curve. However, the evidence on this last point is limited and inconclusive.[11]

Policy Implications
These stylized facts--that inflation has become less persistent and now responds less to aggregate demand and supply shocks--can lead to inappropriate policy advice. If we take these facts to be structural and attributable to factors other than monetary policy, we might interpret them as suggesting that the Federal Reserve could respond less to shocks and yet be confident that inflation would remain at a low level. In addition, the smaller coefficient on unemployment gaps in traditional Phillips-curve models, which seems to imply that the sacrifice ratio has gone up, might lead us to think that reducing inflation is very costly because it requires long periods of high unemployment. As a result, policymakers might decide that such an attempt would not be worthwhile and so would be less likely to try to reduce inflation if it were undesirably high. If used in model simulation exercises, the flatter Phillips curve might also suggest that getting inflation down to a particular desired level could take an inordinately long time.

However, if we instead attribute the observed changes in inflation dynamics to better monetary policy and a resultant better anchoring of inflation expectations, then such policy conclusions are unwarranted. Under this alternative interpretation, the reason that inflation has become less persistent is that monetary policy, in carrying out its dual mandate of promoting both price stability and maximum sustainable employment, has been vigilant in maintaining a low rate of inflation on average. But if the monetary authorities were to become complacent and to think that they could get away with not reacting to shocks that, in their mistaken view, no longer have the potential to cause inflation to rise persistently, then inflation expectations would surely become unhinged again. In short, complacency that might arise from the current low inflation persistence might result in deja vu all over again and return us to an era like the Great Inflation. These are exactly the concerns expressed in Tom Sargent’s (2000) book on the rise and fall of U.S. inflation, in which he worries that a misunderstanding of the inflation process might again lead to a high-inflation equilibrium.

If inflation has indeed become less persistent because better monetary policy has anchored inflation expectations more solidly, the monetary authorities may find that they have less need to induce large swings in economic activity to control inflation. This is a key benefit of establishing a strong nominal anchor. Because the public has become confident that the Fed will do the right thing, expectations now behave in a manner that makes the economy more stable to begin with. If this hypothesis is correct, cyclical movements in interest rates need not be as great as was necessary when expectations were unanchored. However, these favorable circumstances will persist only so long as the monetary authorities continue to ratify the public’s expectations. Consistent with its dual mandate, the Fed must therefore continue to respond aggressively to shocks that have potentially persistent adverse effects on both inflation and real activity. Because long-run inflation expectations are a key driver of trend inflation, the monetary authorities need to monitor long-run inflation expectations closely. If they find that they are losing credibility with the markets, so that inflation expectations begin to drift and rise above (or fall below) a desired level, they must take actions to restore their credibility.

At the Federal Reserve, we understand the importance to the health of the economy of anchoring inflation expectations. This is why Federal Reserve officials continually reiterate our commitment to maximum sustainable employment and price stability, why we have been willing to make preemptive strikes against both inflationary and deflationary pressures, and why we remain vigilant about developments in the economy that could lead to persistent departures of inflation from levels that are consistent with price stability.

My view--that recent changes in inflation dynamics result primarily from better-anchored inflation expectations and not from structural change or simply the achievement of a persistent low rate of inflation--implies some very good news: Potentially inflationary shocks, like a sharp rise in energy prices, are less likely to spill over into expected and actual core inflation. Therefore, the Fed does not have to respond as aggressively as would be necessary if inflation expectations were unanchored, as they were during the Great Inflation era. Indeed, this can help explain why the recent sharp rises in energy prices have had a much more benign effect on the real economy than they did in the 1970s--a point that then-Governor Ben Bernanke made three years ago.

Although solidly anchored inflation expectations are indeed highly desirable, they could pose a bit of a problem for monetary policy if they were at a level somewhat above or below the rate preferred by policymakers. Under such circumstances, the central bank would likely be interested in shifting the public’s expectations in a more favorable direction. Whether such adjustment would be easy or difficult is, unfortunately, quite uncertain because we do not understand the expectations-formation process very well. In some early models that used the rational expectations assumption, changing inflation expectations was relatively easy and thus implied sacrifice ratios that were extremely low, suggesting that the monetary authorities could shift inflation at little cost. However, the historical record suggests that permanently lowering inflation expectations may require keeping monetary policy tight for a substantial period, resulting in considerable output and employment losses for a time.

On the other hand, as Christy and David Romer (2002) have pointed out, the Federal Reserve in the 1970s overestimated the cost of reducing inflation because estimates of sacrifice ratios by Arthur Okun and other economists at that time proved to be too high.[12] As a historical note, this provides one explanation for the Federal Reserve’s tolerance of such high inflation at the time. The disinflation after October 1979, carried out by the Federal Reserve under the leadership of Paul Volcker using words, procedures, and actions that were a sharp break from the past, produced a much lower cost of disinflation than policymakers had anticipated during the 1970s.

I think these considerations leave us with fairly wide bounds on what the costs might be of permanently shifting long-run inflation expectations that are already anchored. On the one hand, the historical record gives us little reason to think the costs would be as minimal as the simplest models with rational expectations might suggest. On the other hand, overly pessimistic estimates have proved to be wrong in the past.

