Interview: Jason Furman, former chair of the Council of Economic Advisers
The Harvard economist talks about inflation, wages, and where the Obama administration fell short
Economist Jason Furman was chair of the Council of Economic Advisers during Barack Obama’s second term. Now he’s a professor at Harvard’s Kennedy School. He is also one of the most prolific and well-read contributors on Econ Twitter. His ability to see the big picture of the American economy is matched by very few.
In recent months, Jason has expressed ambivalence about the direction the economy is headed. At a time when inflation worries are rising in the media but the economy is still nowhere near full recovery, we need such sensible and knowledgeable voices to help us chart a path back to good times.
In the interview that follows, I talk to Jason about inflation, wage growth, productivity, and what the Obama administration could have done better.
N.S.: So, you've been fairly ambivalent about where the macroeconomy is headed. You're not sounding the alarm about inflation as much as Larry Summers or Olivier Blanchard, but you're taking the possibility pretty seriously. Has your outlook on near-term inflation and growth changed recently, and if so, how?
J.F.: Before getting to inflation, we should not lose sight of the big picture. What is most important is real economic growth and employment. We don’t have numbers for the second quarter yet but assuming GDP comes in at a 10 percent annual rate we will have had more growth in the first half of 2021 than the Congressional Budget Office had forecast for the entire year. Most forecasters expect GDP to be above pre-pandemic projections by the end of the year. If that happens it would be a remarkable contrast not just to the overly slow recovery from the last crisis but also to previous recessions all of which were associated with output losses.
I believe this is because of the interaction of two factors: first, this was a massive negative supply shock that is going away comparatively rapidly due to vaccination. There is no reason the economy cannot just pick up where it left off now that it is getting safer to do so. But this factor interacts with a second one: the remarkable public policy response that drove a wedge between GDP which fell dramatically and real disposable personal incomes which rose sharply. This has enabled people to increase their spending as it has become safer to do so. The fiscal response was much larger in the United States than in Europe, their comparative economic outlooks over the next year will provide evidence for whether this matters. I expect it will and that the U.S. recovery will be further and faster than Europe’s.
Now that we’ve gotten the very important part out of the way let’s talk about a somewhat important issue: inflation. So far inflation has vastly exceeded what most people expected. It may be transitory (more on that in a moment) but it is worth admitting that even if it is transitory it is much more transitory inflation than almost anyone expected. Back in January and February a number of forecasters did not think we would even hit two percent inflation this year. As recently as May the Survey of Professional Forecasters was expecting 2.1 percent PCE inflation for the entire year, we have already gotten more than that in just the first five months of the year. For the year as a whole inflation is very likely to end up above 3 percent and could even be above 4 percent, amounts that would have—and were—dismissed as scaremongering just a few months ago.
You asked what has surprised me about inflation. First let’s talk about what is not surprising: base effects, pandemic-related prices like restaurants and travel returning to normal, and transitory increases for certain bottleneck-related goods (for example most everything that uses lower-end microchips). Everyone knew those were coming and built them into their forecasts. Other countries are experiencing something similar but the contrast with the United States is telling: over the last twelve months consumer prices have risen 2.0 percent in the Euro area, 2.1 percent in the United Kingdom, and 5.0 percent in the United States (part, but only part, of this difference is due to the fact that we have reopened more). Prices in the United States have been rising at an 8 percent annual rate, definitely a lot of that is special factors but even stripping those special factors out and you are left with something more like a still unusually rapid 4 percent annual rate.
The big surprise to me has been the strength of labor demand and the comparative weakness of labor supply. You see this in record job openings and quits juxtaposed to the labor force participation rate which has barely increased from last summer. This disconnect is evidenced in the most rapid nominal wage gains since the 1980s. Moreover these rapid nominal wage gains come on top of a recessionary period where nominal wage growth never slowed and was even higher for lower-wage workers, something that is extremely unusual and possibly unprecedented. So instead of thinking about the economy as having the substantial amount of slack you would expect from an economy with an unemployment rate of around 6 percent it may have less slack than it has had at almost any point in recent history.
