Larry Summers’ blog

What you need to know about the next recession


How should we respond to the next recession? That was the topic of an event held by the Brooking’s Hamilton Project where I spoke on Monday in Washington with White House budget director Sean Donovan. I argued a number of points that address issues of current concern.

First, I argued that the possible election of “Demagogue Donald” dwarfs congressional dysfunction as a threat to American prosperity. I argued that beyond lunatic and incoherent budget and trade policies Trump would for the first time make political risk of the kind usually discussed in the context of Argentina, China or Russia relevant to the USA. How else to interpret threats to renegotiate debt, prosecute insubordinate publications, and rip up treaties? Creeping fascism as an issue dwarfs macroeconomic policy!

Second, I cautioned that while 2009 could have seen a repeat of 1929-1933 and that did not happen there are no grounds for complacency. As the picture below illustrates on current forecasts the economy will have performed as badly over the 2007-2018 period as it did over the Depression 1929-1940 period. The single most important issue for containing government debt burdens, increasing us national security, encouraging more generosity towards the poor and raising middle class standards of living is accelerating US economic growth.

Graph 1

Third, I argued — following my secular stagnation thesis — that fiscal policy is now important as a stabilization policy tool in a way that has not been the case since the Depression. Historical evidence suggests a better than even chance of an officially declared recession in the next three years.graph 2


When recessions come, the Federal Reserve normally reduces real rates by four to five percentage points.

graph 3

But there will in all likelihood be nothing like this amount of room when the next recession comes.

Graph 4

I say this with full awareness that the Fed has unconventional tools at its disposal in addition to simply lowering rates. But I think it very unlikely that more than 150bps of Fed funds equivalent additional stimulus is feasible. After all rates below minus 50 or 75 bps are impracticable in a society with cash and might actually hurt financial intermediation. Forward guidance is fine in principle but when the next recession comes expected forward rates will be very low far into the future. And QE is surely already hitting diminishing returns with the yield curve flattening and markets functioning without the illiquidity premia of the early recovery period. “Helicopter money” is basically a form of fiscal policy as I shall argue in a subsequent post and cannot be carried out autonomously by the central bank.

There is an additional case for fiscal policy. The economy as it now stands requires remarkably low interest rates to grow adequately. These rates are an invitation to leverage, to reaching for yield, to financial engineering and to bubbles. Raising rates significantly as many suggest without doing anything else risks recession. So the right strategy is to raise demand so as to make financially sustainable growth. This comes back to fiscal policy along with measures like tax, regulatory and immigration reform to spur private demand.

Fourth. I suggested a number of areas for expansionary fiscal policy both to make recession less likely and to respond when the next one comes. The decline in US infrastructure investment is indefensible in light of recent declines in interest rates, employment opportunities and materials costs.

Graph 5

Other areas in which fiscal support seems desirable include housing — where residential investment still lags badly — and support for social security. I referred to the economists argument going back to Paul Samuelson that pay as you go social security can help all generations in a world where growth rates exceed interest rates. Further it raises demand without enlarging government deficits.

As I expect to discuss in subsequent posts, much of what economists thought they knew about macroeconomic policy needs to be reassessed in light of events. Just as the events of the 1970s and emergence of stagflation throughout the industrial world, led to new policy paradigms, I believe that recent events will force us to develop new approaches to thinking about economic fluctuations and inflation which will, in turn, drive major changes in thinking about fiscal and monetary policy.


Data collection is the ultimate public good.


On Wednesday I spoke at a World Bank conference on price statistics.  While price statistics are not usually thought of as a scintillating subject, I got a great deal of satisfaction out of preparing and presenting my remarks.  In part this was because my late father Robert Summers focused his economic research on International price comparisons.  It was also because I am convinced that data is the ultimate public good and that we will soon have much more data than we do today.

I made four primary observations.

First, scientific progress is driven more by new tools and new observations than by hypothesis construction and testing.  I cited a number of examples: the observation that Jupiter was orbited by several moons clinched the case against the Ptolemaic system, the belief that all celestial objects circle around the Earth.  We learned of cells by seeing them when the microscope was constructed.  Accelerators made the basic structure of atoms obvious.

Second, if mathematics is the queen of the hard sciences then statistics is the queen of the social sciences.  I gave examples of the power of very simple data analysis. We first learned that exercise is good for health from the observation that in 1940s London bus conductors had much lower death rates than bus drivers.  Similarly data demonstrated that smoking was a major killer decades before the biological processes were understood.  At a more trivial level, Moneyball shows how data-based statistics can revolutionize a major sport.

Third, I urged that what “you count counts” and argued that we needed much more timely and complete data. I noted the centrality of timely statistics to meaningful progress towards Sustainable Development Goals.  In comparison to the nearly six year lag in poverty statistics, it only took the USA about three and a half years to win World War II.

Fourth, I envisioned what might be possible in a world where there will soon be as many smart phones as adults.  With the ubiquitous ability to collect data and nearly unlimited ability to process it will come more capacity to discover previously unknown relationships. We will improve our ability to predict disasters like famines, storms and revolutions.  Communication technologies will allow us to better hold policymakers to account with reliable and rapid performance measures.  And if history is any guide we will gain capacities on dimensions we cannot now imagine but will come to regard as indispensable.

This is the work of both governments and the private sector.  It is fantasy to suppose data, the ultimate public good, will come into being without government effort.  Equally, we will sell ourselves short if we stick with traditional collection methods and ignore innovative providers and methods such as the use of smart phones, drones, satellites and supercomputers. That is why something like the Billion Prices Project at MIT, which can provide daily price information, is so important. That is why I am excited to be a Director and involved with Premise – a data company that analyzes information people collect on their smartphones about everyday life, like the price of local foods – in its capacity to mobilize these technologies as widely as possible. That is why Planet Labs, with its capacity to scan and monitor environmental conditions, represents such a profound innovation.

If we really valued excellence, we would single it out.


Catherine Rampell’s column this week reminded me of an issue that has long interested me as an economist and as a president of Harvard.

I remember many years ago listening to some monetarist quote Milton Friedman one too many times with “inflation is always and everywhere a monetary phenomenon.”  I responded “what about grade inflation?”

I could never quite decide whether this was just a wisecrack or it captured something important.  The idea that it might well reflect is that inflation has much to do with different actors seeking to leapfrog each other and in the process setting off a spiral.

In any event, I think that the pervasiveness of top grades in American higher education is shameful. How can a society that inflates the grades of its students and assigns the top standard to average performance be surprised when its corporate leaders inflate their earnings, its generals inflate their body counts, or its political leaders inflate their achievements?

More than ethics classes this is a matter of moral education. And America’s universities are failing when “A” is the most commonly-awarded grade.  If we really valued excellence, we would single it out.

I did succeed in a small way as Harvard president in reducing the fraction of students graduating with honors from a ludicrous 90 percent to an excessive 55 percent.  I wish I had been able to do more. Even more I wish that today’s academic leaders would take up this issue.

Vast double standard on American college campuses


It has seemed to me that a vast double standard regarding what constitutes prejudice exists on American college campuses.  There is hypersensitivity regarding prejudice against most minority groups but what might be called hyper-insensitivity with respect to anti-Semitism.

At Bowdoin College, holding parties with sombreros and tequila is deemed to be an act of prejudice against Mexicans.  At Emory, the chalking of an endorsement of the likely Republican Presidential candidate on a sidewalk is deemed to require a review of security tapes.  The existence of a college named after widely admired former US President has under the duress of a student occupation been condemned at Princeton.  At Yale,  Halloween costumes are the subject of administrative edict.  The dean of Harvard Law School has acknowledged that hers is a racist institution, while the Freshman Dean at Harvard College has used dinner placemats to propagandize the student body on aspects of diversity.   Professors acquiesce as students insist that they not be exposed to views on issues like abortion that make them uncomfortable.

All I have discussed in the past, this is in my view inconsistent with basic American values of free speech and open debate.  It fails to recognize that the a proper liberal education should cause moments of acute discomfort as cherished beliefs are challenged.

But, if comfort is elevated to be a preeminent value, the standard should be applied universally.  Unfortunately, there is a clear exception made on most university campuses for anti-Semitic speech and acts.

The State Department has made clear that it regards demonizing Israel or  “applying double standards by requiring of it a behavior not expected or demanded  of any other democratic nation” as anti-Semitism.   This makes obvious good sense.  Does anyone doubt that applying standards to African countries that were not applied to other countries or singling them out for sanction when other non-African countries were guilty of much greater sins would be deemed racism?

Instances of anti-Semitism by this standard are ubiquitous in American academic life.  Nearly a dozen academic associations have enacted formal boycotts of Israeli institutions and in some cases Israeli scholars.  Student governments at dozens of universities have demanded the divestiture of companies that do business in Israel or the West Bank.   Guest speakers and even some faculty in their classrooms compare Israel with Nazi Germany and question its right to continued existence as a Jewish state.