Implications for Inflation Forecasts
I have argued here that the most attractive explanation for recent changes in inflation dynamics is that expectations have become better anchored. Inflation is now less persistent and more likely to gravitate to a trend level that is determined by where long-run inflation expectations have settled. What implication does this have for forecasts of future inflation? In this framework, the first priority is to determine where inflation expectations may be anchored. This task is somewhat tricky. According to the latest reading from the Survey of Professional Forecasters (SPF), long-term inflation expectations are currently around 2 percent, as measured by the PCE price index that the FOMC has emphasized. The private forecasters appear to have held this view for many years, given that long-run expectations for CPI inflation in the SPF have been running at around 2-1/2 percent since the late 1990s, and that the average historical wedge between the CPI measure and the PCE measure of inflation has been about 50 basis points.[13]

Of course, the views of professional forecasters may not be representative of what the average person or firm is thinking.If we turn to the financial markets, we find that inflation expectations extracted from a comparison of the yields between Treasury inflation-protected securities (TIPS) and standard Treasury securities seem consistent with an anchor at or perhaps a little above 2 percent in terms of PCE inflation.[14] I say “seem consistent,” because we cannot be sure about at least one of the items used in the extraction, the premium paid to investors to compensate them for inflation risk. In the case of households, long-term inflation expectations from the Reuters/Michigan survey have been running much higher for a number of years, at around 3 percent. However, this figure seems overstated in light of the persistent bias found in the short-term inflation expectations reported by this survey. Correcting for the bias, these survey results are probably more in line with PCE inflation closer to 2 percent.[15] As for firms, we unfortunately have no good data on their inflation expectations. More information on this score would be particularly welcome, but expectations in general is an area worthy of further study.

Taken together, the data suggest to me that long-run inflation expectations are currently around 2 percent. That said, I think it should be clear that the evidence points to a range of estimates; moreover, this range is itself uncertain because of the assumptions needed to tease point estimates from the available data. So, although I think that 2 percent is a reasonable estimate of current long-run expectations, I don’t want to overstate the precision of this figure. We still face some uncertainty in this regard, and policymakers must be cautious about placing too much confidence in any one estimate.

If long-run expectations are in fact about 2 percent, where is actual inflation likely to be headed in the next year or two? While recognizing how embarrassingly wrong such prognostications often turn out to be, I think that we can be reasonably optimistic that core PCE inflation will gradually drift down from its latest twelve-month reading of 2-1/4 percent. This process may take a while in light of the recent rebound in prices for gasoline and other petroleum products. These price increases have boosted the cost of producing many non-energy goods and services, and as firms gradually pass on these higher costs to their customers, monthly readings on the change in core prices are likely to be higher than they otherwise would be. Once this process is completed, however, we might expect consumer price inflation to move into better alignment with long-run expectations and thus settle in around 2 percent. Of course, our understanding of the empirical links between our measures of expected inflation and actual inflation is sufficiently poor that things could well go awry with this forecast. Moreover, many things could happen in the coming months to alter the outlook, as the recent fluctuations in energy markets and swings in GDP growth illustrate.

Looking to the medium term, I am less optimistic about the prospects for core PCE inflation to move much below 2 percent in the absence of a determined effort by monetary policy. For the most part, this assessment--which I should stress is subject to considerable uncertainty--flows from my view that long-term expectations appear to be well anchored at a level not very far below the current rate of inflation. If so, a substantial further decline in inflation would require a shift in expectations, and such a shift could be difficult and time consuming to bring about, as I noted earlier.

As I mentioned at the start, central bankers are acutely interested in the inflation process. This is why I have thought about the topic a lot and chosen to talk about it today. I hope you have found my musings on this subject useful.


Bernanke, Ben (2004). "The Great Moderation," speech delivered at the meeting of the Eastern Economic Association, Washington, D.C., February 20.

Boivin, Jean, and Marc Giannoni (forthcoming) "Has Monetary Policy Become More Effective?" Review of Economics and Statistics.

Borio, Claudio, and Andrew Filardo (2006).  "Globalisation and Inflation: New Cross-Country Evidence on the Global Determinants of Domestic Inflation," unpublished paper, Bank for International Settlements, Basel, Switzerland (March).

Cecchetti, Stephen G., Peter Hooper, Bruce C. Kasman, Kermit L. Schoenholtz, and Mark W. Watson (2007). "Understanding the Evolving Inflation Process," presentation at the U.S. Monetary Policy Forum, Washington, D.C., March 9.

Clarida, Richard, Jordi Gali, and Mark Gertler (2000). "Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory," Quarterly Journal of Economics, vol. 115 (February), pp. 147-80.

Cogley, Timothy, and Thomas Sargent (2005). "Drifts and Volatilities: Monetary Policies and Outcomes in the Post WWII U.S.," Review of Economic Dynamics, vol. 8 (April), pp. 262-302.

Friedman, Milton (1957). Theory of the Consumption Function. Princeton, N.J.: Princeton University Press.

Hanson, Bruce (1999). "The Grid Bootstrap and the Autoregressive Model." Review of Economics and Statistics, vol. 81 (November), 594-607.

Hellerstein, Rebecca, Deirdre Daly, and Christina Marsh (2006). "Have U.S. Import Prices Become Less Responsive to Changes in the Dollar?" Federal Reserve Bank of New York, Current Issues in Economics and Finance, vol. 12 (September), pp. 1-7.

Hooker, Mark A. (2002). "Are Oil Shocks Inflationary? Asymmetric and Nonlinear Specifications versus Changes in Regime," Journal of Money, Credit, and Banking, vol. 34 (May), pp. 540-61.