Looking forward, everyone expects the inflation rate to fall from its 8 percent annual rate as bottlenecks get resolved, labor supply returns more fully, and demand cools—particularly for goods. But I still expect it to be elevated because some sectors will see rising prices (shelter is the best candidate), sticky wages and prices will take some time to adjust up, and most importantly, demand is likely to continue to exceed supply because of the combination of lagged effects of past fiscal support, another 3-4 percent of GDP in already enacted fiscal support coming next year, and unprecedentedly loose financial conditions.
The inflation rate could temporarily dip over the remainder of this year as used car and other bottleneck prices come down but then my best guess would be about 3 percent inflation next year—with a massive margin of error around that prediction.
N.S.: OK, so let's talk about inflation a bit more. First of all, I've written that the most likely cause of sustained high inflation is neither supply constraints nor excess spending on its own, but a regime shift -- a public perception that the Fed no longer cares as much about fighting inflation as much as it used to. Do you agree with that? My argument was based on the idea that the Philipps curve is usually quite flat, but can shift substantially in the event of a policy regime shift.
J.F.: I like to think of inflation in terms of micro or bottom-up factors (e.g., what is going on with cars and houses), macro or top-down factors (e.g., what is the balance of supply of demand), and expectations (e.g., what is the regime). I already talked about the previous two and they are a good way to think about inflation over the next year or two but let’s now turn to the regime shift. We may already have had somewhat of a regime shift. It is possible we have had one on fiscal policy given how massively expansionary it has been relative to the past and how no one is talking about reducing the deficit. But I am skeptical given how much the political sytem has generally returned to the view that everything must be paid for. In fact my own view is that we have not had enough of a regime shift on fiscal policy.
Monetary policy is a better candidate for a possible regime shift—one that we might already have had and might go further in the future. Officially the goal has shifted to flexible average inflation targeting, a small shift from inflation targeting that allows a persistent overshoot. Some argue that framework should allow a sustained overshoot given that inflation undershot for over a decade prior to it. More importantly Fed Chair Powell and the other governors (but not the regional Presidents) are talking very differently about monetary policy, one that has been widely celebrated by the full employment community which has talked about how as a non-economist Powell is avoiding the mistakes macroeconomists have led us into. You cannot celebrate those changes (as many have) and then be indignant at the claim of a regime shift. Just look at how the Fed has declared it will not raise rates to even 37 basis points until we are already at maximum employment and inflation is poised to overshoot 2 percent for “some time”. That is a very, very different approach than we have seen in the past.
It is still uncertain, however, how much the monetary policy regime has shifted because the FOMC appears to believe inflation will come down to its 2 percent target next year and stay there. At the current moment it is hard to distinguish between them having a very dovish reaction function or a very dovish forecast. If it is a very dovish forecast then if inflation comes in higher than they expect they will tighten faster and we will be back to something more like the old regime. I view this as the most likely. But another possibility is that they have a very dovish reaction function and even if inflation is closer to 3 percent next year (the center of my very uncertain forecast) they will find another reason not to raise rates.
Ultimately there is a better possibility than at any point in the last three decades that we will have a few years of inflation well above 2 percent. If this happens I don’t think the Fed will try to get inflation back to its target. And then at its next framework review about four years from now it could end up ratifying reality with a higher inflation target like a 2 to 3 percent range or 3 percent inflation. So far the bond markets and professional forecasters are giving very little weight to this possibility and expect them to keep inflation to 2 percent over the longer term. I’m much less sure and place at least a 20 percent chance of a higher inflation target, but possibly that is wishful thinking on my part because I would like them to raise the inflation target.
N.S.: Second of all, why should we worry about inflation? Is it because inflation that's not caused by rising wages will end up reducing real wages, because wages are nominally sticky and can't adjust to keep up? Or some other reason?
