7 Powerful Ways Business Cycles Expose the Limits of Rational Expectations

Last fall, a senior economist at a major US bank told clients to brace for a hard landing. October data arrived softer than expected, unemployment inched upward, and the financial press anointed him a prophet. Then fourth-quarter GDP printed at 3.1%. The same forecast suddenly looked like a haircut on a bald man. Does that make the economist a fool? Not at all. It makes him a human being trying to model an economy built by other human beings.

This little drama plays out every single quarter, yet it cuts to something deeper. Macroeconomics has spent the last fifty years building elegant castles on a concept called rational expectations. The theory says people form forecasts using every relevant piece of information, make no systematic mistakes, and therefore cannot be fooled by predictable policy. It’s a beautiful assumption, logically clean, and mathematically satisfying. The problem? Business cycles refuse to read the memo.

Let’s unpack why. If you spend any time with modern macro models, you’re meeting agent who resembles a supercomputer, but actual economies still lurch from boom to bust, panic to euphoria, with alarming regularity. Here’s the kicker: the very existence of that rhythm tells us rational expectations captures only part of the story.

This article walks through seven powerful ways business cycles break the rational-expectations fairytale. You won’t find a call to abandon the theory here, not entirely. But if you forecast, trade, or set policy, you need to know where the machinery stops working. What follows may change the way you read every jobs report that ever crosses your screen.

The short version of the argument, if you only remember one thing from the article:

• Rational expectations assumes free information, but data is expensive and slow
• Recessions exist even though perfectly rational agents shouldn’t let them happen
• Herding turns micro-level sense into macro-level nonsense
• Self-fulfilling beliefs can shift GDP without any change in fundamentals
• One rational agent, one forecast; millions of agents means messy reality
• Politicians playing games inject distortions no clean model can defend
• Real humans learn by trial and error, not by instant revelation

Read on. The first point alone will tell you why your favorite forecast model always seems one recession late.

1. Information Is Not Free, Coordination Is Not Automatic

Rational expectations theory starts with the premise that people use all relevant information available to them. Sounds reasonable, right? Until you remember that information costs money, time, and cognitive energy to acquire. The clean version of the theory assumes agents behave as if they already know the model generating the economy, right down to its parameters. The real version of you does not have that model. You read headlines, scroll through LinkedIn posts, and check your 401(k) quarterly.

This gap between textbook information and real-world information produces systematic forecasting errors, which produce waves in economic activity. Businesses invest based on what they think demand will do next year. But because demand data releases with a lag, your “rational” forecast today relies on data from three months ago. By the time you adjust, the cycle has already turned. Everyone adjusts together, and that synchronized response amplifies the swing.

The rational expectations hypothesis looks elegant on a whiteboard, but real markets suffer from sticky expectations. Firms pricing a new product usually check their own costs, last quarter’s sales, and what their biggest competitor is charging. That’s rational, but locally, not globally. Throw in the fact that business owners are also reading election polls and worrying about shipping rates from the Red Sea, and you’ve got a forecast cocktail that’s one part data, two parts vibes.

Perception lives in the rearview mirror. When you treat information as cheap, you treat cycles as avoidable. When you admit it’s costly, the path becomes clearer: recessions happen because humans legitimately don’t know what’s coming.

2. The Existence of Recessions Is Embarrassing

Here is a thought experiment that keeps macroeconomists humble. If every agent in the economy holds rational expectations, then everyone knows with certainty that a depression can’t happen without a genuine shock to fundamentals. People keep spending, firms keep hiring, and a dip in sentiment never snowballs into a genuine contraction, because no one believes in self-fulfilling panics.

Recessions exist anyway. Historians count no fewer than 50 recessions in the United States since 1775, and dozens abroad since 1990 alone. You’d be forgiven for thinking the world economy never got the memo about rationality.

The hard part is reconciling that observable cyclicality with the “policy ineffectiveness proposition,” the most famous implication of rational expectations. It says anticipated monetary policy changes shift only prices, never real output. But when the Federal Reserve signals a tightening cycle and construction slows a quarter later, it doesn’t feel like a price-level effect. It feels like lost jobs.