Ihrig, Jane, Steven Kamin, Deborah Lindner, and Jaime Marquez (forthcoming). “Some Simple Tests of the Globalization and Inflation Hypothesis,” International Finance Discussion Papers. Washington: Board of Governors of the Federal Reserve System.

Levin, Andrew T., and Jeremy M. Piger (2004). "Is Inflation Persistence Intrinsic in Industrial Economies?" (1.0 MB PDF) ECB Working Paper Series 334. Frankfurt, Germany: European Central Bank, April.

Marazzi, Mario, Nathan Sheets, Robert J. Vigfusson, Jon Faust, Joseph E. Gagnon, Jaime Marquez, Robert F. Martin, Trevor A. Reeve, and John H. Rogers (2005). "Exchange Rate Pass-Through to U.S. Import Prices: Some New Evidence," International Finance Discussion Papers 833. Washington: Board of Governors of the Federal Reserve System, April.

Nason, James M. (2006). "Instability in U.S. Inflation: 1967-2005," Federal Reserve Bank of Atlanta, Economic Review, vol. 91 (Second Quarter), pp. 39-59.

Okun, Arthur M. (1978). "Efficient Disinflationary Policies," American Economic Review, vol. 68 (May), pp. 348-52.

O’Reilly, Gerard, and Karl Whelan (2005). "Has Euro-Area Inflation Persistence Changed over Time?" Review of Economics and Statistics, vol. 87 (November), pp. 709-20.

Roberts, John M. (2006). "Monetary Policy and Inflation Dynamics," International Journal of Central Banking, vol. 2 (September), pp. 193-230.

Romer, Christina D., and David H. Romer (2002). "The Evolution of Economic Understanding and Postwar Stabilization Policy," in Rethinking Stabilization Policy, proceedings of a symposium sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, August 29-31, pp. 11-78.

Rudebusch, Glenn D. (2005). "Assessing the Lucas Critique in Monetary Policy Models," Journal of Money, Credit, and Banking, vol. 37(April), pp. 247-72.

Sargent, Thomas J. (2000). The Conquest of American Inflation. Princeton, N.J.: Princeton University Press.

Sekine, Toshitaka (2006). "Time-Varying Exchange Rate Pass-Through: Experiences of Some Industrial Countries," BIS Working Paper 202. Basel, Switzerland: Bank for International Settlements, March.

Stock, James, and Mark Watson (2007). "Why Has U.S. Inflation Become Harder to Forecast?" Journal of Money, Credit, and Banking, vol. 39 (February), pp. 3-34.

Williams, John C. (2006). "The Phillips Curve in an Era of Well-Anchored Inflation Expectations," unpublished working paper, Federal Reserve Bank of San Francisco, September.


1. I want to thank Michael Kiley, Jean-Philippe Laforte, Deborah Lindner, David Reifschneider, John Roberts, and Jeremy Rudd for their extremely helpful comments and assistance on this speech.

2. Two recent studies that report a marked decline in inflation persistence in the United States are Nason (2006) and Williams (2006). Their findings should be treated carefully, however, as analysis by the staff of the Federal Reserve Board suggests that modest changes in methodology, such as lengthening the sample period, correcting for small-sample bias, or changing the particular price index used in the analysis, can alter both the magnitude and the statistical significance of the estimated decline in persistence.

3. Levin and Piger (2004) find significant declines in inflation persistence since the 1980s for the major European economies as well as for Japan, Canada, Australia and New Zealand. However, O’Reilly and Whelan (2005) find little or no evidence for a recent decline in persistence for the euro area as a whole.

4. Studies that present evidence of a marked decline in the sensitivity of U.S. inflation to unemployment and other measures of resource utilization include Roberts (2006) and Williams (2006). Unpublished work by staff at the Federal Reserve Board indicates that this result generally holds across a variety of regression specifications, estimation methods, and data definitions. Other studies find similar declines in many foreign industrial economies; see, among others, Borio and Filardo (2006) and Ihrig and others (forthcoming). However, the empirical evidence on this question is such that the exact magnitude, timing, and statistical significance of these changes remain a subject of debate.

5. Hooker (2002) finds that oil price movements during the 1980s and 1990s had little or no effect on core inflation, in contrast to a substantial influence in previous decades. Recent empirical work carried out at the Federal Reserve Board confirms Hooker’s findings using data through 2006; however, some of this work also hints that the surge in energy prices since 2004 may have had a larger influence on core inflation.

6. Interestingly, Ihrig and others (forthcoming) find that the sensitivity of inflation to movements in import prices has fallen in several foreign industrial economies.

7. For evidence on U.S. exchange rate pass-through effects, see Marazzi and others (2005). Hellerstein, Daly, and Marsh (2006) also find that pass-through has declined in the United States, although by a considerably smaller amount. Empirical results for other developed countries are reported in Sekine (2006).

8. The FRB/US series splices data from two surveys of expected long-run inflation--the Hoey survey of financial market participants from 1981 through 1989, and the Survey of Professional Forecasters from 1990 on. After splicing, a 50-basis-point constant factor is subtracted from the series to put it on a PCE price index basis, on the assumption that survey respondents reported their expectation for CPI inflation, not PCE inflation. (The average wedge between CPI inflation and PCE inflation was about 50 basis points from 1980 through 2006.) I have made a similar adjustment to the Livingston survey data plotted in figure 3 to put it on a PCE price basis.