J.F.: I would rather we were in a regime of steady about 3 percent average inflation instead of the steady sub 2 percent average inflation we were in for the last decade. The higher inflation target would have some costs in terms of greater inconvenience of adjusting prices and understanding them over time—I would never advocate that we annually change the definition of a foot. But unlike the definition of the foot, changing price levels every year has meaningful benefits in terms of fighting recessions and the terrible toll they take on the most vulnerable workers and possibly even average growth rates over time. Specifically, interest rates have come down over time so the Fed has less ability to cut rates, a higher inflation rate would mean that when interest rates were cut to zero that would be more powerful, technically the real rate would be more negative—but you could think of that as it being cheaper to borrow because you can pay back the debt more easily with all the price and wage increases you are going to get. In addition, an older and less fashionable argument—but no less true argument—is that inflation “greases the wheels of the labor market” where nominal wages cannot be cut then inflation leads to larger real wage reductions in recessions and thus helping to maintain employment.
But in the current environment inflation raises a number of potential concerns and problems, mostly because we are talking about unexpected and inconsistent inflation that falls outside of the generally accepted regime. Those potential concerns and problems include:
First, real wages. We are accustomed to think of a hot labor market leading to larger rises in wages than prices. That is probably true for the types of gradually heated labor markets we had in the late 1990s and prior to the pandemic. It is much less clear if that is the case now. So far nominal wages have risen a lot but prices have risen even more so this has not been a good year for real wages. For it to be a better year we will need to see price increases translating into wage increases (which has not happened for decades but very well might now) and then see wage increases not translate back into price increases.
Second, problems associated with the monetary policy reaction. The Fed massively underestimated inflation so far this year and I fear they continue to put much too little weight on the possibility that inflation above 2 percent is persistent. If we get this persistent inflation then it could cause them to react in ways that surprise markets (as they did at the last FOMC meeting), more abruptly raise rates, and possibly even cause a recession. If the Fed ultimately is committed to 2 percent inflation then it could be painful to bring inflation back down again if it ends up around 3 percent for another year or two.
Third, we have benefited a lot from the very flat Phillips curve, a curve that was flat in part due to anchored expectations. If these expectations become less stable then it could be harder in the future to run experiments like we did in the late 1990s and just prior to the recession where we push the unemployment rate lower and lower. The Fed’s credibility may have allowed those better outcomes. I think and hope we can shift to the same situation but 1 percentage point higher. But it certainly has risks.
Fourth, in the current environment the same factors giving rise to inflation could also be giving rise to housing bubbles and asset price bubbles that threaten to destabilize the economy in the future.
The Fed is operating under substantial uncertainty. Its most important job should be to avoid recessions with their terrible and persistent consequences for the most vulnerable workers. The argument for pushing further is strong in terms of bringing vulnerable workers into jobs and possibly raising real wages. But the argument on the other side is not about the wealthy—they do just fine in recessions—it is about those same workers who are the first to suffer in a recession. So it is really important to get this right and understand it as a balancing of risks for vulnerable workers and not some sort of battle between capital and labor.
N.S.: What about the possibility of "fiscal dominance"? The federal government's debt level is just much higher than it was in the early 80s, when Paul Volcker tamed inflation. In 1980, federal government debt held by the public was about 24% of GDP; now it's about 100%. The average maturity of this debt has not changed much over the years, meaning that much of this is still short-term debt that needs to be rolled over. This means that if the Fed raises rates to fight inflation, it could severely increase the amount of interest that the federal government is required to pay each month. That could force the government into a punishing austerity program that would hurt the economy, on top of the monetary effect of higher interest rates. Is it possible that this will tie the Fed's hands? In other words, even without explicit political interference -- even with perfect central bank independence -- is it possible that the threat of higher interest payments will make the Fed very hesitant to raise rates to fight inflation? And might that lead to the kind of monetary regime change, and spiraling inflation, that you're worried about? Or is fiscal dominance not something that should concern us?
