Has the last thirty years of macroeconomics been a waste of time?
World Bank chief economist, Paul Romer, has again written a very important critique of modern macroeconomics – or more specifically the academic mainstream variant which dominates graduate teaching at the world’s leading universities and typically forms a basis for formal central bank research, if not actual policy practice. Aspects of Romer’s article have been discussed insightfully by Simon Wren Lewis and Chris Dillow.
As always with big-picture observations on an area of study, there will be exceptions that prove the rule. Tony Yates lists a number of them (storified here by Toby Nangle), and no doubt has more in his back-pocket. It should also be said that observing that the broad trend in macroeconomics has been regressive, does not mean that there are not some great macro-economists in academia (there clearly are), nor that profoundly interesting and important work has been done.
I want to tie-in Romer’s observations, with a related but different set of arguments I made several years ago the lack of impact of macro research on policy-making. I concluded that Milton Friedman (and Robert Lucas) were the last policy-relevant macro economists.
My own perspective is that the most important change in macroeconomics, as it relates to policy, in the last thirty years concerns the role of expectations. Of course, this started with Friedman and Phelps in the 1960s, and arguably earlier. Expectations gained further prominence in the 1970s and 1980s following the Lucas critique and the development of rational expectations.
But something changed in the last 30-years. The focus on expectations has become much more deeply ingrained in modelling and policy-thinking. When Samuelson, Solow, Friedman, Phelps thought about expectations, they were pragmatic, and sceptical. Expectations were not always and everywhere the heart of the matter, either. Rational expectations was also heavily critiqued in the mainstream when I was taught economics. What has happened in the last thirty years is that inflation expectations and the real policy rate, which is simply the nominal interest rate less expected inflation, dominate models.
Never before in economic history did the entire macroeconomic framework hang on these three premises:
1. Economic agents have inflation expectations.
2. The policy rate, adjusted for these expectations, determines demand.
3. Central banks (or in the case of the FTPL, central banks and the fiscal authorities) control these expectations.
Romer has something to say about this, although it is not his primary example. His core argument concerns the “identification problem”, which reveals “how difficult it is to make reliable inferences about causality from observations on variables that are a part of a simultaneous system”. Consider a simple example: do falling stock prices cause recessions, or do recessions cause stock markets to fall?
The case Romer focuses on is the relationship between money and output, and specifically the causality between the actions of the Volcker Fed the subsequent US recession. More relevant, from my perspective, however, is his brief discussion of expectations:
“… allowing for the possibility that expectations influence behaviour makes the identification problem at least twice as bad. This may be part of what Sims (1980) had in mind when he wrote, ‘It is my view, however, that rational expectations is more deeply subversive of identification than has yet been recognised’. Sim’s [sic] paper, which is every bit as relevant today as it was in 1980, also notes … that the number of parameters that need to be pinned down scale as the square of the number of variables in the model; and attributes the impossibility of separating the constant term from the expectations term to Solow (1974).”
Let’s relate this back to points 1-3 above. I think the assumption that policymakers control inflation expectations, and that economic agents all have economically-relevant inflation expectations, is totally ingrained and central to all mainstream macro modelling and thinking. In this regard there is no real difference between the fiscal theory of the price level, new Keynesian macro, and even significant parts of the heterodoxy, such as market monetarists – who believe expectations about nominal GDP determine everything.
To be very clear that these assumptions lie at the heart of contemporary macroeconomics, consider the following observations from Greg Mankiw, and Chris Sims. Mankiw and Weinzierl conclude a very useful overview of modern macroeconomics with a set of implied policy conclusions:
“The second level of the [policy] hierarchy applies when the short-term interest rate hits against the zero lower bound. In this case, unconventional monetary policy becomes the next policy instrument to be used to restore full employment. A reduction in long-term interest rates may be sufficient when a cut in the short-term interest rate is not. And an increase in the long-term nominal anchor [expected inflation, nominal GDP, money growth] is, in this model, always sufficient to put the economy back on track. This policy might be interpreted, for example, as the central bank targeting a higher level of nominal GDP growth.”
Mankiw continues by making explicit that the central bank controls expectations and that expectations are everything, saying that the only circumstances where monetary policy is redundant is where, “the central bank is unable to commit to future monetary policy actions.”
Sims, who together with Olivier Blanchard, and Gautti Eggertsson is mentioned in Mankiw’s acknowledgements, concludes his recent talk of the Fiscal Theory of the Price level (FTPL), with a straightforward expectations-based policy prescription:
“What is required is that fiscal policy be seen as aimed at increasing the inflation rate, with monetary and fiscal policy coordinated on this objective.”
Sims’s prescription for Japan is noteworthy:
“In Japan, this might be achieved by explicitly linking planned future increases in the consumption tax to hitting and maintaining the inflation target.”