9. Clarida, Gali, and Gertler (2000) make a similar argument, but for a dissenting view, see Rudebusch (2005).

10. This is why Milton Friedman’s Theory of the Consumption Function is one of my favorite pieces of empirical research. What is extraordinary about this book is that it has only a handful of regressions but nonetheless shows that the basic idea of permanent income has to be right because the theory explains numerous facts derived from numerous studies. This book is not widely read by graduate students nowadays, but it should be.

11. See Borio and Filardo (2006) and Ihrig and others (forthcoming) for opposing views on this issue. Borio and Filardo provide evidence that global output gaps may be just as important as conventional domestic output gaps in the determination of inflation; moreover, they argue that these effects have been rising over time. However, Ihrig and her co-authors find that Borio and Filardo’s results are sensitive to small changes in specification; they also find little support for an independent role of global output gaps and no evidence that globalization can account for falling coefficients on domestic gaps.

12. Okun (1978) estimated that a 10 percent reduction in real GDP for one year would reduce the inflation rate only 1.0 percentage point. By the mid-1980s, however, Federal Reserve Board staff estimates of the sacrifice ratio were roughly half as large.

13. Recently, the Survey of Professional Forecasters (SPF) has begun to report long-run expectations for the PCE price index as well as the CPI. Long-run expectations for PCE price inflation were 2 percent in the latest release, the same rate as indicated by the FRB/US estimate since the late 1990s. (The FRB/US estimate since 1991 derives from the SPF reading on expected long-run CPI inflation less an average historical wedge between the CPI measure and the PCE measure of inflation of about 50 basis points.) Interestingly, long-run expectations for CPI inflation in the SPF ticked down to 2-1/3 percent in the latest survey, suggesting that private forecasters may have lowered their estimate of the average CPI-PCE wedge to about 30 basis points. Such a revision would be in line with current Board staff estimates, which indicate that the wedge has declined in recent years and now stands at about 30 basis points.

14. Inflation compensation--the difference between nominal and indexed Treasury yields--has recently been about 1/2 percent. However, two adjustments are necessary to translate inflation compensation into an estimate of expected long-term inflation on a PCE price basis. First, because the difference between nominal and indexed yields equals the sum of inflation expectations and an inflation term premium, we must subtract an estimate of the premium. Given that term premiums on nominal Treasury yields are extremely low at present, the inflation term premium is probably less than 25 basis points currently, and could even be zero. Second, because the inflation series used to index the Treasury securities is the CPI, not the PCE price index, we must subtract an estimate of the wedge between the two measures of inflation. If we use the same 30 basis points wedge expected by private forecasters in the latest reading from the SPF, indexed and nominal yields appear to be consistent with expected long-term PCE inflation of about 2 percent or a little higher.

15. According to the Reuters/Michigan survey, long-term inflation expectations of households are currently around 3 percent, as they have been for the past few years. This survey does not specify a price index. However, expectations from the companion one-year-ahead expectations have come in about 75 basis points higher than actual PCE inflation since 1990, suggesting that there may be a systematic bias in the responses to the Reuters/Michigan survey relative to this measure of inflation. If this bias also applies to longer-run inflation expectations, then household expectations may be in line with PCE inflation running in the vicinity of 2-1/4 percent in the long run.

What Conundrum?

Mohamed A. El-Erian of the Harvard Business School and Nobel laureate in economics Michael Spence say there's nothing puzzling or hard to understand about global imbalances, declining risk spreads, flattened yield curves, and declining market volatility. However, their analysis of these changes leads them to conclude that global imbalances raise "considerable challenges, as does the ability to maintain an orderly global reconciliation process over time":

Capital Currents, by Mohamed A. El-Erian and Michael Spence, Commentary, WSJ: For the past few years, the U.S. has generated insufficient domestic savings to cover its investment needs. The difference has been covered by large capital inflows from abroad, the counterpart of which is the much-discussed current account deficit, which has been running at unprecedented rates of 6%-7% of GDP. ...

This has raised concerns about its sustainability, including whether it will end in a sudden and disorganized manner that sharply reduces growth in the global economy and causes problems in global capital markets. And underlying the concern is a kind of puzzlement about the configuration of global savings -- one that runs counter to virtually every text book description: The world's richest country appears to be saving at a low rate and has to borrow from poorer, developing countries to maintain its consumption and investment.

Let's analyze this puzzle, beginning with the U.S. The financial assets of U.S. households have risen rapidly in the past 10 years, at rates well above inflation. The most dramatic increase occurred in household real estate, principally housing. At least some of these increases in asset values were not anticipated, relative to ... long-term ... savings and consumption plans... [I]t ... seems reasonable that U.S. households would consume a portion of their windfall ... over time.

The recent shakiness in the subprime mortgage market has created uncertainty as to whether this dynamic will be sustained. ... More generally, the potential pressure on house prices could also reduce household's propensity to consume out of their accumulated equity windfall. And ... there is concern in ... capital markets that global growth could be negatively impacted.

These concerns are worth monitoring carefully -- and they highlight a more general issue...: While individual consumption and savings decisions may have been largely rational, that does not mean that the decisions of individual households "add up" properly in the aggregate... In fact they easily might not have. ...

[W]e can consume and invest more than output by importing more than we export -- and we did. Hence the trade deficit. However, this ability is dependent on the ability and willingness of the rest of the global economy to accommodate the US desire for higher consumption by investing in the US. That accommodation has been forthcoming from emerging economies generally, including OPEC and China, as well as from Japan.