Yet, with very few exceptions, university leaders who are so quick to stand up against microagressions against other groups remain silent in the face of anti-Semitism.  Indeed, many major American universities including Harvard remain institutional members of associations that are engaged in boycotts of Israel.  The idea of divesting Israel is opposed only in the same way that divesting apartheid South Africa was opposed—as an inappropriate intrusion into politics, not as immoral or anti-Semitic.

That is why the recent statement of the University of California Regents is so welcome.  It is forceful and clear on anti-Semitism, while at the same time recognizing the importance of free speech.  It holds that “Anti-Semitism, anti-semitic forms of anti-Zionism and other forms of discrimination have no place at the University of California”.  Let us hope that similar statements will be made by the leaders of private and public universities across the country.

Corporate profits are near record highs. Here’s why that’s a problem.


As the cover story in this week’s Economist highlights, the rate of profitability in the United States is at a near-record high level, as is the share of corporate revenue going to capital.   The stock market is valued very high by historical standards, as measured by Tobin’s q ratio of the market value of the nonfinancial corporations to the value of their tangible capital.  And the ratio of the market value of equities in the corporate sector to its GDP is also unusually high.

All of this might be taken as evidence that this is a time when the return on new capital investment is unusually high.  The rate of profit under standard assumptions reflects the marginal productivity of capital.  A high market value of corporations implies that “old capital” is highly valued and suggests a high payoff to investment in new capital.

This is an apparent problem for the secular stagnation hypothesis I have been advocating for some time, the idea that the U.S. economy is stuck in a period of lethargic economic growth.  Secular stagnation has as a central element a decline in the propensity to invest leading to chronic shortfalls of aggregate demand and difficulties in attaining real interest rates consistent with full employment.

Yet matters are more complex.  For some years now, real interest rates on safe financial instruments have been low and, for the most part, declining.  And business investment is either in line with cyclical conditions or a little weaker than would be predicted by cyclical conditions.  This is anomalous, as in the most straightforward economic models the real interest rate is the risk adjusted rate of return on capital.  And an unusually high rate of investment would be expected to go along with a high rate of return on existing capital.

How can this anomaly be resolved?  There are a number of logical possibilities.  First, the riskiness associated with capital investment might have gone up and so higher rates of return could be simply compensating for higher risk rather than implying attractive investments.  There are two major problems with this story.  One is that available proxies for risk are not especially high in recent years.  The chart below depicts realized stock market volatility and the VIX measure of expected volatility as implied by options.  Another problem is that if capital returns have become far more uncertain, then the stocks should have become less attractive in recent years rather than more.  In the last 7 years, the stock market has risen to 250 percent of its spring 2009 levels.

VIX and Realized Stock Mkt Volatility







A second explanation could be that a heightened demand for liquidity and a shortage of Treasury instruments, perhaps created by quantitative easing programs, has driven down bond yields, widening the spread between the rate of profit and these yields.  This story does not provide a natural explanation for the relatively weak behavior of business investment.  Further as Sam Hanson, Robin Greenwood, Joshua Rudolph and I pointed out in earlier work, the market is today being asked to absorb an abnormally high rather than an abnormally low level of long term Federal debt.

Long Term Treasury







On top of that, if Treasuries were in short supply, one would expect that they would have their yield bid down relative to market synthesized safe instruments.  Yet the so-called swap spread is actually negative and Treasury yields (vs swaps) are unusually high relative to history.

Ten Year Swaps






Third, it could be that higher profits do not reflect increased productivity of capital but instead reflect an increase in monopoly power.  If monopoly power increased one would expect to see higher profits, lower investment as firms restricted output, and lower interest rates as the demand for capital was reduced.  This is exactly what we have seen in recent years!

Is the increased monopoly power theory plausible?  The Economist makes the best case I have seen for it noting that (i) many industries have become more concentrated (ii) we are coming off a major merger wave (iii) there is some evidence of greater profit persistence among major companies (iv) new business formation has declined (v) overlapping ownership of companies that compete has become more common with the rise of institutional investors, (vi) leading technology companies like Google and Apple may be benefiting from increasing returns to scale and network effects.

The combination of the fact that only the monopoly power story can convincingly account for the divergence between the profit rate and the behavior of real interest rates and investment, along with the suggestive evidence of increases in monopoly power makes me think that the issue of growing market power deserves increased attention from economists and especially from macroeconomists.


Trump’s rise illustrates how democratic processes can lose their way


While comparisons between Donald Trump and Mussolini or Hitler are overwrought, Trump’s rise does illustrate how democratic processes can lose their way and turn dangerously toxic when there is intense economic frustration and widespread apprehension about the future. This is especially the case when some previously respected leaders scurry to make peace in a new order — yes Chris Christie, I mean you.

The possible election of Donald Trump as President is the greatest present threat to the prosperity and security of the United States.  I have had a strong point of view on each of the last ten presidential elections, but never before had I feared that what I regarded as the wrong outcome would in the long sweep of history risk grave damage to the American project.

The problem is not with Trump’s policies, though they are wacky in the few areas where they are not indecipherable. It is that he is running as modern day man on a horseback—demagogically offering the power of his personality as a magic solution to all problems—and making clear that he is prepared to run roughshod over anything or anyone who stands in his way.

Trump has already flirted with the Ku Klux Klan and disparaged and demeaned the female half of our population.  He vowed to kill the families of terrorists, use extreme forms of torture, and forbid Muslims from coming into our country.  Time and again, he has claimed he will crush those who stand in his way; his promised rewrite of libel laws, permitting the punishment of The New York Times and Washington Post for articles he does not like, will allow him to make good on this threat.

Lyndon Johnson’s celebrated biographer, Robert Caro, has written that while “power doesn’t always corrupt…[it] always reveals.” What will a demagogue with a platform like Trump’s who ascends to the presidency do with control over the NSA, FBI and IRS?  What commitment will he manifest to the rule of law? Already Trump has proposed that protesters at his rallies “should have been roughed up.”

Nothing in the way he campaigned gave Richard Nixon a mandate for keeping an enemies list or engaging in dirty tricks.  If he is elected, Donald Trump may think he has such a mandate.  What is the basis for doubting that it will be used?

To be sure there are precedents in American politics for Trump. Precedents like Joe McCarthy, George Wallace, and Huey Long.  Just as Trump does, each mined the all too rich veins of prejudice, paranoia and excess populism that lie beneath American soil.  Yet even at their highest points of popularity, none of these figures looked like plausible future Presidents.  One shudders to think what President Huey Long would have done during the Depression, what President Joe McCarthy would have done at the height of the Cold War, or what President George Wallace would have done at the end of the turbulent 1960s.

My Harvard colleague, Niall Ferguson, suggests that William Jennings Bryan is the right precursor for Trump. This comparison seems unfair to Bryan who was a progressive populist but not a thug, as evidenced by the fact that he ended up as Secretary of State in the Wilson Administration.  Trump’s election would threaten our democracy.  I doubt that democracy would have been threatened if Bryan had beaten McKinley.

Robert Kagan and others have suggested that Trump is the culmination of trends under way for decades in the Republican Party.  I am no friend of the Tea Party or of the way in which Congress has obstructed President Obama.  But the suggestion that Trump is on the same continuum as George W. Bush or even the Republican congressional leadership seems to me to be quite unfair.

Even the possibility of Trump becoming President is dangerous.  The economy is already growing at a sub-two percent rate in substantial part because of a lack of confidence in a weak world economy.  A growing sense that a protectionist demagogue could soon become President of the United States would surely introduce great uncertainty at home and abroad. The resulting increase in risk premiums might well be enough to tip a fragile US economy into recession.  And a concern that the US was becoming protectionists and isolationist could easily undermine confidence in many emerging markets and set off a financial crisis.

The geopolitical consequences of Donald Trump’s rise may be even more serious. The rest of the world is incredulous and appalled by the possibility of a Trump presidency and has started quietly rethinking its approach to the United States accordingly.  The US and China are struggling over influence in Asia.  It is hard to imagine something better for China than the US moving to adopt a policy of ‘truculent isolationism.’  The Trans-Pacific Partnership, a central element in our rebalancing toward Asia, could collapse.  Japan would have to take self-defense, rather than reliance on American security guarantees more seriously.  And others in Asia would inevitably tilt from a more erratic America towards a relatively steady China.

Donald Trump’s rise goes beyond his demagogic appeal. It is a reflection of the political psychology of frustration – people see him as responding to their fears about the modern world order, an outsider fighting for those who have been left behind. If we are to move past Trumpism, it will be essential to develop convincing responses to economic slowdown.