J.F.: It would take a large institutional shift for “fiscal dominance,” or really any fiscal considerations at all, to affect the Fed’s policymaking. They’ve been effectively independent of the Treasury for nearly 70 years now and have acted in a highly independent, anti-inflationary manner for forty years. It’s very far from the DNA of the institution. Look at the way Chair Powell stood up to tremendous public pressure from President Trump, that is what the Fed is mostly about. It is possible we will see such an institutional shift but given that 5 FOMC voters are reserve bank Presidents, the governors have long terms, the Senate needs to confirm them, it would be a difficult and unlikely—but not impossible—process.
Note that the budget forecasts do assume that interest rates rise. CBO expects the federal funds rate to rise to 2.6% by the end of the window and the 10-year Treasury to rise to 3.6%. Even with these assumptions they expect the debt to only grow slowly as a share of GDP and the real interest payments on the debt, the metric I like to look at, to be much lower than they have been historically. Real interest rates have been so low that even if they end up higher they are still likely to be low in historical context.
I should add, if I were sitting in the White House right now I would be more worried about the political implications of sustained higher inflation than about the magnitude of interest payments and the deficit. The later is abstract and any number would still seem (and be) quite large. The former is something people directly experience. So I would probably be rooting for (but not doing anything to achieve) the Fed to at least sound more hawkish to better keep inflation expectations in check. In the short run this motivation is the opposite of the one presumed to underly fiscal dominance.
N.S.: OK, let's hope you're right about that, because fiscal dominance is scary! OK, so to sum up the macro conversation, you think the Fed can and will signal that it's ready to raise interest rates to curb inflation if necessary, and that fiscal discipline of some sort hasn't been completely thrown to the winds, and because of this, inflation will come back down in a reasonable time frame?
J.F.: That’s a pretty good summary, I wasn’t as concise in my answers because I have a tremendous amount of uncertainty around everything. But yes, that all seems like the most likely scenario. And to be clear, I wish fiscal discipline was thrown a little more to the winds than I think it will actually be.
N.S.: Anyway, while I have you here, let's talk about some longer term problems. How do we raise productivity growth in America? Are recent tech breakthroughs going to be enough? How can policy help?
J.F.: My idea for speeding up productivity growth: everything. Probably the single most important policy lever is immigration which both increases population and also brings with it ideas and innovations. This is also a conceptually straightforward, but admittedly politically fraught, lever for the federal government to pull. Another straightforward step is spending more basic research by massively increasing the budgets for the NSF, NIH and other scientific research. Ensuring that people can live and work where they are most productivity is another important step, something that requires land use and occupational licensing reforms by states and cities across the country. I don’t know how much extra productivity growth we would get from corporate tax reform but I think we should try because it could even be done in a revenue-raising way, including the proposals President Biden has made plus allowing businesses to expense their investments, disallowing interest deductions, and possibly expanding the R&D tax credit.
N.S.: And how do we raise real wages?
J.F.: Real wages have grown relatively slowly since the 1970s. They have grown in real terms, you have to take some pretty poor measures of inflation overly seriously to believe anything other than that. But they have not grown as fast as they did in the decades before and they have not grown as fast for workers at the middle or the bottom as for those at the top. The biggest cause of the slowdown in real wage growth is the productivity slowdown, but the increase in inequality has played a larger role as well. Note that causes are not always the same as solutions, but understanding how big a role the productivity slowdown has played does lead me to think much harder about the productivity component of the solution.
