Irrational Expectations
Irrational Expectations
Irrational Expectations
viewpointmag.com/2020/04/29/irrational-expectations/
April 29,
2020
However much upheaval the global COVID-19 pandemic has generated, a great deal
more is coming. The economic disaster is already the object of frantic analysis, much of
which tells us we can expect a bottom that matches or exceeds the Great Depression of
the 1930s, at least as measured by conventional economic indicators like GDP,
unemployment, and bankruptcies. This narrative provides the backbeat to the
competing attempts to organize our attention during the passage through present and
future trials.
While we are endlessly reminded that “we are all in this together”—a blatant act of false
solidarity—many have also pointed out that we were never “all in this together” before
the pandemic, we are not now, and it is quite possible we will emerge from this even less
together than we were. At least in terms of wealth and income inequality, the prospects
do not look good. The relatively well-off will weather the lockdown more comfortably and
without the threat of eviction, debt default, and hunger, and they will return to better-
paid and stable work more quickly.
The only way to avoid that is to think hard about the modes of economic organization we
might take this terrible chance to transform. Green New Deals, permanent universal
basic income (and health care where it does not already exist), and more radically
decentralized means of provisioning and dignifying our collective lives—all of these
things and more are possible. A background condition of these possibilities, however, is
an economics that can underwrite and make sense of them. No economics could have
“predicted” the pandemic; what matters is what directions it offers us in its wake. Even
more than the financial crisis that began in 2007-2008, the pandemic has exposed the
glaring inadequacies of existing economic policy, the absurdly fragile economics on
which it relies, and the stunted conception of the economy that underwrites it.
1/12
In particular, it has turned a spotlight on the extraordinary uncertainty of this moment, of
our state institutions’ and elites’ criminal underappreciation of the precariousness of the
social regime over which they rule, and their utter and complete lack of a political
economic analysis or vocabulary to understand the scale and substance of the current
crisis. The economics that has dominated mainstream knowledge for decades cannot
provide a fix for any of this. The insuppressible fact of real uncertainty, its kryptonite, has
again proven all its expectations irrational. In an effort to explain how we can avoid
reviving it, and what we can do to fight the efforts to reimpose it that are sure to come, it
is worth thinking about how we got here, and why we can’t go back.
Macroeconomic “Revolution”
Twice in the last century, the world has forced macroeconomics to reinvent itself. First,
the Great Depression dissolved economists’ widely-held faith in markets’ capacity to self-
correct: if left to their “freedom,” not only were capitalist societies subject to “sub-
optimal” conditions, but the freer the markets, the worse the conditions got. In the space
created by John Maynard Keynes’ General Theory of Employment, Interest and Money of
1936 and the subsequent development of what would come to be called “Keynesian”
economics, economists developed a new way of analyzing aggregate dynamics that
understood crisis and instability as endogenous to market-based economies. The
managed capitalism of the quarter century following the war put this “new” economics in
the disciplinary and political driver’s seat.
Second, in the late 1960s and early 1970s, this Keynesian consensus collapsed as the
global economy stalled in ways it not only suggested was unlikely, but that it could not
understand. The most glaring, and most famous, example of this failure is the standard
Keynesian model of the relationship between inflation and unemployment—the so-
called Phillips curve— which suggested the two were inversely related: rising prices
reduce unemployment by stimulating production and vice versa. But by the early 1970s,
both inflation and unemployment were rising in tandem, and attempts to slow one only
seemed to exacerbate the other. By the Volcker coup of 1979–80, the “Keynesian
revolution” had been put down, with most of the credit claimed by the so-called
monetarist counter-revolution.
At least at a broad scale, these revolutions look a lot like the scientific revolutions
Thomas Kuhn described in the early 1960s. Kuhn argued that science advances in a kind
of punctuated equilibrium: most of the time, a field knowledge is characterized by a
“paradigm” through which the main principles and theories are widely accepted.