The United States has always been governed by the authority of ideas, rather than the idea of authority.  Nothing is more important than to be clear to all Americans that the tradition of vigorous political debate and compromise will continue.  The sooner Donald Trump is relegated to the margins of our national life, the better off we and the world will be.


In defense of killing the $100 bill


By Peter Sands and Lawrence Summers

Our advocacy for ending the printing of high denomination notes — first in a working paper by Peter and colleagues and, later in a post by Larry — have been attacked on the ground that this proposal represents an infringement on liberty (for example, see here and here).  Most prominently, the Wall Street Journal  concludes an editorial with the remarkable assertion “Beware politicians trying to limit the way you can conduct private economic business.  It never turns out well.”

There are two levels of illogic here.  First, even the most ardent libertarian recognizes the need for antifraud statues, limits on what can be contracted, and requirements of tax withholding, so the general principle that government should not affect the conduct of private business is absurd.

Second, it is surely a stretch to assert that ceasing government’s active effort to circulate and print 500 euro notes or $100 bills constitutes a government infringement on the conduct of private economic business.  Do the Journal editors believe that liberty was constrained by the US decision in the 1960s to stop printing $1000 dollar bills or to stop issuing bearer bonds?  Surely it is not a government’s obligation to provide every means of payment or store of value that someone might choose to use.   Nor, it should be emphasized, did the abolition of the $1000 put us on any kind of slippery slope to monetary perdition.

Our advocacy for the elimination of high denomination notes is based on a judgment that any losses in commercial convenience are dwarfed by the gains in combatting criminal activity, not any desire to alter monetary policy or to create a cashless society.  We take no position on the desirability of negative interest rates but are convinced by the arguments of JPMorgan,  Miles Kimball and others that significantly negative rates can, if desired, be maintained without any limitation on currency through bank withdrawal fees.  And we believe that for the foreseeable future there will be a role for cash in modern economies though we would not be surprised if in many contexts its transactions costs come to exceed those of various electronic payment schemes.

No ATMs in Europe offer 500 euro notes and very rarely do American ATMs offer $100 bills, so we are highly skeptical that legitimate commerce for the vast majority of law abiding citizens will be impacted.  Only the affluent will be affected at all and we do not think that giving a grandchild five $20 dollar bills rather than a single $100 or tipping a caddy with three $50 bills is too great an inconvenience relative to the benefit of inhibiting criminal activity.

There are 30 US $100 bills in circulation for every man woman and child, and more than 300 billion euro in 500 euro notes.  As Peter’s paper documents, the vast majority of this currency is involved with activity that is at a minimum problematic, and often criminal. This inference for Europe is supported by the observation that high denomination note issuance relative to GDP is nearly 20 times as high in Luxembourg, a traditional haven for illicit activity, relative to the rest of Europe.

We agree with the philosophical position of the Wall Street Journal that governments should do only what is necessary and that excessive government activity is dangerous.   That is why we favor an end to the further printing of high denomination notes.

Peter Sands is a Senior Fellow at the Mossavar-Rahmani Center for Business and Government at the Harvard Kennedy School and the former Chief Executive Officer at Standard Chartered Bank.

Lawrence Summers is a professor at and past president of Harvard University. He was treasury secretary from 1999 to 2001 and an economic adviser to President Obama from 2009 through 2010. He serves as an advisor or board member to a number of financial technology and payments companies.




Increasingly Convinced of the Secular Stagnation Hypothesis


Foreign Affairs has just published my latest on the secular stagnation hypothesis.  I am increasingly convinced that it captures what is going on in the industrialized world and that the risks of long term weakness on the current policy path are growing.

Unfortunately since I put forward the argument in late 2013, the data have been all too supportive.  Despite monetary policy being much more expansionary than was expected and medium term interest rates falling rapidly, growth and inflation throughout the industrial world have been much lower than anticipated.  This is exactly what one would expect if structural factors were increasing saving propensities relative to investment propensities.

Bond markets are now saying that neither inflation rates approaching 2 percent targets or real interest rates substantially above zero are on the horizon anytime in the foreseeable future.  Growth forecasts are being revised downwards in most places and there is growing evidence in the United States that inflation expectations are becoming unanchored to the downside.

I would put the odds of a US recession at about 1/3 over the next year and at over ½ over the next 2 years.   There is a substantial chance that widening credit spreads, a strengthening dollar as Europe and Japan plunge more deeply into the world of negative rates, and lower inflation expectations will be tightening financial conditions even as recession looms.  And while there is certainly scope for QE, for forward guidance and possibly for negative rates it is very unlikely that the Fed can take steps that are nearly the functional equivalent of 400 basis point cut in Fed funds that is normally necessary to respond to an incipient recession.

If I am right in these judgements, monetary policy should now be focused on avoiding an economic slowdown and preparations should be starting with respect to the rapid application of fiscal policy.  The focus of global coordination should shift from clichés about structural reform and budget consolidation to assuring an adequate level of global demand.   And policymakers should be considering the radical steps that may be necessary if the US or global economy goes into recession.

Of course, real time economic theorizing is problematic and I cannot be certain that I am reading the current situation accurately.  If the Fed succeeds in significantly raising rates over the next two years without a growth slowdown and if inflation accelerates to 2 percent, I will conclude that the secular stagnation hypothesis was overly alarmist in confusing cyclical elements with long term problems.  If on the other hand, the economy turns down even at current interest rates, secular stagnation will have to be taken more seriously by policymakers than is currently the case.



It’s Time To Go After Big Money


The Mossavar-Rahmani Center for Business and Government at Harvard that I am privileged to direct has just issued an important paper by Senior Fellow Peter Sands and a group of student collaborators.  Sands’ paper makes a compelling case for stopping the issuance of high denomination notes like the 500 euro note and 100 dollar bill or even withdrawing them from circulation.

I remember that when the euro was being designed in the late 1990s, I argued with my European G7 colleagues that skirmishing over seigniorage by issuing a 500 euro note was highly irresponsible and mostly would be a boon to corruption and crime.  Since the crime and corruption in significant part would happen outside European borders, I suggested that to paraphrase John Connally it was their currency, but would be everyone’s problem.  And I made clear that in the context of an international agreement, the US would consider policy regarding the $100 bill.  But because the Germans were committed to having a high denomination note, the issue was never seriously debated in international fora.

The fact that – as Sands points out — in certain circles the €500 is known as the “Bin Laden” confirms the arguments against it.  Sands’ extensive analysis is totally convincing on the linkage between high denomination notes and crime. He is surely right that illicit activities are facilitated when a million dollars weighs 2.2 pounds as with the €500 note rather than more than 50 pounds as would be the case if the $20 was the high denomination note. And he is equally correct in arguing that technology is obviating whatever need there may ever have been for high denomination notes in legal commerce.

What should happen next?  I’d guess the idea of removing existing notes is a step too far.  But a moratorium on printing new high denomination notes would make the world a better place.  In terms of unilateral steps, the most important actor by far is the European Union.  The €500 is almost six times as valuable as the $100.  Some actors in Europe, notably the European Commission, have shown sympathy for the idea and ECB chief Mario Draghi has shown interest as well.  If Europe moved, pressure could likely be brought on others, notably Switzerland.

I confess to not being surprised that resistance within the ECB is coming out of Luxembourg (see also), with its long and unsavory tradition of giving comfort to tax evaders, money launderers, and other proponents of bank secrecy and where 20x as much cash is printed relative to GDP compared to other European countries.

These are difficult times in Europe with the refugee crisis, economic weakness, security issues and the rise of populist movements.  There are real limits on what it can do to address global problems.  But here is a step that will represent a global contribution with only the tiniest impact on legitimate commerce or on government budgets.  It may not be a free lunch, but it is a very cheap lunch.

Even better than unilateral measures in Europe would be a global agreement to stop issuing notes worth more than say $50 or $100.  Such an agreement would be as significant as anything else the G7 or G20 has done in years.  China, which is hosting the next G-20 in September, has made attacking corruption a central part of its economic and political strategy. More generally, at a time when such a demonstration is very much needed, a global agreement to stop issuing high denomination notes would also show that the global financial groupings can stand up against “big money” and for the interests of ordinary citizens.

Global security and pandemic risk


I was privileged this morning to deliver keynote remarks at the release event for the Global Health Risk Framework Commission report on “The Neglected Dimension of Global Security: A Framework to Counter Infectious Disease Crises”.  The commission, convened by the National Academy of Medicine and chaired by Peter Sands, has delivered a very important report on what I think is the issue with the highest ratio of seriousness to policy preparation in the global system.  Indeed, for reasons I sketched in my remarks, I believe the threat to global well-being from pandemics over the next century is comparable to the threat from global climate change.