The other part of boosting real wages is improving the distribution of income. I don’t personally feel any passion in the holy wars over the causes of the increase in inequality because it I think it has been large enough that there is room for all of the explanations to play a role. Moreover, just because something caused inequality does not necessarily mean that addressing it is the best solution (e.g., to the degree you think technology is the cause there may be not feasible or desirable solution). The amount a worker can produce for an hour of labor plays an important role in how much they are paid, getting people who can make more per hour is clearly part of the solution, and education is the main tool we have here. But pay is determined by more than just marginal products so we should be raising the minimum wage, strengthening labor unions, addressing issues like wage collusion and non-competes. Finally, one of the exciting intellectual developments over the last five years is an increased awareness that reduced product market competition can play a role in labor markets too—so a revitalized competition agenda would help as well.
Macroeconomic policy can play a role as well. Running hot labor markets—ideally one log on the fire at a time—can increase real wages and reduce inequality.
Finally, even if we did everything I wanted—let alone everything that is feasible or actually gets done—we still would probably not have sufficient wage growth. A good set of policies would likely raise real wage growth by a few tenths of a percent per year. A sustained increase of 1 percent per year would be amazing and surprising. Even with that it would take a very long time, if ever, to get to incomes I would view as fully desirable or that make up for the large increase in inequality over the last forty years. So tax and transfer programs of whatever form you like should play a role as well. Note that the distinction between tax and transfer programs and programs for market incomes may be a little blurrier than one would traditionally think, at least when it comes to programs for children whether child allowances, housing vouchers or healthcare, all of which appear to have substantial benefits for longer-run mobility.
N.S.: Regarding pay and productivity, what do you think of the famous EPI graph showing a divergence between the two?
J.F: I’m not a fan of the famous graph because I think it creates a misleading impression about the sources of inequality, the productivity slowdown and the causal relationship between productivity and pay. Before getting to these let me mention a number of well-known technical issues with various iterations of that picture. Most important is that different price indices are generally used for productivity and pay and the former is lower for technical and substantive reasons, overstating the gap between them. Productivity is an average concept for the economy as a whole but pay is often shown as a median or a mean of a subset of the economy. Sometimes the graph is done with just wages so it misses the faster growing sources of compensation like health benefits. EPI acknowledges these issues but many less sophisticated users of the graph do not realize the importance of this nuance. But let me get to the three bigger points.
First, the source of inequality. A portion of inequality reflects the fact that income has shifted between capital and labor. This is the gap between compensation and productivity. But a much larger portion of inequality reflects increased inequality within labor income. That is higher-paid managers getting paid more and regular workers less. The picture glosses over this entirely.
Second, the picture does not emphasize the dramatic slowdown in productivity growth from around 2.8 percent annually from 1948 to 1973 to around 1.8 percent annually from 1973 to the present. That is a huge deal and quantitatively accounts for the majority of the slow wage growth. In part it is hard to see the slowdown on the graph for the technical reason that the Y axis is not shown on a ratio scale (or in logs) so it makes it seem like productivity growth is faster later on than it actually is (moving 1 unit from the baseline to 1% is a much bigger deal than going from from 250% to 251% but both take up the same space on the y axis).
Third, and most important, the graph is often interpreted to mean that productivity and pay are unrelated. But the graph shows aggregate data so cannot tell you anything about what is ultimately a microeconomic claim. Moreover, even the macro data leads to strange conclusions—like if productivity had been 2.8% annually instead of 1.8% annually do you really think pay would have been unchanged? To believe that is to believe the labor share would have fallen from around two-thirds of income to less than one-third of income. That is very unlikely because when businesses are more productive it is not that they want to pay their workers more but they are forced into it by the (imperfect) competition we have.
Other than that… Seriously, inequality is really important, graphs that show pay for workers at different percentiles or incomes for households at different percentiles are powerful, important, and do tell us something real about the economy. Would just use those.
N.S.: Next question: What important things do you think we've learned from the Obama years, in terms of economic policy and/or the politics of getting economic policy passed? What are some mistakes you think the Obama administration made on this front, and how can we do better this time around?