Scientific labour in these periods consists in the “normal” science of filling in the details
or working out the specifics of this or that dimension of the system. Occasionally,
however, the paradigm is undermined, and a “revolution” shakes the overall framework,
during which a whole new way of knowing or explaining emerges, and the old is either
left behind or shrinks from a general theory to a more provincial understanding of the
world. The revolution is followed by a period in which the new paradigm’s “normal”
2/12
dominates, and the pattern continues. In physics, for example, these revolutions are
relatively well-known—those associated with Copernicus, Newton, Einstein, etc.—and
each time, according to Kuhn, physics had to work itself out anew.
The fundamental claim of these rational expectation revolutionaries was not merely that
we need to understand the role of agents’ expectations in shaping economic activity.
That had been obvious to economists for decades. Much of Keynes’s General Theory is
concerned with expectations, and many of his best-recognized contributions are
essentially insights into the role perceptions of the future play in current economic
decision-making. Effective demand, for instance, among Keynes’ most well-known
conceptual innovations, is purely expectational: it is the volume of current production
determined in light of businesses’ expected sales. Keynes’ point was that no business will
purposefully produce more than its best guess at what it will sell: the aggregate estimate
of future demand is the “effective” level for which firms produce. By the late 1960s, this
was taken for granted even by non-Keynesians (it is part of what Friedman meant when
he said “in some ways, we are all Keynesians now”).
The revolutionaries’ point was not only that actors develop expectations which
determine their behaviour based on past experience, but that those expectations are
rational in the economic sense: self-interested, anticipatory, and adaptive. In other
words, in the interests of profit-maximization, businesses will not only try to estimate
future market conditions, they will also anticipate policy changes that might affect their
3/12
own and consumer decisions, and will get better and better at this as they learn over
time. These were the “microfoundations” appropriate to a macroeconomics of capitalist
individualism. This might not sound like much of a discovery, but its proponents insisted
it meant (in the words of Sargent) “that no less than the field of macroeconomics must be
reconstructed in order to take account of this principle of human behavior.” 1 In
particular, it had to be reconstructed in non-Keynesian form because, like their teacher
Friedman, they hated Keynesianism for justifying social constraints on the Darwinian
dynamics of “economic freedom.”
It would be difficult to overstate the influence this way of thinking has had on
macroeconomics, macroeconomic policy, and through it on the world over the last forty
years. It sounds like the most minor of tweaks, but Sargent was right. It was definitely
more than a “turn.” In fact, a lot of the basic principles of market-friendly economic policy
since 1980 find their fundamental theoretical justification in the economics of rational
expectations.
Their first and in some ways still most famous achievement was the construction of a
more-than-merely-monetarist account of the failures of 1970s Keynesianism. Building on
Friedmanite foundations, the militia diagnosed it as largely a result of an inadequate
theory of expectations. The so-called “Lucas critique” demonstrated that the driving
force behind the inflation-unemployment spiral of the early 1970s was that as the rules
of the game changed—as the government tried to stimulate the economy by spending—
market participants recognized it and changed their behaviour accordingly. They
stopped being “fooled” by the stimulus and started expecting inflation, pricing
anticipated increases into their plans. In the process, expansionary policy and
expectations made inflation a self-fulfilling prophecy: people expected a certain level of
inflation, priced it in to the market, and consequently realized their own expectations.
The government, in turn, could only stimulate by exceeding expectations, which required
further and further expansionary policy, which changed expectations, only driving
inflation higher. As economists would say, the long-run Phillips curve is vertical: you can
push prices up all you want but, in the end, unemployment will be the same. Policy
cannot actually do more than pull off short-term “tricks”—beyond that, rational
expectations neutralize it.
Rational expectations is also the foundation of the “efficient markets hypothesis,” which
claimed that asset prices reflect all available information: rational market participants’
taking this information into account means that, adjusting for risk, prices reflect true
asset values (they are “efficient”). The policy lessons of these two insights are enormous:
(a) that in anything but the short term, expansionary fiscal and monetary policy is
useless at best, and likely bad; and (b) financial markets efficiently allocate risk across
market participants, and there is no such thing as a “bubble” (or at least no bubble that
requires regulatory attention).