Their initial reaction to their improved external trade balances has been primarily to recycle the funds to the "risk free assets," U.S. Treasuries and Agencies. By financing U.S. consumption, many surplus countries are also meeting their domestic objectives, to promote exports, increase employment and build up significant reserve cushions to deal with the possibility of sudden disruptions in global capital markets.

This constellation of conditions was largely unanticipated by both markets and policy makers, and as a result it has been reflected in a host of unusual economic and market outcomes -- referred to as conundrums, aberrations, puzzles, etc. The most visible is ... global "imbalances"; also of note is the excessive compression in risk spreads, the unusual collapse in market volatility, the inverted shape of the U.S. yield curve...

Now to the future. Over time, emerging markets will inevitably divert more of their assets to more sophisticated investments abroad. ... One effect will surely be to put upward pressure ... on the cost of capital in the U.S., as the incremental demand for treasuries declines.

While the shift is inevitable, it would be unlikely that the emerging economies as a group would deliberately ... undermine global economic markets. There will also be domestic pressure on policy makers in emerging countries to gradually shift their emphasis away from the producer and towards the consumer. That will mean lowering the savings rate relative to investment, increasing consumption and letting it assume a more important role (relative to exports...) in driving growth.

Under this state of the world, domestic consumption in the rest of the world picks up over time, facilitating the needed adjustment in the U.S. The result is a gradual journey to a more normal relationship between assets and income returns, with savings moving to a more normal long-run pattern.

But this process is not automatic and faces significant disruption risks; and it is particularly sensitive to "policy mistakes." Among these policy mistakes, protectionism measures in the U.S. would derail the global adjustment So, too, would the inability of emerging economies to navigate their complex policy challenges.

Geopolitical shocks would also be a problem... Finally, significant "market accidents" ... associated with excessive leverage and ... sudden and large portfolio changes and credit rationing, would add to the policy complexity.

So where does all this leave us? The current configuration of global imbalances, while highly unusual is not a real puzzle. It is the result of a series of individual decisions in both advanced and emerging economies that were largely rational when considered at the micro level. ...

The aggregation of these decisions at the national and international levels raises considerable challenges, as does the ability to maintain an orderly global reconciliation process over time. The fundamental question, therefore, is whether these global considerations will be sufficient to minimize the risk of "policy mistakes" in a world that is subject to geo-political risk and bouts of excessive leverage.

One thing that bothers me about this story is that it begins with a run-up in asset prices as the source of global imbalances:

Let's analyze this puzzle, beginning with the U.S. The financial assets of U.S. households have risen rapidly in the past 10 years, at rates well above inflation. The most dramatic increase occurred in household real estate, principally housing.

However, prices are endogenous variables, and therefore this explanation leaves the primary driving force behind the run-up in asset prices unexplained. Jim Hamilton looks at the housing market in "Bubble, bubble, toil, and trouble." His analysis is concerned with whether or not housing prices have departed from underlying fundamentals, and he doesn't believe that they have. He concludes:

Low interest rates and rapid population and employment growth relative to the supply of available housing were the main factors driving house prices up...

To that, Jim adds:

The one thing to which I think I was not paying enough attention two years ago was the role of lax credit standards and even fraud ([1], [2]) in addition to low interest rates as factors fueling the boom. I have been coming around to the view that there may have been some significant market failures behind that. My first worry here is about Fannie Mae and Freddie Mac, and the second concerns whether some of our institutions have the right incentives for fund managers to properly value lower-tail risks. This ready availability of credit, over and above the low interest rates themselves, I now believe was an important factor contributing to the real estate boom.

I would also mention regulatory restrictions as an important factor, e.g. see Edward Glaeser's on zoning regulations or Krugman on Flatlands and Zoned Zones. But many people blame (or thank) the Fed for driving interest rates so low.

Which opens the door to "Did the Fed Do It?" from David Altig. David isn't so sure that the Fed is to blame for the escalation in housing prices:

Did The Fed Do It?, macroblog: The ISI Group's Andy Laperriere, writing on the opinion page of yesterday's Wall Street Journal, says the answer is yes (at least in part):

Federal Reserve officials and most economists believe the problems in the subprime mortgage market will remain relatively contained, but there is compelling evidence that the failure of subprime loans may be the start of a painful unwinding of a housing bubble that was fueled by easy money and loose lending practices...

The ... fallout from the second major asset price bubble in the last decade should prompt some broader questions. For example, what role did the Fed's loose monetary policy from 2002-2004 play in fueling the housing bubble? Should the Federal Reserve reexamine its policy of ignoring asset bubbles?

I know that the easy money claim has become something of a meme, but I often find myself pondering this picture:


What's the story here?  That the long string of federal funds rate cuts beginning in January 2001 caused the decline in long-term interest rates -- including mortgage rates -- that commenced a full half-year (at least) before the first move by the FOMC?  That low levels of short-term interest rates have kept long-term rates well below their pre-recession peaks?  Then what to make of the fact that rates at the longer end of the yield curve have barely budged in the face of a 425 basis point rise in the funds rate target?  Maybe it's "long and variable lags"? Should we then be expecting that big jump in long-term rates any day now?  I guess it's still a conundrum. But maybe, then, we should be a little circumspect about the finger pointing?