I began by expanding on why I think pandemics are such an important issue.  The global mortality rate from the flu pandemic of 1918 was 7000 times as large as from the recent Ebola outbreak.  AIDS profoundly changed the human experience in Africa.  No one knows the probability of a recurrence of these kinds of disasters.  History is too short to permit reliable estimates and in any event conditions are rapidly changing because of scientific improvements on the one hand, and huge increases in global interconnection on the other.

In the context of the Global Health 2035 report that Dean Jamison and I developed, we considered the economic benefit of mortality reduction using approaches derived from economic theory.  We noted that the well established fact that people demand higher pay to accept riskier jobs demonstrates that reduced mortality risk has economic value.  And we further showed that these benefits are very large relative to standard estimates of the economic impact of health interventions.

In related work, Dean and I with Victoria Fan calculate the potential cost of a 1918 flu recurrence, discounted by its chance of happening.  We expect to publish this work soon. On plausible assumptions we find an expected flu cost approaching $1 trillion a year going forward into the 21st century. This underscores the urgency of doing all that can be done to counter pandemic risks.

Eisenhower famously said that “in preparing for battle, I have always found that plans are useless but planning is indispensable”. So also with the pandemic threat, which is what makes the GHRF report so important.  I have three main takeaways from its consideration of the Ebola experience and its stock-taking of the current global architecture.

First, this is an area where the urgent has crowded out the profoundly important.  In too many countries, pressing near term needs and budgetary pressures have prevented the establishment of necessary infrastructures for public health. Such resilient health systems would be a high payoff investment even if catastrophe never comes and would be transformative if and when catastrophe comes.

Second, the international system needs to do more to clarify responsibilities and authorities when the next emergency comes and to put in place financial mechanisms for rapid action.  Given the nonlinear exponential character of pandemic processes this is of the highest urgency.

Third, there is a need for more and better science both ex-ante and ex-post after a pathogen is identified.  I will never forget my friend Barry Bloom, former Dean of Harvard’s School of Public Health, asking me how I would prefer to have spent money in the 1950s: on scholarships for iron lungs or on supporting Salk and Sabin.  There is much here that science could be doing but currently is not.  Given that profiting hugely in time of pandemic is unacceptable, markets provide insufficient incentives for gaps to be filled. As such, pre-emptive research is fundamentally important.


A postscript to Delong and Krugman





Brad is unpersuaded by my response.  He is broadly right in my view that models function both as discovery tools and as ways of organizing and codifying thought. The danger comes when too primitive a model is regarded as too powerful a discovery. Thus Krugman was right to recognize that first generation models of currency crises were an inadequate basis for making policy in response to many actual currency crises.

On the issue at hand my judgement is that excessive spending and fear of a sudden stop can push the IS curve back leading to contractionary capital outflows. I cite the work of Blanchard et al in support of this proposition. With a bit of political economy, the argument can be extended. Suppose countries in danger of high inflation are more likely to turn populist.  It was Keynes after all who introduced animal spirits and warned Roosevelt about business confidence.

Again, I don’t think any of this is of concern for the US right now and I think my track record in favor of fiscal expansion notably in my work with Brad is clear enough.

The point though is that responsible policy makers like Janet Yellen and Stan Fischer may sometimes come to judgements different from mine. I find it unhelpful and likely wrong to attribute this to a failure to understand and appreciate basic macroeconomics.


Paul now offers some observations. He is right that we are not all very far apart. I agree with him that one needs more than attitude, one needs a logically consistent view of how the world works. As my response to Mike Spence and Kevin Warsh illustrates I too am impatient with fear mongering attitude as an approach to analysis.

We have I think two remaining disagreements. First, I am more willing than Paul to credit the possibility that people with substantial experience and even a track record of making money by predicting markets, have important insights even if they cannot speak the language of “models” in the way I teach in economics. Hyman Minsky is an example of a scholar whose warnings were ignored in part because they were not formalized not because they were incoherent or illogical.

Second, on the issue of confidence crises I think Paul is way too serene for reasons Blanchard’s work makes clear. If loss of confidence in their government or economy makes people less wealthy, they will spend less and that is contractionary. This objection may for some reason be wrong but I do not see why it should be dismissed apriori. This is a case where I think reliance on formalism may lead people astray.

Thoughts on Delong and Krugman blogs


Brad Delong and Paul Krugman accept my criticisms of Fed thought regarding their monetary policy strategy but disagree with my assertion that it reflects an excessive attachment to existing models and modes of thought.

Their argument is that standard IS-LM leads to the conclusion you should not raise rates in the present environment so no move away from orthodoxy is necessary to reach this conclusion.  I think the issue is more on the supply side than the demand side.  If I believed strongly in the vertical long run Phillips curve with a NAIRU around five percent and in inflation expectations responsiveness to a heated up labor market, I would see a reasonable case for the monetary tightening that has taken place.

Since I am not sure of anything about the Phillips curve and inflation is well below target I come down against tightening.  The disagreement does it seems to me come down to the Fed’s attachment to the standard Phillips curve mode of thought.  My disagreement is reinforced by other judgmental aspects that are outside of the standard model used within the Fed.  These include hysteresis effects, the possibility of secular stagnation, and the asymmetric consequences of policy errors.

I am sure Paul and Brad are right that a desire to be “sound” also influences policy.  I am not nearly as hostile to this as Paul.  I think maintaining confidence is an important part of the art of policy.  A good example of where market thought is I think right and simple model based thought is I think dangerously wrong is Paul’s own Mundell-Fleming lecture on confidence crises in countries that have their own currencies.  Paul asserts that a damaging confidence crisis in a liquidity trap country without large foreign debts is impossible because if one developed the currency would depreciate generating an export surge.

Paul is certainly correct in his model but I doubt that he is in fact. Once account is taken of the impact of a currency collapse on consumers’ real incomes, on their expectations, and especially on the risk premium associated with domestic asset values, it is easy to understand how monetary and fiscal policymakers who lose confidence and trust see their real economies deteriorate as Olivier Blanchard and his colleagues have recently demonstrated.  Paul may be right that we have few examples of crises of this kind but if so this is perhaps because central banks do not in general follow his precepts.

I do not think this is a pressing issue for the US right now.   But the idea that policymakers should in general follow the model and not worry about considerations of market confidence seems to me as misguided as the view that they should be governed by market confidence to the exclusion of models.

The Fed and Financial Reform – Reflections on Sen. Sanders op-Ed


Bernie Sanders had an op Ed in the New York Times on Fed reform last week that provides an opportunity to reflect on the Fed and financial reform more generally. I think that Sanders is right in his central point that financial policy is overly influenced by financial interests to its detriment and that it is essential that this be repaired. At the same time, reform requires careful reflection if it is not to be counterproductive. And it is important in approaching issues of reform not to give ammunition to right wing critics of the Fed who would deny it the capacity to engage in the kind of crisis responses that have judged in their totality been successful in responding to the financial crisis.  The most important policy priority with respect to the Fed is protecting it from stone age monetary ideas like a return to the gold standard, or turning policymaking over to a formula, or removing the dual mandate commanding the Fed to worry about unemployment as well as inflation.

Sanders is right that Fed governance has been and is overly tied up with the financial sector. Each of the 12 regional Fed s has a board of directors that is made up of 9 people—three banking representatives, three private sector non-banking representatives and three public interest representatives. The fact that a member of Goldman Sachs’ board at the time of the 2008 crisis was the “public interest” Chairman of the New York Fed board is to put it mildly indefensible.

More generally, it is not clear why a government institution with vital policy responsibilities should have a role in governance for the private sector. Yes, advice on market conditions and the like is needed but this can be sought from advisory committees. Yes, Dodd Frank did clean things up some by removing bankers from the selection process for regional bank presidents and orienting regulatory responsibility towards the Washington. But it is hard to imagine an appropriate governance activity for business figures with respect to the Federal Reserve System. Nor is it clear why banks should in any sense be “shareholders” in the Federal Reserve System.

I am less certain as to whether Sanders particular reform of making regional presidents senate confirmed is wise given the dysfunctionalities of the Senate process. Also, I suspect that this reform would strengthen the NY Fed with its close ties to Wall Street and hawkish regional presidents in ways that Sanders would not like.

A separate issue from Sanders concern with governance is his concern with who holds the full time senior positions within the Fed.  He asks rhetorically whether the CEO of Exxon should be head of the EPA or FCC should be stuffed with former Verizon executives. This is a difficult issue.  There is a tension between acquiring expertise and avoiding cooptation or cognitive capture. In fact many FCC chairs have had industry backgrounds including both Obama FCC chairs who pushed net neutrality approaches that industry loathed. It would be a valuable study but it is not my impression that as a general matter officials who come from industry have been systematically softer on industry than those who have come from other backgrounds.   It is worth pondering that figures much respected for their commitment to aggressive financial regulation like Arthur Levitt came into their positions from full time roles in industry whereas the principal regulators in place before the 2008 crisis—Bernanke, Geithner, Cox– came from public sector backgrounds. Franklin Roosevelt was hardly a pushover to the financial industry. He famously made former stock-operator Joseph Kennedy the inaugural leader of the SEC on the theory that it takes a thief to catch a thief.