4/12
Since the “revolution,” the competition for macroeconomic theoretical hegemony has
involved a series of pretenders to the throne—new classical economics, real business
cycle theory, and most recently, “New Keynesian” macro—but all eagerly adopt rational
expectations assumptions. These are now so central to the epistemological arsenal of,
say, central banks, that to formulate a modern macro model without them would be
unthinkable. For each of these variations on contemporary macroeconomics, regardless
of its flavour, the market is the priority, precisely because it is the realm in which
economic rationality is optimally expressed. It doesn’t matter if that market is
characterized by perfect competition and information, as in real business cycle theory, or
if it is subject to sticky prices and missing markets, as it is with New Keynesianism. In all
cases, the market, and the market alone, is the medium through which stability is
produced by adaptive, self-interested expectations in combination with clear and
predictable rules of the game. So-called “credible commitment” to these rules is the
government’s one inviolable responsibility. This is an economics for the production and
maintenance of a market-based “normal,” in which prices are true, expectations are
“anchored,” policy is “time-consistent” and institutions operate according to strict and
universally recognized decision rules.
A dozen years on, however, the “new” economics has still not arrived. In most of the
economies of the global capitalist core there are new, mostly quite mild, regulations
intended to soften future blows; central banks have developed a few new tricks of
varying effectiveness, and the state’s backstop function is better institutionally
5/12
embedded, if only in weak form. Policymakers’ models now try to account for financial
“frictions”. But the main tenets of macroeconomics have not changed since the crisis,
and most of them are constrained by the rational expectations straightjacket because it
is still a key ingredient in the cookbook used to make sense of the world, despite its brief
but stern dressing down.
This extraordinary influence operates especially through a particular way of dealing with
uncertainty, an issue that Keynes did more than anyone to put at the centre of
macroeconomics. In fact, one of rational expectations economics’ greatest achievements
was to tackle uncertainty—something earlier classical and neoclassical economics had
for the most part ignored. Part of the power of Keynesian political economy in the post-
war world was due to the fact that in the wake of the Depression and war, it was the
economic thinking that took the glaring vagaries of history seriously. More orthodox
efforts to reclaim the throne of macroeconomic common sense were continually
compromised by the fact that post-war anti-Keynesians had basically no compelling way
to deal with uncertainty on the terrain of policy. On this front, the rational expectations
militia came to the rescue, and have held the territory ever since.
6/12
It is one thing to say this about a rock dropped on the surface of a pond. It is quite
another to suggest that the whole economy has a “natural” tendency toward general
equilibrium. If this were indeed the case, then one might be tempted to describe the
various possible “states of the world” with a probability distribution, any of which is
understood as less likely, and less stable, the greater its distance from the central mass
of the distribution. Consequently, it is possible to take uncertainty regarding future states
of the world as accurately represented by probabilistic measures like “risk,” since states
of the world can be assigned a measure of likelihood over time. Such measures assure
us that, while the outcome of existing developments is uncertain, we can paradoxically
be quite certain just how likely any particular outcome is. This allows us to operate with
some confidence in our expectations about future states of the world, based on their
relative probable impact on “normal” economic activity.
From a rational expectations perspective, these models are useful tools for several
reasons. First, they appear to embrace uncertainty. Not only is the future not assumed to
be perfectly known, as in some earlier non-Keynesian models, but the whole apparatus
is assembled to take account of the effect of “shocks” to the overall system. 2 Second,
because economic agents are assumed to have rational expectations, they are in effect
assumed to know how the model works. In other words, they know how things work,
they know the rules of the game, and they know, or at least can learn, how to make
decisions if and when those rules change due to stochastic events like weather-induced
“supply shocks” or trade-war related “demand shocks.”