OK, here's part of the Laperriere article I can get behind:

It's not the size of foreclosure losses as a share of the economy that matters, it is the effect those losses have on the availability of credit.

Like I said.

In a recent speech, Fed Chair Ben Bernanke says the Fed still has the ability to affect long-term rates:

The empirical literature supports the view that U.S. monetary policy retains its ability to influence longer-term rates and other asset prices. Indeed, research on U.S. bond yields across the whole spectrum of maturities finds that all yields respond significantly to unanticipated changes in the Fed’s short-term interest-rate target and that the size and pattern of these responses has not changed much over time (Kuttner, 2001; Andersen and others, 2005; and Faust and others, 2006). ...

[G]lobalization of financial markets has not materially reduced the ability of the Federal Reserve to influence financial conditions in the United States. But, ... globalization has added a dimension of complexity to the analysis of financial conditions and their determinants, which monetary policy makers must take into account.

I'm also intrigued by David's suggestion, and hopefully more evidence can settle whether previous research has this wrong. But for now, my policy recommendations will still account for the possibility that the Fed can affect long-term rates.

May 17, 2007

Milton Friedman in China

Here's another response to Paul Krugman's "Who Was Milton Friedman?" [previous response from Anna Schwartz and Edward Nelson along with and Paul Krugman's reply]:

Milton Friedman in China, by Bertrand Horwitz, Reply by Paul Krugman, In response to Who Was Milton Friedman?: To the Editors:

Paul Krugman's "Who Was Milton Friedman?" [NYR, February 15] is the best brief popular summary I have thus far read of the contributions of Professor Friedman, "the economist's economist" and the greatest exponent of free markets since Adam Smith as Krugman rightly contends. ...

Friedman's contributions ... were clearly described..., but a few other important results of his influence were missed. Much has been written about his and his students' ("the Chicago Boys") impact on the significant improvement of the economy during the Chilean dictatorship, but neither Krugman nor others even in memoriam have noted Friedman's effect on Chinese economic policy at crucial periods in its development. When he first visited China in 1980 the only policy guidelines the authorities set down following the breakdown of Mao's crumbling disorder were so-called pragmatic rules: "cross the river and feel the rocks" and then "seek the truth from facts." The "river" was not named and the place to "look" was not specified. In the meetings he had with the Chinese leaders, Friedman strongly emphasized the importance of unfettered markets, pointing to China's neighbor, Hong Kong, as a model to be followed.

This indeed, up to a point, is the road that has been taken. Again in 1988, the Chinese authorities, deeply worried by double-digit inflation which they knew undercut the Nationalists before 1949, sought his advice. Rumors had been spreading then that in Shanghai there was a run on the banks and even ordinary people were approaching foreigners with shouts of "wai hui, wai hui" (foreign exchange). The greatest spokesman for monetarism and his students had analyzed the causes of inflation in more than a dozen countries and had persuasively shown that the quantity theory of money works, that inflation indeed is a phenomenon of "too much money chasing too few goods," and that the application of price controls and rationing was a "cure" which would only worsen the situation. Friedman's advice was taken and since then China's inflation has been within a small, acceptable range.

Further missing from Krugman's summary is Friedman's theoretical contribution to the adoption of flexible foreign exchange rates. Until his analysis, flexibility was rejected on the grounds that such rates would be unacceptably unstable because of speculation. And it was Friedman whose faith in free markets led to his promulgating the negative income tax. When families' incomes fall below a certain level, they should be sent checks, using the money as they see fit. Missing also is Friedman's contribution to statistics—a nonparametric test is named after him. Whether his contributions will stand up to the test of time remains to be seen, but as he was so fond of saying: "the proof of the pudding is in the eating." Or as Einstein similarly said about theoretical physics: "the test of truth is experience." -- Bertrand Horwitz Asheville, North Carolina

Paul Krugman replies: I wasn't aware of the China story, and am glad to have it out there.

I didn't bring up exchange rate policy because I don't think Friedman can be said to have made a deep intellectual contribution on the subject. Nonetheless, his advocacy of flexible rates does illustrate two of his great virtues.

First, on this as on other issues he showed himself much less doctrinaire and much more realistic than many of his acolytes: many conservative economists are drawn to visions of a restored gold standard or a world currency, dismissing the problems such a system would create; Friedman knew better.

Second, his famous paper on flexible rates is a masterpiece of writing, with a brilliant analogy: achieving international adjustment by changing the exchange rate, rather than by depending on thousands of firms to change their prices, is like shifting to daylight savings time, rather than depending on thousands of firms to change their working hours.

More on the reference to "cross the river and feel the rocks" is in "Mo zhe shi tou guo he," or, "Cross the River by Groping the Stone Under Foot." A comparison of the "Beijing Consensus" with the "Washington Consensus [that] emerged from the neoliberal revolution that swept the globe with the arrival of the Thatcherite and Reaganite schools of thought and power" is part of the discussion.

Geithner: Credit Markets Innovations and Their Implications

New York Fed president Tim Geithner, who hasn't been shy about warning about financial risks from financial market innovation, doesn't seem too concerned that problems in the subprime mortgage market will spread and cause wider disruptions. Here's part of a longer speech:

Credit Markets Innovations and Their Implications, by Timothy Geithner, NY Fed President: ...The latest wave of credit market innovations has elicited some concerns about their implications for the stability of the financial system, concerns similar to those associated with earlier periods of rapid change in financial markets. Will the most recent credit market innovations amplify credit cycles, contributing to "excessive" lending in times of relative stability, and then magnify the contraction in credit that follows? Will they introduce greater volatility in financial markets? Will they create greater risk of systemic financial crisis?