Sanders proposes to make the Fed more transparent and accountable by releasing not just minutes but transcripts six months after meetings rather than the current five years. I am not sure I understand the logic here. Monetary policy accountability is essential but so is accountability for decisions of war and peace or protection of the environment. No one expects transcripts from the Pentagon’s war room or the EPA senior staff meetings. The Supreme Court Justices meet alone, without clerks or stenographers, because the best decisions tend to come when policymakers can deliberate privately before reaching conclusions. This encourages out of the box thinking and forceful dissent while minimizing grandstanding. Why should monetary policy be different? It seems to me that Janet Yellen has done great work over the years in pushing the Fed to be more open and that if further steps are to be taken, they should be in the directions she has pioneered such as more frequent press conferences.

With respect to proposals to audit the Fed, I think the issues are really of truth in labeling. There is no question that the Fed like every other part of government should be subject to independent audit. My understanding is that this is currently the case. If there are lacunae in current procedures, these should be pointed up and repaired. This is not the focus of current proposals like those of Ron and Rand Paul who would prefer to abolish the central bank and whose ideas are not so much about auditing the Fed as subjecting it to political control and straitjacketing it.  Surely Senator Sanders should not wish to curtail the Fed’s ability to act with discretion to insure a continued flow of credit to businesses and households in the event of an economic downturn or financial crisis.

On the substance of monetary policy, I have been clearly on the dovish side of the Fed for quite sometime and think the risks of the last rate increase exceeded the benefits so I agree with Sanders general thrust. I prefer the “do not raise rates until you see the whites of inflation’s eyes” to Sanders rather arbitrary 4 percent unemployment target.

I did not feel strongly but am inclined to agree with Sanders that the Fed should not have paid interest in excess reserves while it was setting the Fed funds rate in the zero range.  However, when rates are raised it is a matter of necessity that the Fed either pay interest on reserves or what is essentially equivalent offer Treasury bills to banks which soak up their excess cash. In a positive interest rate environment, there is no way that banks will or should hold on to zero interest rate cash. I do not think there would be any expert support for Sanders views here.

On regulatory policy, no one is for gambling with insured deposits. But Sanders fails to recognize some of the tensions that make regulatory policy so difficult. Loans to small businesses which he likes are far riskier than holdings of securities that are marked to market on a daily basis so if banks focused on traditional lending they would be riskier than they are today. Indeed the majority of the world’s banking crises over the last three centuries and over the last quarter century have come from traditional lending especially against real estate.  Making banks safer means reducing their dependence on traditional lending activities so balances must be struck.

Sanders asserts as many do that Glass Stegall’s repeal contributed to the crisis. I may not be objective as I supported this measure as Treasury Secretary but I do not see a basis for this assertion. Virtually everything that contributed to the crisis was not affected by Glass Steagall even in its purest form. Think of pure investment banks Bear and Lehman, or the GSEs Fannie and Freddie, or the banks Washington Mutual and Wachovia or AIG or the growth of the shadow banking system. Nor were the principle lending activities that got Citi and Bank of America in trouble implicated by Glass Stegall.

Moreover preventing financial institutions from diversifying into multiple activities can actually make them more likely to fail. Imagine how much better the outcome would have been if Lehman had sold itself to a large bank during 2008. There is the also the further point that without the repeal of Glass Stegall it would have been much more difficult to aggressively use the discount window to contain panic following Lehman s fall.

We currently have a system in which institutions have higher capital requirements if they are larger, more complex or more interconnected. I think there is a good case to be made that capital requirements should be further increased and further graduated with size and complexity. This would tend to encourage deconsolidation and is I think the right 21st century response to the concerns about concentration in banking.

Finally, stepping back from Sanders specifics, I agree with his core claim that the excessive power of financial interests shapes financial policy to an unhealthy extent. I saw this during the debate on Dodd Frank when there were nearly five registered lobbyists for every member of Congress. It’s worth pondering how policy would be different if special interests were kept in check. Here are my top five:

First, the financial regulatory agencies would be adequately resourced and would not be under pressure to kowtow to legislators pushing their contributors interest. CFTC head Tim Massad had it right when he condemned the recent budget agreement as undercutting the ability to regulate derivatives in a serious way.

Second, the Balkanized character of US banking regulation is indefensible and would be ended. The worst regulatory idea of the 20th century—the dual banking system—persists into the 21st. The idea is that we have two systems one regulated by the States and the Fed and the other regulated by the OCC so banks have choice. With ambitious regulators eager to expand their reach, the inevitable result is a race to the bottom.

Third, the current SEC and CFTC would be combined and charged with regulating in a coherent way all financial markets with respect to market integrity, manipulation issues, insider trading, transparency, fairness of execution, and systemic risk. When there were securities markets and commodity markets it may have made sense. It no longer is defensible when the vast majority of “commodity trading” is in financial derivatives. The current system persists only so that multiple congressional committees can maintain jurisdiction over financial regulation and reap the benefits in terms of campaign contributions.  Quite possibly, it would be a good idea to give the new agency if it was strong jurisdiction over fiduciary rules for investment and pension advisors.

Fourth, either the new agency formed out of the SEC and CFTC or the existing FSOC would take on systemic risks associated with asset management in a serious way. While the asset managers have the better side of the argument when they claim not to be systemic in the sense of JP Morgan or Goldman Sachs, their activities are systemic and egregiously under regulated. It took far far too long to reach a still unsatisfactory solution with respect to money market funds. And ETFs are as likely as anything else to be the source of the next major bit of financial drama.

Fifth, there would be appropriate taxation of financial activities and the financial sector. Among the obvious reforms held back only by special interests are the highly preferential treatment of carried interest, the privileged treatment of dividends and capital gains, the ability of financial institutions to reduce their tax liabilities by using off shore tax havens and the tax deductibility of huge fines paid to resolve allegations of wrongdoing.

Senator Sanders is right that there is more to be done to respond to the long standing problems pointed up by the financial crisis and he is right in his concern about the way in which special interests distort the process. These notes will have served their purpose if they help to channel legitimate energy in the most constructive directions.


My views and the Fed’s views on secular stagnation


It has been two years since I resurrected Alvin Hansen’s secular stagnation idea and suggested its relevance to current conditions in the industrial world.  Unfortunately experience since that time has tended to confirm the secular stagnation hypothesis.  Secular stagnation is a possibility.  It is not an inevitability and it can be avoided with strong policy.  Unfortunately, the Fed and other policy setters remain committed to traditional paradigms and so are acting in ways that make secular stagnation more likely.

The core idea behind secular stagnation was that the neutral real rate had for a variety of reasons fallen and might well be below zero a substantial part of the time going forward.  The inference was that economies might be doomed to oscillate between sluggish growth and growth like that of the 2003-2007 period that rested on an unstable financial foundation.

In more technical economic language secular stagnation is the hypothesis that the IS curve has shifted back and down so that the real interest rate consistent with full employment has declined.  More straightforwardly if you see weaker growth despite lower real interest rates that tends to confirm the secular stagnation idea.

The table below shows what has happened since the Fall of 2013 in the industrial world.  Growth has been consistently weaker than was expected.  This has occurred even though interest rates have fallen substantially suggesting that it reflects a reduction in demand.  Further evidence comes from the decline in inflation expectations.  If the dominant shock were slower productivity one might expect to see an increase in inflation.  Whatever risk one saw of secular stagnation two years ago, one has to be more concerned today.

graph for secular stagnation blog posting 122215










Here is a transcript of a lecture (along with the associated power point) given at the Bank of Chile Research Conference that updates my thinking on secular stagnation.  A central argument I make is that even in the United States it is unlikely that we have left the zero lower bound behind.  Indeed I would judge that there is at least a two-thirds chance that we will experience zero or negative rates again in the next five years.

This pessimism is strongly confirmed by the landmark Rachel and Smith study which shows that there is little basis for believing that neutral real rates will rise in the next few years.

To its credit, judging by the “dots” embodying long run rate projections as well as Chair Yellen’s press conference, the Fed has started to recognize the idea of a declining neutral real rate.  Yet I believe its decision to raise rates last week reflected four consequential misjudgments.