All the while, though, both the overall economic system and market participants’ own
rationality are pulling things toward “normal”: if people expect things correctly, then they
will learn that in general they are correct, and in acting rationally they will produce the
outcome they expected. Rational expectations solve the problem of history by erasing it,
because in the end, they are built on the premise that over time, what is expected will
turn out to be true. Which is to say that on Kuhn’s terms, the rational expectations
revolution’s “normal science” is the science of a free-market normal, and its models are
only tractable if we assume that come what may, the train remains on the tracks and the
engine is always pulling us forward.
7/12
This is the conceptual possibility frontier of contemporary macroeconomics, and in an
era in which “normal” is an increasingly spurious category, it is an economics that is no
help at all. It suggests that given current dynamics, possible futures can be assigned a
probability in a system that tends toward stability, at least in the long-run. Consequently,
it cannot anticipate or even conceive of what we might call catastrophic outcomes, which
involve the disintegration or suspension of the system itself. There is no point inside the
probability distribution indicating the likelihood that the probability distribution itself will
become unmoored from the world it signifies, no point within the distribution that can
represent an outcome beyond it. The best it can offer is to categorize calamity as a “fat-
tail” exogenous shock—this is cutting edge climate economics—but even that is absurd,
since what it tells is basically something like where x is the present and y is the future, y =
ax ± who knows? 3 Anyone paying attention already knew that.
When, in the wreckage of the financial crisis of 2008 onward, even the Queen of England
wondered out loud why the top economists in the world never saw it coming, the
rational expectations economists were outraged. If their models were side-swiped and
blown to pieces, that had nothing to do with the models themselves. The whole point of
an unpredictable event, they fumed, was that it was unpredictable. No model of the
economy could have told us what was coming. What they could not shout down,
however, was the fact that rational expectations macroeconomics and its DSGE models
could not only not predict the crisis, it actually insisted it was virtually impossible, or so
unlikely that we needn’t worry about it. As Robert Lucas put it in his 2003 address to the
American Economic Association: “the central problem of depression prevention has been
solved, for all practical purposes, and has in fact been solved for many decades.” 4 If this
is uncertainty, it is of a remarkably certain kind.
There is nothing in any of this, however sensitive to a limited range of external shocks,
that can handle crisis, let alone a crisis endogenous to the system. All it could ever do,
and all it ever did, was optimize capitalist “normal” for capital. This is why even so non-
radical a policymaker as Jean Claude Trichet, President of the European Central Bank
from 2003-2011, remarked that “in the face of the crisis, we felt abandoned by
conventional tools.” Rational expectations theory and DSGE were foremost among these
tools.
The difference between 2008 and today, however, is that it is possible that when the
virus has run its deadly course, economics and economic policy will not be able to go
back to the last four decades’ devastating “normal.” The virus might—I want to say
should—mean the death of modern macroeconomic orthodoxy, and of the policy regime
it has underwritten. Perhaps it might even mean the end of equilibrium, at least in an
epistemological sense. When climate change jumps back out of the political shadows
when this emergency has receded, how can equilibrium, let alone general equilibrium,
survive?
Economics’ defense when confronted with something like the pandemic is of course
partly understandable. It is true that it would have been basically impossible to “predict,”
if that means tell us it’s coming and when. It is equally true, consequently, that much if
not most of modern economics, as a way of understanding and anticipating the
unfolding of history, has little to offer right now. The best it can do is provide some basic
statistical descriptions of patterns that are currently unfolding: the flight of capital from
so-called emerging markets, soaring unemployment rates, or vague estimates of GDP
losses over the next six months or a year.