This entire approach to macroeconomics is, in many ways very odd. And Romer’s criticism of the legacy of Friedman’s methodological observations – which is a very tired and old debate – does seem apposite. These assumptions are testable. An obvious question to ask is ‘Do people have inflation expectations based on what policy-makers announce, and do these expectations change in response to policy-maker commitments about the future?’ Similarly, do expectations about future fiscal policy affect inflation expectations? Do inflation expectations affect wages? Do changes in central bank nominal targets affect wages?
A macroeconomist, who has done a lot of very interesting work in the last thirty years, who is broadly supportive of modelling of inflation expectations, Chris Carroll, points out that, “there has been almost no recent effort to model actual empirical expectations data […]”. In a similar vein, it is worth quoting at length, Olivier Blanchard, who writes in his recent, superb, review of DSGE models (discussed here by Simon Wren-Lewis):
“Go back to the benchmark New Keynesian model, from which DSGEs derive their bone structure. The model is composed of three equations: an equation describing aggregate demand; an equation describing price adjustment; and an equation describing the monetary policy rule. At least the first two are badly flawed descriptions of reality: Aggregate demand is derived as consumption demand by infinitely lived and foresighted consumers. Its implications, with respect to both the degree of foresight and the role of interest rates in twisting the path of consumption, are strongly at odds with the empirical evidence. Price adjustment is characterized by a forward-looking inflation equation, which does not capture the fundamental inertia of inflation [italics and bold added].”
It is somewhat ironic, that I am questioning the relevance of expectations, given that it is in financial markets where expectations might be most relevant. Even there, however, practice tells me that myopia and risk preferences and behavioural factors are as important if not more so than expectations about the future. Does anyone really have an expectation for the outlook for interest rates over thirty years?
I am very sceptical that price and wage setting in the competitive private sector is affected by policy commitments by the central bank and fiscal authorities. As I have argued before, a forward-looking, semi-rational description of inflation determination and expectations formation, affected by policy actions, may well have described economic reality in the 1970s, and parts of the world today which are heavily indexed and have centralised wage bargaining and other institutional features typical of some emerging markets. But in the developed world, Greenspan’s description of price stability as “an environment in which inflation is so low and stable over time that it does not materially enter into the decisions of households and firms” may well describe the prevailing regime.
A consequence of the persistence of this regime, and the institutional and structural changes that have occurred in our economies, could well leave us operating in a world where most economic agents have no relevant inflation expectations, and to the extent that some do, they do not believe that policymakers, using existing policy tools, control prospective inflation. Other factors may be far more relevant to their economic beliefs – such as what they are actually experiencing.
The recent policy announcements by the Bank of Japan are a profound test of Romer’s hypothesis. There can now be no doubt that the BoJ is following as bluntly and aggressively as it is reasonable to expect the policy prescriptions of modern macroeconomics, as Martin Sandbu outlines. It is committing to print money until inflation rises above its current target of 2%, and since 2013 it has demonstrated an unstinting determination to pursue an increase in inflation expectations.
The Bank of Japan introduced QQE in 2013. In their thorough assessment of the effects of their policies this month, they leave no doubt about their persistent attempts to raise inflation expectations, they state:
“The main transmission channel of QQE [the set of policies introduced in 2013] would be the reduction in real interest rates. Namely, (1) people’s deflationary mindset would be dispelled and inflation expectations would be raised through the Bank’s large-scale monetary easing under its strong and clear commitment to achieving the price stability target of 2 percent. At the same time, (2) downward pressure would be put on nominal interest rates across the entire yield curve through the Bank’s purchases of JGBs. (3) Together, these developments would reduce real interest rates. (4) The decline in real interest rates would lead to an improvement in the output gap. (5) The improvement in the output gap, together with rising inflation expectations, would push up the observed inflation rate. (6) Once people experienced an actual rise in the inflation rate, they would adapt their inflation expectations, resulting in higher inflation expectations and further reinforcing this process.”
This month, they also augmented their huge QE programme with “an “inflation-overshooting commitment” in which the Bank commits itself to expanding the monetary base until the year-on-year rate of increase in the observed consumer price index (CPI) exceeds the price stability target of 2 percent and stays above the target in a stable manner.”
How this experiment ends will not just be important for the Japanese economy, but for the future of economics.
Claudia Sahm pointed out this piece of research by the Fed, reinforcing Blanchard’s observations above.
This research from the Boston Fed is also important: what limited evidence there is suggests the opposite sign in the relationship between expected inflation and consumption, in the US, at least.
This BoJ study is more ambiguous in its conclusions, but certainly undermines assumption that raising inflation expectations raises consumption growth via a real interest rate effect, or that higher inflation expectations do not result in negative income effects.
An interesting discussion ensued on Twitter, with comments from Paul Romer, Narayana Kocherlakota, James Mackintosh, Danny Blanchflower and others.