First, the Fed assigns a much greater chance that we will reach 2 percent core inflation than is suggested by most available data.  Inflation swaps suggest inflation on the Fed’s preferred PCE deflator measure will average only 1 percent over the next 3 years, 1.2 percent over the next 5 years and 1.5 percent over the next 10 years.  Survey measures of expected inflation are falling not rising.  Moreover, if account is taken of quality change inflation measures would have to be further reduced.

Second, the Fed seems to mistakenly regard 2 percent inflation as a ceiling not a target.  One can reasonably argue that after years of below target inflation, it is appropriate to have a period of above target inflation.  This is implied by arguments for price level targeting.  Alternatively, it seems reasonable to simply suggest that the Fed should run equal risks of over and under shooting its inflation target.  I would actually argue given the observed costs of deflation that the costs of under shooting the target exceed the costs of overshooting it.

Third, the Fed seems to be in the thrall of notions that might be right but do not to my knowledge have analytic support premised on the idea that the rate of change of interest rates as distinct from their level influences aggregate demand.  It is suggested that by raising rates the Fed gives itself room to lower them.  This is tautologically true but I know of no model in which demand will be stronger in say 2018 if rates rise and then fall than if they are kept constant at zero.  Nor conditional on their reaching say 3 percent at the end of 2017 do I know of a reason why recession is more likely if the changes are backloaded.  I would say the argument that the Fed should raise rates so as to have room to lower them is in the category with the argument that I should starve myself in order to have the pleasure of relieving my hunger pangs.

Fourth, the Fed is likely underestimating secular stagnation.  It is failing to recognize its transmission from the rest of the world and it is overestimating the degree of monetary accommodation now present and likely to be present in the future by overestimating the neutral rate.  I suspect that if nominal interest rates were 3 percent and inflation were far below target there would be much less pressure to raise them than there has been of late.  The desire to raise rates reflects less some rigorous Philips curve analysis than a sense that zero rates are a sign of pathology and an economy creating 200,000 jobs a month is not diseased.  The complexity is that zero rates may be less abnormal than is supposed because of fundamental shifts in the saving investment balance,

Why is the Fed making these mistakes if indeed they are mistakes? It is not because its leaders are not thoughtful or open minded or concerned with growth and employment.  Rather I suspect it is because of an excessive commitment to existing models and modes of thought.  Usually it takes disaster to shatter orthodoxy.  We can all hope that either my worries prove misplaced or the Fed shows itself to be less in the thrall of orthodoxy than it has been of late.


What Should the Fed Do and Have Done?


The Federal Reserve meets this week and has strongly signaled that it will raise rates.  Given the strength of the signals that have been sent it would be credibility destroying not to carry through with the rate increase so there is no interesting discussion to be had about what should be done on Wednesday.

There is an interesting counterfactual discussion to be had.  Should a rate increase have been so clearly signaled? If rates are in fact going to be increased the answer is almost certainly yes.  The Fed has done a good job of guiding expectations towards a rate increase while generating little trauma in markets.  Assuming that the language surrounding the rate increase on Thursday is in line with what the market expects, I would be surprised if there are major market gyrations after the Fed statement.

But was it right to move at this juncture? This requires weighing relative risks.  A decision to keep rates at zero would have taken several risks.  First, since monetary policy acts only with a lag failure to raise rates would risk an overheating economy and an acceleration of inflation possibly necessitating a sharp and destabilizing hike in rates later.  Second, keeping rates at zero would risk encouraging financial instability particularly if there became a perception that the Fed would never raise rates.  Third, keeping rates at zero leaves the Fed with less room to lower rates in response to problems than it would have if it increased rates.

Finally, perhaps zero rates have adverse economic effects.  Perhaps economic actors take the continuation of zero rates as evidence that the Fed is worried and so they should be as well.  Some believe that zero rates are a sign of pathology and we no longer have a pathological economy and so no longer should have zero rates.  Or perhaps there is a fear that when rates go up something catastrophic will happen and this source of uncertainty can only be removed by raising rates.

These arguments do not seem hugely compelling to me.  Inflation is running well below 2 percent and there is not yet much evidence of acceleration.    Decades of experience teaches that the Phillips curve can shift dramatically so reasoning from the unemployment rate to inflation is problematic.  Declining prices of oil and other commodities suggest inflation expectations may actually decline.  Furthermore, if one believes that productivity is understated by official statistics one has to as a matter of logic believe that inflation is overstated.  I have recently argued that this is quite likely the case given the rising importance of sectors like health care where quality is difficult to measure.

Even if one assumes that inflation could reach 2.5 percent, this is not an immense problem.  There is no convincing evidence that economies perform worse with inflation marginally above 2 percent than at 2 percent. Then there is the question of whether it is better to target the annual rate of inflation or the price level.  On the latter standard it is relevant that inflation over any multiyear interval would still have averaged less than 2 percent.  And I am not sure why bringing down inflation would be so difficult if that were desired especially given that it would surely take a long time for expectations to become unanchored towards the high side of 2 percent.

It seems to me looking at a year when the stock market has gone down a bit, credit spreads have widened substantially and the dollar has been very strong it is hard to say that now is the time to fire a shot across the bow of financial euphoria.  Looking especially at emerging markets I would judge that under-confidence and excessive risk aversion are a greater threat over the next several years than some kind of financial euphoria.

The Fed does not have special information on where the economy is going and I find it highly implausible that its not acting would scare market participants if it explained its decision making. Given the now large body of research suggesting substantial declines in neutral real rates, it is an anachronism to believe that zero rates are only appropriate in pathological economies.








A very important recent study from two Bank of England economists suggests that on a global basis neutral real rates are unlikely to rise much if at all in the next few years.  As I noted in yesterday’s blog, the “temporary headwinds” interpretation of low neutral real rates has been wrong for the last few years and is not hugely plausible as a basis for predicting the next few.  Historically volatility has been a bit higher  for stocks and for the dollar and a bit lower for bonds after the Fed starts hiking than immediately before so I’m not sure of the basis for the belief that “getting it over with” would reduce uncertainty.

3 year graphic






Finally, the reload the cannon argument seems to me entirely specious.  I know of no economic model where raising the average level of rates over the next few years raises the average level of output.  Contractionary policy is contractionary and if there is a risk of a slowdown or recession in the next few years that is surely an argument against contractionary policy.

What about the risk of raising rates?  Certainly the risks of exacerbating financial instability or hitting the brakes when the economy is slowing look much less serious than they did in September.  Nonetheless growth in the second half of 2015 may well come in at less than 2 percent.  There is certainly a real risk that slow speed becomes stall speed becomes recession.  On average mature recoveries like the present one last less than an additional 3 years.  And given how low rates are and the political aversion to the use of fiscal policy a substantial slowdown could have very severe consequences.  It is important to recognize in this regard that once the decline in neutral real rates is recognized policy is much less accommodative than is often supposed.  Indeed on some measures policy rates now are above neutral.

There is also the risk that inflation expectations start to anchor below 2 percent.  This will happen if economic actors conclude as they reasonably can from what they are hearing that the Fed will cap inflation at 2 percent but allow it to fall below 2 percent in periods of economic slack.  If anything the picture coming from survey measures of expectations, from the indexed bond market and from inflation swaps suggests that inflation is expected to remain below 2 percent for a decade.  If projections were conditioned on the Fed’s intended path of monetary policy rather than what the market expects they would be even lower.  And if account was taken of inflation mismeasurement they would be lower still.

All of the argument about appropriate inflation targeting in recent years has focused not on whether 2 percent is too high but on whether it is too low a target.  I am not yet ready to abandon a 2 percent target but wage rigidity and zero lower bound on interest rate considerations suggest whatever the right target was a decade ago, a higher target might be appropriate today given the slowdown in productivity and reduction in the neutral rate.  Surely the risks of missing the target on the low side over a prolonged interval should be a major policy concern today.

This is a time of considerable financial and geopolitical fragility around the world.  There are real risks of serious capital flight and associated dislocation in many emerging markets.  Any change in policy and financial conditions carries with it at least some chance of setting off instability which could snowball given the current high degree of illiquidity in many markets.  The risks are magnified by the asymmetry between what the Fed is doing and what most other major central banks are doing.  While in principle exchange rate movements should not affect the level of global demand, further dollar appreciation is likely to be contractionary for the global economy because of the uncertainty it engenders.

On balance the risks of raising rates seem a little more likely to play out and much more serious than the risks of standing still on rates.   Moreover, given the inevitability of mistakes prudence dictates tilting towards making errors that are reversible.   An excessive delay in raising rates can be remedied eight weeks later at the next FOMC meeting by raising them then.  On the other hand, if rates are raised and it proves to be a mistake there are likely to be substantial costs as inflation expectations move down, financial turbulence ensues, and the economy possibly tips towards recession.  Reversing the rate increase would be unlikely to eliminate these consequences.  Moreover, reversing the direction of policy would hardly be helpful for central bank credibility as the central banks around the world who raised rates and then were forced to reverse themselves have discovered.