J.F.: Obama’s proposals were better than what actually passed Congress. If he had gotten what he had proposed we would have had a significantly larger stimulus in 2009 and another round in 2011, a cap-and-trade system for climate change, a tax on oil, universal preschool, support for childcare, a large infrastructure program, a substantial set of green investments, a higher minimum wage, laws that made it easier for unions to organize, a reformed UI system with greater coverage, higher replacement rates and automatic triggers, and much much more.
Of course Presidents are judged by what they got done and not by what they proposed. But in the “mistakes” question the first order issue is really one of legislative tactics, what would have gotten more of what Obama proposed done? I don’t know the answer to that. The more common view among progressives is that he should have proposed even more, started from a higher negotiating position and then would have gotten more. A less common but also possible perspective is sometimes when you try for too much you end up with nothing so he should have limited his ambitions. I confess I do not know which of these is right and suspect it varies from issue to issue.
My broader perspective is that Presidents should get whatever they can to improve the economy done. When we had unified control of the Congress in 2009 and 2010 that would have meant getting more purely Democratic stuff done. We largely lived within the constraint of 60 votes in 2009-10 and that might have been a mistake. Freeing ourselves from the constraint earlier on could have led to a bigger fiscal stimulus. But unless we had gotten rid of the filibuster entirely, you would still have needed 60 votes for the Affordable Care Act, cap and trade and Wall Street Reform. Still, I wish we had done more to push through infrastructure—although it is easy to say that in hindsight, at the time it seemed almost impossible to get as much done as we did.
After 2011 we had a Republican Congress. At that time I was all for picking the subset of our agenda that we could do with Republicans and compromising to get that done. I would even have compromised more in some of the fiscal negotiations if it had meant more revenue and smaller discretionary spending cuts than we ultimately got.
My biggest regret is that we didn’t pass immigration reform. Maybe we could have taken advantage of the large majorities in 2009 and 2010? Maybe we could have compromised more with a Republican leadership that genuinely wanted to do it but could not get their rank and file to go along. There is no issue that is more important to the more than 10 million people in the United States without documentation or to the future of the country. It might have been too politically costly in 2010 and too impossible from 2011-16, we certainly tried, but if you ran history again 10 times we should have been able to get immigration done in at least 2 of those re-runs. Unfortunately our actual run of history was not one of those.
N.S.: And final question: Assuming that the Biden administration has limited political capital and can't pass all the policies it would like, what would be your top priorities right now for economic policies? What should we really be pushing for, first and foremost?
J.F.: On his fiscal proposals, I’m even more enthusiastic about the families plan than the jobs plan. The evidence is very clear about investments in children and the beneficiaries of those investments are relatively clear too. So I would put preschool at the very top of the list and then some of the other elements of the families plan including childcare and caregiving (the later is in the jobs plan). The most important piece of the jobs plan is the clean energy standard which would bring net emissions to zero in the power sector in a very efficient manner, unfortunately this may be the piece that has the longest odds, but anything that could be done for it would be good. Anything related to green innovation is good. Biden has not proposed a formal health plan but here the highest priority should be the coverage gap, ensuring that low-income people in the states that have not expanded Medicaid can get coverage, this should come well before anything like expanding Medicare benefits or the Medicare age.
Finally, the fiscal envelope should be expanded as much as possible to fit as many priorities as possible—which is to say, we should both be comfortable with adding a substantial amount to the deficit for well-designed investments and we should also be very willing to raise taxes on high-income households and corporations, including just about everything Biden has proposed.
On the non-fiscal proposals, I would raise the minimum wage as much as possible. Senators Romney and Cotton have proposed $10 per hour, Biden could start the negotiation from there. In fact, the higher you think the minimum wage should be the larger is the problem associated with it being too low. If you happen to think the minimum wage should be $20 an hour then it being $7.25 instead of $10/hour is a much bigger problem than if you think it should be $10/hour. It would be nice to make it easier to unionize as well, although I’m less sure what the compromise would look like there. Finally, I would reform our antitrust rules and put in place a pro-competition digital regulator, something that has at least some support from both sides of the aisle.