Rational expectations theory can explain surprise-driven recessions, but it has a hard time explaining long, grinding recoveries. Consider the 2010–2016 period in Europe. European Central Bank policy was widely telegraphed; every tightening and easing was anticipated. Yet output in Italy and Greece spent a full decade below trend. Anticipated policy shouldn’t matter under the theory. Real people say otherwise.

Truth be told, unemployment stubbornly high, output gaps steadily negative, these are anomalies that rational expectations simply explains away with clever language. Don’t let the clever language fool you.

3. Herding Turns Smart People Into a Dumb Crowd

One of the most uncomfortable facts about financial markets is that information cascades happen. Your average fund manager does not make independent forecasts. They watch what other managers are buying. They track the consensus. They benchmark their portfolio against the S&P 500, not against their honest view of the economy.

That produces herding, and herding produces bubbles. When a tech stock doubles, fund managers who own it look like geniuses. Managers who don’t own it face client withdrawals. So even the most rationally grounded investor joins the crowd to survive. Herd behavior drove housing prices to absurd levels in 2006. It drove dot-com valuations into fantasy in 1999. And it produced a banking system where every single lender made the same dumb subprime bet.

Business cycles don’t come from rational agents in equilibrium. They come from rational agents copying one another until the collective position becomes insane. At some point, one hedge fund quietly dumps its position. Nobody notices for a month. Then two more funds sell, and suddenly the whole market catches the same cold. Boom turns to bust. Rationality never had a chance.

You will sometimes hear an industry joke that sums this up: an economist sees a $20 bill lying on the sidewalk and walks right past it. When asked why he didn’t pick it up, he replies, “If it were really there, someone would have taken it already.” A beat later, the business cycle runs it over. That’s the problem with equilibrium thinking, it can only explain away the $20 bill; it cannot explain why everyone crowds around the same door on the way out.

4. Self-Fulfilling Prophecies and the Sunspot Problem

Here’s a phrase that makes even seasoned economists twitch: sunspots.

In macro theory, a sunspot is a random variable that has no real effect on productivity, trade, or preferences, yet still moves the economy. How can that happen in a world of rational agents? Simple. People expect a downturn, so they cancel discretionary purchases. Firms see falling orders and shelve investment plans. Banks tighten credit standards because borrowers look shaky. Then the downturn shows up.

Nothing fundamental changed. The only change was collective belief.

Rational expectations does not smooth over this problem. In fact, the theory can actually help create it. If everyone believes an announcement of higher interest rates signals a recession, then the announcement itself becomes the cause. Central bankers understand this intuitively. That’s why the Federal Reserve spends enormous energy communicating its “dots” and forward guidance documents. They are trying to manage expectations explicitly. They are speaking to millions of irrational, nervous, overstimulated citizens and hoping nobody bolts for the door.

Modern DSGE models have tried to build unique equilibria around this issue, but they resort to straitjacketing expectations with assumptions. Remove a technical condition and you end up with what economists call multiple equilibria. The economy can leap between worlds for no observable reason.

As an aside, this is where John Maynard Keynes’ “animal spirits” stopped being avant-garde and started looking suspiciously practical. Business cycles driven by mood swings are cycles driven by non-fundamental news. Irrational? Undeniably. Powerful? Catastrophically. That reality is precisely why credible central bank communication requires a central bank to say what it means, and mean what it says.

5. Heterogeneous Agents Make Equilibrium a Mirage

One of the sharper critiques of rational expectations models is that they usually feature one representative agent. The entire economy collapses into a single household that consumes, invests, and works. That agent knows the model. But a single-agent model hides all the interesting problems of coordination and disagreement.

The real economy contains young spenders and older savers, risk-tolerant entrepreneurs and risk-averse retirees, borrowers drowning in debt and lenders holding all the cards. They never agree on where inflation is heading or what the Fed will do next. Their differences generate trades, and their trades generate cycles.