Reasonable people can come to different judgements on all of this.  I think on balance it was a mistake to lock in a December rate increase though the argument is closer than it was in September.  But that decision has been made.  I hope the Fed will not now invest its credibility in signaling further increases until and unless there is much clearer evidence of accelerating inflation.  I hope it will also emphasize the two sided character of the 2 percent inflation target to mitigate the risk that markets will think the US has an inflation ceiling rather than target.  Finally, I hope the Fed will signal its awareness of  instability and risk of growing problems in emerging markets.

Breaking new ground on neutral rates


Lukasz Rachel and Thomas Smith have a terrific new paper on world neutral real rates. The fact that it supports a variety of arguments that I have been making on secular stagnation for the last two years may contribute to my enthusiasm but the paper breaks new ground in a number of respects.

First, Rachel and Smith document compellingly the near universality of sharply declining real rates and also the length and breadth of the decline. They show the phenomenon is a very broad based 450 basis point trend decline in real rates occurring over 25 years. Framing the problem this way is significant because it shows the inadequacy of shorter-term explanations of low neutral real rates such as those of Ken Rogoff that focus on the financial crisis and its aftermath. It also suggests the need to look beyond monocausal.

Second, Rachel and Smith do not find that slowing growth is the main explanation for declining real rates. Rather they attempt to quantify most of the factors that I and others have enumerated in accounting for declining real rates. They note that since the global saving and investment rate has not changed much even as real rates have fallen sharply there must have been major changes in both the supply of saving and demand for investment. They present thoughtful calculations assigning roles to rising inequality and growing reserve accumulation on the saving side and lower priced capital goods and slower labor force growth on the investment side. They also note the importance of rising risk premia associated in part with an increase financial frictions. Rachel and Smith’s work is not the last word but it is the first important word on decomposing the causal factors behind declining real rates.

Third, Rachel and Smith use their analysis of the determinants of neutral real rates to predict their future evolution. Here they reach the important conclusion that there is little basis for assuming a significant increase in neutral real rates going forward. This conclusion differs sharply from the “headwinds” orthodoxy prevailing in the official community. As the figure below illustrates, in the United States at least the Fed and the forecasting community has been consistently far behind the curve in recognizing that the neutral real rates has fallen. If, as I suspect, Rachel and Smith are right there will be much less scope to raise rates in the industrial world over the next few years than the world’s central banks suppose.

neutral rate news







Fourth, Rachel and Smith recognize that their findings are highly problematic for the existing central banking order. They imply that the zero lower bound is likely to be a major issue at least intermittently going forward. After all, we will have future recessions and when we do, there will be a need to drop rates by 300 basis points or more. Perhaps QE or negative rates or forward guidance will be availing. I am skeptical that they will be efficacious if a recession comes in an economy without a heavily disrupted financial system in need of repair.

Rachel and Smith also share my concern that a world of chronic very low real rates is going to be a world of high volatility, imprudent risk taking, excessive leverage and frequent financial accident. We may be about to get a taste of this in emerging markets and US high yield markets. It is fashionable to invoke the brave new world of macro-prudential policy in response. To borrow from Wilde, I fear that enthusiasm for macroprudential policy is the triumph of hope over experience. In the last wave of enthusiasm for such policies the poster-child was Spain’s countercyclical capital requirements. They did not work out so well. As best I can tell US macroprudential policy as currently practiced has meaningfully impaired liquidity in some key markets and damaged the credit availability for small and medium sized businesses while not touching excessive flows to emerging markets and high yield corporate issuance. To work, macroprudential policy has to reduce financial vulnerabilities without to an equal extent reducing credit flows that stimulate demand. This is logically possible. I doubt that actual regulators who after all were proclaiming the health of the banking system in mid 2008 are capable of pulling it off consistently given the pressures they face.

There are no certainties here. It is possible that neutral real rates will rise over the next several years. But there is a high enough chance that they will not to make contingency planning an urgent priority. That has been and is the main thrust of the secular stagnation argument.

This $13 trillion question is more important than ever


The Hutchins Center for Fiscal and Monetary Policy at Brookings is having a conference launching an important new volume on federal debt management policy.  Just as in the Great War in became clear that war is too important to be left to generals, so too in the Great Recession it became clear that (government) debt management is too important to be left to the parochial world debt managers. The composition of federal debt is itself often a useful tool for economic policy, particularly in the current low rate environment in which the Federal Reserve will frequently be unable to cut rates as much as it would like and will instead be reliant on “unconventional” policies intended to effect the price of government debt.

Edited by David Wessel the volume contains two chapters that I coauthored with Robin Greenwood, Sam Hanson, and Josh Rudolph as well as some separate comments of mine.  There is also a provocative paper by John Cochrane and a variety of perceptive commentaries by people with experience in debt management policy.

I think the volume makes a case for quite radical revisions in thinking about debt management policy.  Here are my 10 main takeaways starting where I am most confident.

  1. Debt management is too important to leave to Federal debt managers and certainly to leave to the dealer community. This is because, especially when interest rates are near zero, it implicates directly monetary and fiscal policy and economic performance in the short run, and questions of financial stability in the medium run.
  2. Whatever one’s view about desirable policy, it is fairly crazy for the Federal Reserve and Treasury, which are supposed to serve the national interest, to pursue diametrically opposed debt management policies. This is what has happened in recent years with the Federal Reserve seeking to shorten outstanding maturities and the Treasury seeking to term them out.
  3. Standard discussions of QE — which focus on the size of Fed purchases of long term bonds and ignore the scale of Treasury sale of these instruments — are intellectually incoherent. It is the total impact of government activities on the stock of debt that the public must hold that should impact financial markets.
  4. The preceding point is highly significant for the United States. Despite QE the quantity of long term debt that the markets had to absorb in recent years was well above, rather than below, normal.  This suggests that if QE was important in reducing rates or raising asset values it was because of signaling effects regarding future monetary policies not because of the direct effect of Fed purchases.
  5. The standard mantra that Federal debt management policies should seek to minimize government borrowing costs is some combination of wrong and incomplete. It is wrong because it is risk adjusted expected costs that should be considered.  It is incomplete because it is hard to see why the effects of debt policies on levels of demand and on financial stability should be ignored.
  6. The tax smoothing aspect, which is central to academic theories of debt policy, is of trivial significance. Even far larger levels of tax variability than we observe, or than could be offset by altered debt management policies, have only trivial impacts on levels of income.  It is much more important to understand debt management policy impacts on financial stability than on tax variability.
  7. The idea of rollover risk that is ever present in policy discussions is very confused. If there is the possibility that a period will come when the government’s borrowing rate will be very high this obviously needs to be considered in setting policy.  But the problem is not one of rollover.  To see this, think about long term floating rate debt.  Such debt does not offer insulation in the hypothetical circumstance where rollover would be difficult, because in such a situation floating rate debt yield will rise precipitously.
  8. Yield curves typically slope upward. The “carry trade” of borrowing short and lending long is a hedge fund staple.  Rather than providing this opportunity Treasury should reverse the trend towards terming out the debt.  Issuing shorter term debt would also help meet private demands for liquid short term instruments without encouraging risky structures like banks engaged in maturity transformation.
  9. Institutional mechanisms should be found to insure that in the future the Fed and Treasury are not pushing debt durations in opposite directions. The Treasury terming out the debt which the Fed then buys in an effort to quantitatively ease serves only to enrich the dealer community.
  10. Now that we are in a “secular stagnation” world of low interest rates, it is likely that debt management tools will be more important to stabilization policies in the future than in the past.

Advanced economies are so sick we need a new way to think about them


An ebook containing the papers and presentations from the European Central Bank’s central banking forum conference in Sintra Portugal is now available. Mario Draghi and his colleagues are to be greatly commended for running a forum that is so open to profound challenges to central banking orthodoxy.

The volume contains a paper by Olivier Blanchard, Eugenio Cerutti and me on hysteresis and separately some of my reflections asserting the need for a new Keynesian economics that is more Keynesian and less new. Here I summarize these two papers.

Hysteresis Effects

Blanchard Cerutti and I look at a sample of over 100 recessions from industrial countries over the last 50 years and examine their impact on long run output levels in an effort to understand what Blanchard and I had earlier called hysteresis effects. We find that in the vast majority of cases output never returns to previous trends. Indeed there appear to be more cases where recessions reduce the subsequent growth of output than where output returns to trend. In other words “super hysteresis” to use Larry Ball’s term is more frequent than “no hysteresis.”