But if the two most recent supposedly once-in-lifetime global crises have taught us one
thing, it is that modern economics is built for an increasingly illusory and brutally
unequal normal, in which uncertainty and expectations are bounded by “reasonable”
limits, for a world in which a well-managed (market-based) normal chugs along,
occasionally buffeted by “shocks” that are always framed as exogenous. How useful is
this in a world in which the line between normal and crisis has broken down, and
natural, social and geological processes are so clearly interdependent? Beyond the
simple inside/outside rigidities of modeling, it makes no sense to say the coronavirus
pandemic is “exogenous,” when it is clearly a product of the elaborate and unequal
network that constitutes global political economy. As people like Mike Davis have
recently remarked, the pandemic is better understood as a product of the political
economy and ecology of global capitalism: marginalized workers and consumers in the
global South were necessarily the ones likely to initiate transmission, because they are
the ones who still do the risky low-wage work close to nature. The instantaneous and
unregulated networks that constitute global exchange penetrate the entire hierarchy,
and make transmission virtually unstoppable. And forty years of liberalization,
privatization, and the abandonment of public institutions of collective welfare meant that
when it arrived, there was very little left to mobilize.
The result is bad for everyone except Amazon profiteers, obviously, but as Keeanga-
Yamahtta Taylor and many others have pointed out, it is particularly vicious for those the
system has long done its best to crush. As David Harvey puts it, “the progress of COVID-
9/12
19 exhibits all the characteristics of a class, gendered, and racialized pandemic”—and as
Indigenous peoples the world over struggle to manage the coronavirus contagion on top
of everything else, we must of course add “colonial” to this list.
For a variety of reasons both logical and mythological, this is pretty much how it worked
out after the financial crisis. As Adam Tooze and many others have said, however, this
time there can be no going back to normal: “If radical uncertainty was a concern before, it
will now be an ever present reality.” But it is not just true uncertainty that will hinder a
return to an economic normal. The pandemic has destabilized not just those dimensions
of the world we conventionally think of as subject to economic policy. As the days pass,
the breach is increasingly experienced as a vacuum, and that vacuum will be filled. On
the exciting side, movements are organizing to reinvigorate the collective project, and
prevent the reimposition of market discipline and unequal access. But we are already
hearing rumblings of the belt-tightening that will be required when the worst has passed,
the bleating of inflationary hawks who some people still seem to listen to. Other forces
are eagerly seizing the opportunity to extend the surveillance apparatus, build even
more inaccessible islands of privilege, lock-in disastrous fossil fuel infrastructure, and
villainize immigrants and others to inscribe a mythical “purity” on the national body.
Normal won’t return because it is already passed, another victim of the pandemic.
What comes next, however, is by no means clear, and by no means necessarily welcome.
Just because the last “revolution” in economics has turned out to be so catastrophic and
ill-suited to the unequal, ecologically-degraded and undemocratic world it gleefully
helped produce and angrily defends, that in no way means that what’s coming must be
better. History suggests that a scientific “revolution” can leave some of the most
pernicious aspects of its common sense untouched.
Much of the challenge on the economic front right now lies in how to transform that
common sense. For instance, one of the things COVID-19 has made irrefutable is that at
least in the realm of life’s necessities, the market is a totally inadequate instrument of
distribution. But without considerable work at both the institutional and political-
ideological levels, that lesson will not stick, but will instead fit easily into the narrative of a
normal temporarily undone by an exogenous “shock.” Contemporary macroeconomists
would say that this is a “market failure”—the failure of markets to play their efficient
10/12
distributional role—but it is actually a failure of markets, and a direct result of the work
of more than forty years of economic policy. When you believe no one else will support
you, and when markets are not just the central, but the only resource allocation
mechanism in societies, panic-buying and hoarding make perfect sense.
If this experience gets logged into the historical record book as mere “shock,” then next
time, one of the first things many people will have learned from the pandemic is that
because the market is so cruel, the smart response is to hoard earlier and faster. They
are a product of, not an aberration from, market-dependent life in terrible times. What’s
worse, the point of contemporary markets in those times is not to prevent panic, but to
fan the flames. In our just-in-time world there is only ever enough medicine, masks, life
support, or toilet paper because there is never supposed to be enough. The point is only
to plan for a profit-maximizing normal. Firms don’t respond to panics to bravely or
altruistically restock the shelves for the old lady who went with nothing, but to feed the
panicked mouth. If the goal was to get necessities to those who need them, no one
would be burying tons of “surplus” foodstuffs because restaurants are no longer buying
leeks and onions.