These concerns have been heightened in some quarters by the problems currently being experienced in the subprime mortgage sector. It will take some time before the full implications are understood and the full impact can be assessed. As of now, though, there are few signs that the disruptions in this one sector of the credit markets will have a lasting impact on credit markets as a whole.

Indeed, economic theory and recent practical experience offer some reassurance against both these specific concerns and more general worries about the implications of credit market innovations for the performance of the financial system. ...

There are ... compelling arguments in favor of a generally positive assessment of the consequences of innovation. Does experience provide support for these arguments, or are these changes too new for us to know? ...

We are now well into the third decade of experience with the consequences of these earlier innovations, and this history offers some useful lessons for evaluating the probable impact of the latest changes in credit markets.

The ease with which the U.S. financial system absorbed the substantial scale of corporate defaults that peaked in recent years in 2002 provides some support for the argument that broader and deeper capital markets make the system more resilient. In general, there does not seem to be strong empirical support for the proposition that derivatives increase volatility in financial markets. ...

Credit market innovation does not appear to have resulted in a large increase in leverage in the corporate sector, as some had feared. ...

Default rates do not appear to have risen, nor recovery rates fallen as these credit innovations have spread, despite concerns they might lead to excess lending, the mis-pricing of credit risk and more messy and more complicated workouts, resulting from the greater diffusion of the investor base.

And although the sources of the broad moderation in GDP volatility observed in the United States over the past two decades are still the subject of debate, the fact that this moderation occurred during a period of extensive innovation in credit and other financial markets should provide some comfort for those who expected the opposite.

Innovations in credit markets are inevitably accompanied by challenges. Indeed, the history of innovation in financial markets provides many examples of periods of rapid change accompanied by fraud and abuse, by challenges in assessing value and risk, by concerns about the adequacy of investor and consumer protection, and by unexpected behavior of prices, defaults and correlations. To some degree, these types of problems are the inevitable consequence of change and innovation.

Although recent experience as well as theory provide some reassurance..., these judgments require qualification. Some aspects of this latest wave of innovation are different in substance ... from their predecessors. ... [B]road changes in financial markets may have contributed to a system where the probability of a major crisis seems likely to be lower, but the losses associated with such a crisis may be greater or harder to mitigate.

What should policymakers to do mitigate these risks?

We cannot turn back the clock on innovation or reverse the increase in complexity around risk management. We do not have the capacity to monitor or control concentrations of leverage or risk outside the banking system. We cannot identify the likely sources of future stress to the system, and act preemptively to diffuse them.

The most productive focus of policy attention has to be on improving the shock absorbers in the core of the financial system, in terms of capital and liquidity relative to risk and the robustness of the infrastructure. ...

The Federal Reserve is actively involved in a range of efforts... The stronger these shock absorbers, the more resilient markets will be in the face of future shocks, and the more confident we can be that banks will be a source of strength and of liquidity to markets in periods of stress and that the financial system will contribute to improved economic performance over time.

Here's more from the Fed from the last few days:

Update: See also "Toothless Fed, Part 2 (Risk Management Shortcomings)" from Yves Smith at naked capitalism.

The U.S. vs. Other Systems of Health Care

Jane Galt says:

I have a new bloggingheads up with Jonathan Chait, during which I complained about the general tendency for health care books to engage in "argument through anecdote", where the data plays a distant second fiddle to the heartrending stories about x person who didn't get good treatment. So single-payer advocates drag out some American woman who didn't get a breast exam until it was too late, and opponents counter with the Canadian guy who died on the waiting list to see an oncologist.

I'm sympathetic to the data-based versus anecdote point she is making, but in this case I wondered if her impression wasn't partly due to what she chooses to read. Ezra Klein comments along these lines, and he also gives some useful comparisons between the U.S. and single-payer systems - the main reason for the post:

Arguing Health Care, by Ezra Klein: Is Megan's problem with health care writing really that the literature is too narrative-driven? Yikes. She should read some issues of Health Affairs, or the Annals of Internal Medicine, or the New England Journal of Medicine. This is not a debate that lacks for data.

Meanwhile, Megan actually gets a few things wrong in her argument with Chait. She suggests that waiting lines are longer in Europe. That's, uh, untrue. France and Germany don't have waiting lists. Americans do, by the way, with around 40% of patients waiting one month or more for elective surgery. She then suggests that moving to a French or Canadian system would require walking back the medicine we actually provide, telling people they can't have MRIs anymore. That's similarly incorrect. Care utilization in France and Germany is as high -- and in France, higher -- than it is in America. But they pay less per unit of care. And the technology isn't radically different. Germany actually has more CT scanners per million than we do, while the French have three less. The French and the Germans both have more physicians per capita and more acute care beds. Oh, and the French and Germans pay less, and don't have 47 million uninsured.

All this information -- and more! -- can be found in various data-heavy books on the subject, like Thomas Bodenheimer and Kevin Grumbach's wonderful Understanding Health Policy. The thing is, they tend to point towards the same conclusions Jon Cohn's book does, albeit with fewer anecdotes. One reason I spend less time arguing health care with libertarians these days is that it doesn't seem productive. If you really don't want to believe that other system's in the world are better, you won't. If the costs, outcomes, access, and equity advantages offered by the French, German, Japanese, Scandinavian, or Veteran's Affairs systems don't convince you, you simply don't want to be convinced. There are issues, like card check, where I see how the counterargument could be convincing and understand it is, to some degree, a values judgment. Health care isn't one of those issues.