This finding does not in and of itself establish the importance of hysteresis effects. It might be that when underlying growth rates fall recessions follow but that recessions have no causal impact going forward. In order to address this issue we look at the impact of recessions with different precursors. We find that even recessions that are associated with disinflationary monetary policies or the drying up of credit have substantial long run output effects suggesting the presence of hysteresis effects.

In subsequent work Antonio Fatas and I have looked at the impact of fiscal policy surprises on long run output and long run output forecasts using a methodology pioneered by Blanchard and Leigh. Since fiscal policy effects operate primarily through aggregate demand, this provides a way to avoid the causation question. We find that fiscal policy changes have large continuing effects on levels of output suggesting the importance of hysteresis.

I was struck that in a vote taken at the conference close to 90 percent of the participants indicated that they believe there are significant hysteresis effects. While there is much more work to be done, I believe that as of right now the right presumption is in favor of hysteresis effects despite their exclusion from the standard models used in almost all central banks.

Towards a New Macroeconomics

My separate comments in the volume develop an idea I have pushed with little success for a long time. Standard new Keynesian macroeconomics essentially abstracts away from most of what is important in macroeconomics. To an even greater extent this is true of the DSGE (dynamic stochastic general equilibrium) models that are the workhorse of central bank staffs and much practically oriented academic work.

Why? New Keynesian models imply that stabilization policies cannot affect the average level of output over time and that the only effect policy can have is on the amplitude of economic fluctuations not on the level of output. This assumption is problematic at a number of levels.

First, if stabilization policies cannot effect average levels of employment and output over time they are not nearly as important as if they can. Beginning the study of stabilization with this assumption takes away much of the motivation for doing macroeconomics.

Second, the assumption is close to absurd.  It is surely reasonable to assume that better policy could have avoided the Depression or the huge output losses associated with the financial crisis without having shaved off some previous or subsequent peak.

Third, contrary to the now common view that macroeconomics is best understood by studying the stochastic properties of stationary time series, the most important macroeconomic events are in some sense one off. Think of the Depression or the Great Recession or the high inflation of the 1970s.

The problem has always been that it is difficult to beat something with nothing. This may be changing as topics like hysteresis, secular stagnation, and multiple equilibrium are getting more and more attention.  As well they should. US output is now about 10 percent below a trend estimated through 2007. If one attributes even half of this figure to the effects of recession and assumes no catch up on this component until 2030 the cost of the financial crisis in the USA is about 1 year’s GDP. And matters are worse in the rest of the industrial world.

As macroeconomics was transformed in response to the Depression of the 1930s and the inflation of the 1970s, another 40 years later it should again be transformed in response to stagnation in the industrial world. Maybe we can call it the Keynesian New Economics.

Where Paul Krugman and I differ on secular stagnation and demand


Paul Krugman suggests that I have had some kind of change of heart on secular stagnation and converged towards his point of view, citing the publication of the transcript of a 2011 debate which we both participated in. I certainly appreciate the gravity of the secular stagnation issue more than I did a few years ago, given the continuing decline in global real interest rates. But I think Paul exaggerates the change in my views considerably.

The topic of the debate was: “North America faces a Japan-style era of high unemployment and low growth.” Paul argued in favor. I opposed the motion – not on the grounds that the US economy was in good shape, but on the grounds that our demand deficiency problems should be easier to solve than Japan’s.

Quoting from my response to Paul’s arguments:

You’re right the United States has a serious demand deficiency. You’re right that not enough is being done to contain that demand deficiency. You’re right that we will suffer needless unemployment and stagnation until more is done to address that demand deficiency….

My thesis is that as serious as that problem is, it is dimensionally much less than the problems that Japan faced in four respects. Japan’s problems were different in magnitude, different in the depth of their structural roots, different in the relative perspective they had — relative to the rest of the world — and different in the degree of resilience their system had for adapting to them…

…It will take time. There are steps that need to be taken but we are a society that works. We are a society whose principal problems — we all up here agree — can be addressed by a change in the printing of money and the creation of infrastructure.”

Paul responded in part by saying:

The question is, are we going to be stuck in a state of depressed demand of the kind that Larry has talked about. Larry and I agree that that is what has been happening… I think Larry and I agree almost entirely on the economics, on what needs to be done”.

(Full transcript available here).

I think we have both been focused on demand and the liquidity trap for a long time.

There are, though, two areas where I have had somewhat different views from Paul. I believe that structural issues are often important for demand and growth. I have often asserted that “business confidence is the cheapest form of stimulus”, and quoted to President Obama Keynes’ famous 1938 letter to Roosevelt:

“Businessmen … are … at the same time allured and terrified by the glare of publicity, easily persuaded to be ‘patriots’, perplexed, bemused, indeed terrified, yet only too anxious to take a cheerful view, vain perhaps but very unsure of themselves, pathetically responsive to a kind word. You could do anything you liked with them, if you would treat them (even the big ones), not as wolves or tigers, but as domestic animals by nature, even though they have been badly brought up and not trained as you would wish… If you work them into the surly, obstinate, terrified mood, of which domestic animals, wrongly handled, are so capable, the nation’s burdens will not get carried to market; and in the end public opinion will veer their way.”

Second, I have never related well to Paul’s celebrated liquidity trap analysis. It has always seemed to me be a classic example of economists’ tendency to “assume a can opener”. Paul studies an economy in liquidity trap that will, by deus ex machina, be lifted out at some point in the future. He makes the point that if you assume sufficiently inflationary policy after this point, you can drive ex ante real rates down enough to stimulate the economy even before the deus ex machina moment.

This is true and an important insight. But it seems to elide the main issue. Where is the deus ex machina? Where is the can opener? The essence of the secular stagnation and hysteresis ideas that I have been pushing is that there is no assurance that capitalist economies, when plunged into downturn, will over any interval revert to what had been normal. Understanding this phenomenon and responding to it seems the central challenge for macroeconomics in this era.

Any analysis that assumes restoration of previous equilibrium is, from this perspective, missing the main issue.  I was glad to see Paul recognize this point recently.  I suspect it will lead to more emphasis on fiscal rather than monetary actions in depressed economies.


This critique of the Fed isn’t backed by logic nor evidence


My friends Mike Spence and Kevin Warsh, writing in yesterday’s Wall Street Journal, have produced what seems to me the single most confused analysis of US monetary policy that I have read this year (Brad DeLong has expressed related views).  Unless I am missing something — which is certainly possible — they make a variety of assertions that are usually exposed as fallacy in introductory economics classes.

My problem is not with their policy conclusion, though I do not share their highly negative view of QE.  There are many harshly critical analyses of QE such as those of Martin Feldstein that are entirely coherent and consistent with the macroeconomics of the last 50 years.  My differences are based on judgements about empirical magnitudes and relative risks not questions of basic logic.

Spence and Warsh do not focus on the inflation risks, financial stability risks or distributional risks from overly expansionary monetary policy.  Instead they assert a proposition that I have not encountered in 40 years as a professional economist—that overly easy monetary policy reduces business investment.  Indeed they blame the weakness of business investment during the current recovery on the Fed.

This line of argument is to say the least surprising.  Every major macroeconomics textbook whether Keynesian monetarist or classical in orientation teaches students that investment increases as real interest rates decline.  This is motivated in a variety of quite compelling ways.  It is noted that lower rates raise the present value of the returns from investment and so make them more attractive.  It is observed that lower rates mean lower borrowing costs or lower costs of drawing down liquid asset holding making the purchase of capital goods more attractive.  For these reasons millions of students have been taught that Hicks famous IS curve slopes downward.  Hundreds of empirical studies have found that investment responds to capital cost as theory predicts though the magnitude is open to debate.

What arguments do Spence and Warsh offer for their heterodox conclusion? They note rightly that monetary policy has been easy and investment has been weak in the current recovery.  This is a little like discovering a positive correlation between oncologists and cancer and asserting that this proves oncologists cause cancer.  One would expect in a weak recovery that investment would be weak and monetary policy easy.  Correlation does not prove causation.

They then argue that low rates somehow promote corporate stock buybacks and this is an alternative to real new investment.  I confess that I cannot follow the logic here.  I would have supposed that the choice between real investments and share repurchases would depend on their relative price.  If as Spence and Warsh assert QE has raised stock prices this should tilt the balance towards real investment.

Likewise I would have thought that by making the return to holding cash less attractive, easy money would tend to drive firms into making real investments.  And for the many firms that do not have huge hoards, it would bring down borrowing costs.

Perhaps Spence and Warsh are on to something that I am missing.  I’m curious whether they can point to any peer reviewed economic research or indeed any statistical work that backs up their views.  I am certainly open to any new evidence or new argument after all that has happened in recent years that easy money reduces business investment.  And there is plenty of room for debate over policy.  For now though I would put the Spence-Warsh doctrine that easy money reduces investment in a class of propositions backed by neither logic nor evidence.