Some of the work to move common sense, to show how “endogenous” the lived reality
of the pandemic is, is already underway. Most of it, of course, is happening outside the
realms of economics proper. Neither the Defence Production Act, nor international swap
lines between central banks for dollar funding, nor bottomless quantitative easing
undermine that common sense; on the contrary, these measures buttress it by framing
the entire rescue operation as an exception that will eventually fade away against a
background of normal. Where the dying economic common sense is losing grip is
outside the categories that economics uses to think about the economy, the result of
work being done closer to the ground. Mutual aid groups have sprung up all over the
world, helping those in need of support. While it might have been obvious to the people
who do the work, many of us are recognizing for the first time in our lives how crucial
supposedly “low-skill” labour is not only to the functioning of the economy, but to our
everyday well-being, and how surprisingly expendable much so-called “skilled” and
professional work turns out to be. The taken-for-granted hierarchies are unsettled, and
ripe for a toppling.
Among the more complex dimensions of the challenge to more radical transformation at
the moment, however, is that it seems to me to require using the institutional, political
and conceptual apparatus at hand, because too many folks are in desperate and
immediate need, and for many of them it is access to this apparatus that appears to
provide their best chance of managing the turmoil. Normal might have been exploitative,
even punishing, but it was also somewhat familiar, and was for lots of people the only
means, however inadequate, through which to confront necessity. A qualitatively
meaningful and just transformation is entirely possible, but in the near-term it must
manage demands on which the capitalist everyday of jobs, housing, education and
health is supposed to deliver, even if it never did for so many.
11/12
This will not be easy to organize and achieve, at least partly because of the hold that
vague thing policy called Keynesianism currently has on the way capitalist societies think
about emergency. The last decade has shown that something like what most people
think of a Keynesianism—government stimulus, rescue funds, low interest rates and
occasionally income or unemployment relief—is the de facto shared panic button of
capitalism’s greatest admirers and some of its staunchest critics. The specifics matter
enormously, of course, but crowded together at the edge of the precipice, it is the one
thing on which a general agreement has emerged. Yet however “revolutionary” its
economics, Keynesianism has rarely been about radical or even significant social change.
It has virtually always been a rescue operation to salvage a civilization of a particularly
bourgeois flavor, an elaborate legitimization program. In Keynes’s own mind, it held open
the door to a far-away post-scarcity economy, but even if we take his rosy projections of
the “economic possibilities for our grandchildren” at face value, that vision is miles from
today’s Keynesianism, and especially from its current New Keynesian macroeconomic
version.
All of which is to say that if we hear about the return of Keynesianism, we should not
relax one bit. It does not mean everything will be fine, and it definitely does not mean
that any of its supportive measures, however discriminatory and inadequate, will last.
Instead, the “new” economy cannot be merely Keynesian, because it has to be a rescue
operation for those who had been thrown off the boat long before the pandemic. The
new economics cannot dream of some Pareto-optimal normal, which has never been
anything more than an arrogant just-so story. Even when we have escaped the clutches
of the virus, we will remain beholden to a concatenation of permanent emergencies for
which most of contemporary economics has already proven itself almost totally useless.
The new economics will have to be more wary, more place and time specific, more
deliberate, and probably a lot slower and more “costly.”
The question, of course, is who bears those costs. I can guarantee you that if the
economics of the last forty years retains its dominance in policy and beyond, it will be the
same people who have been bearing them for so long: the poor, the marginalized, and
the unwelcome. The economics that must replace it won’t be a fully-formed “school,” and
it won’t be “rational” in the narrow disciplinary sense, but an experimental, adaptable
and bold patchwork. If we look around, we can already see people planting its seeds.
The task ahead is to water them, and make sure they get lots of light.
12/12