There's a bit more here.

May 16, 2007

William Easterly: Africa's Poverty Trap

William Easterly has spent much of his career applying economic analysis to very poor countries:

Africa's Poverty Trap, by William R. Easterly, Commentary, WSJ: There is a sad law I have noticed in my economics career: the poorer the country, the poorer the economic analysis applied to it. Sub-Saharan Africa, which this month marks the 50th anniversary of its first nation to gain independence, Ghana, bears this out.

There has been progress in many areas over the last 50 years -- ... yet the same poor economics on sale to Ghana in 1957 are still there today. Economists involved in Africa then and now undervalued free markets, instead coming up with one of the worst ideas ever: state direction by the states least able to direct.

African governments are not the only ones that are bad, but they have ranked low for decades on most international comparisons of corruption, state failure, red tape, lawlessness and dictatorship. Nor is recognizing such bad government "racist"... Instead, corrupt and mismanaged governments ... reflect the unhappy way in which colonizers artificially created most nations, often combining antagonistic ethnicities. Anyway, the results of statist economics by bad states was a near-zero rise in GDP per capita for Ghana, and the same for the average African nation, over the last 50 years.

Why was state intervention considered crucial in 1957? Africa was thought to be in a "poverty trap," since the poor could not save enough to finance investment necessary to growth. Free markets could not get you out of poverty. The response was state-led, aid-financed investment. Alas, these ideas had already failed the laugh test... The U.S. in 1776 was at the same level as Africa today, yet it escaped the poverty trap. The same was also true for the history of Western Europe, Australia, Japan, New Zealand and Latin America. All of these escapes from poverty happened without a state-led, aid-financed "Big Push."

In the ensuing 50 years, there have been plenty more examples of poor countries which grew rapidly without much aid -- China and India (who each receive around half a percent of income in foreign aid) being the most famous recent examples. Meanwhile, aid amounted to 14% of total income year in and year out in the average African country since independence.

Despite these reality checks, blockbuster reports over the last two years by the U.N. Millennium Project (led by Jeffrey Sachs), Mr. Sachs again in his book "The End of Poverty," the U.N. Development Program (UNDP), the Tony Blair Commission for Africa, and the U.N. Conference on Trade and Development (Unctad) have all reached what the UNDP called "a consensus on development": Today Africa needs another Big Push. Do they really think nobody is paying attention?

Africa's poverty trap is well covered in the media, since it features such economists as Angelina Jolie, Madonna, Bono and Brad Pitt. But even Bill Gates ... expressed indifference to Africa's stagnant GDP, since "you can't eat GDP." Mr. Gates apparently missed the economics class that listed the components of GDP, such as food.

The World Bank and the International Monetary Fund have good economists who have criticized state intervention. Under the pressure of anti-market activists, alas, they have soft-pedaled these views lately in favor of ... U.N.-led Millennium Development Goals...

The cowed IMF and the World Bank never mention the words "free market" in thousands of pages devoted to ending poverty. ... World Bank economists are so scared of offending anyone on Africa that they recite tautologies. The press release describing the findings of the 2006 World Bank report "Challenges of African Growth" announces: the "single most important reason" for Africa's "lagging position in eradicating poverty" ... is "Africa's slow and erratic growth." The next World Bank report may reveal that half a dozen beers has been identified as the single most important reason for a six-pack.

Today Unctad (in its 2006 "Big Push" report) still offers to make possible government "infant-industry policies" for "small, fragmented economies" by setting up a regional market, presumably so Burkina Faso and Niger can help absorb the potential output of the Togolese automobile industry.

Unctad lacks everything but chutzpah: All aid to Africa, it said, should be moved into a new U.N. Development Fund for Africa, to which Unctad helpfully offered its "in-house experience"... Unctad will thus permit the economics of Africa to at last "escape from ideological biases," so we can finally understand "why economic activity should not be left entirely to market forces."

The free market is no overnight panacea; it is just the gradual engine that ends poverty. African entrepreneurs have shown what they are capable of. They have, for example, launched the world's fastest growing cell phone industry to replace the moribund state landlines. What a tragedy, therefore, that aid agencies have foisted the poorest economics in the world on the poorest people in the world for 50 years. The hopeful sign is that many independent Africans themselves are increasingly learning the economics of how to get rich, rather than of how to stay poor.

With that attitude you almost want him to be wrong. There seems to be little love lost between Sachs and Easterly. Right or not, Sachs is clearly well-intentioned.

A couple of comments, or questions rather about his examples in support of the free market approach to development. He says "In the ensuing 50 years, there have been plenty more examples of poor countries which grew rapidly without much aid -- China and India ... being the most famous recent examples." Are these examples of free markets at work once government stepped aside, or are they cases where the state has provided substantial direction as the big push to get the ball rolling? Should we wonder why he doesn't mention countries where the strict free-market approach has failed and paved the way for populist alternatives?

Nobody knows for sure what the secret is to escaping poverty, and the answer may lie somewhere between the strict free-market and the heavy-handed state intervention approaches. But whatever the answer, given what we know presently, the case for a strict free-market approach is not as clear as Easterly implies.