When the Fed raises rates, some households interpret that as a mistake and keep spending. Others view it as a signal of tighter conditions and immediately slam the brakes. Consumer credit behavior splits in half. Mortgage originations drop while auto loans keep flowing. Aggregating those wildly different responses into one “rational” household hides crucial imbalance. Supply chokes precisely because firms misread one group’s behavior as the whole economy’s direction.

The deeper point is that heterogeneous expectations create an ecology of forecasters, and in an ecology, booms arrive when bulls outnumber bears. Rational expectations theory assumes everyone has the same information and same model, but the very persistence of trading volume proves people disagree. You’ll never see an entire economy inhabit a single equilibrium at once, no matter what your textbook says.

6. The Political Business Cycle Refuses to Die

Macroeconomists have a phrase for the intersection of elections and the economy: the political business cycle. Governments love to juice the economy right before voters head to the polls. They cut taxes, boost spending, pressure central banks into easing. Voters respond with a temporary surge in optimism, consumption rises, and the incumbent gets another term.

A rational expectations economist would say this can’t work. Voters see the manipulation, discount the binge, and reward no one. Politics, like the market, should be efficient so long as voters process information cleanly.

Tell that to nearly every democratic government on the planet.

Research repeatedly shows incumbent governments engineer favorable economic conditions in election years. Spending increases skew toward public-sector wages and visible construction projects. Public debt balloons more in election years than at any other point in the electoral calendar. The evidence stretches from Brazil to Japan to the United States.

Why do economists call this irrational? Because voters could theoretically learn to see through the manipulation. In practice, humans respond more strongly to recent wage gains than to abstract arguments about malinvestment. The credibility problem is real, and it powers the cycles rational expectations hoped it would eliminate.

Even the classic remedy for this, making central banks independent, only partially fixes it. Fiscal policy remains firmly in the hands of politicians. The business cycle thus feeds on partisan incentives rather than pure fundamentals. As long as elections exist, expect an artificial kick going into them. No amount of “rationality” washes that away.

7. Real Humans Learn, and Learning Produces Cycles

Rational expectations assumes people know the “true” model of the economy from the start. That’s a heavy lift, and it looks even heavier when viewed through the lens of actual experiments.

Nobel laureate Thomas Sargent spent years documenting how countries learned to conquer inflation. A key insight of this work is known as adaptive learning. Agents don’t begin with the correct model. They form rough estimates, test them against the data, and update their beliefs as reality unfolds. This updating process can itself cause persistent cyclical swings.

Here lies the perfect illustration. In the 1970s, the Federal Reserve believed inflation would settle back down after each oil shock. Those expectations were built on estimation errors. When inflation kept climbing, the central bank kept overestimating slack in the labor market and underestimating underlying price pressure. Only after painful recessions did agents, including the Fed, update their models. The learning curve literally created the business cycle that rationalized the inflation of that decade.

Decades later, economists like Christina Romer and David Romer found that professional forecasters’ errors are predictable and persistent, something you would never observe in a rational-expectations benchmark.

Should we scrap the theory broadly? No. That would also risk losing a lot. Rational expectations remains a powerful normative benchmark, a kind of imaginary world where policy mistakes vanish because everyone sees clearly. But benchmark and reality are different animals. The very process of learning, revising, overcorrecting and undercorrecting, creates the volatility the theory claims to eliminate.

This is also why policy makers love central bank transparency now. They recognize that rational learning is brutally slow and wants to shorten the adjustment period. It doesn’t always work, but it beats silence.

Why This Matters More Than Your Latest Forecast

So what should an investor, entrepreneur, or policy analyst do with all this theory bashing?

Respecting rational expectations means recognizing when you should adjust forecasts toward efficiency. Respecting business cycles means recognizing when you should retain a little humility. A healthy macro view uses rational expectations as your base case and then adds layers of behavioral friction on top. Expect that information takes time, that agents copy their neighbors with enthusiasm, and that institutions, from electoral systems to central bank mandates, carve deep grooves into the cycle.

Forecasters who pursue pure rational expectations ride a finely tuned car directly into a brick wall every recession. They assume agents see the brick wall, stop the car, and avoid impact. Forecasters who accept its limits treat the brick wall as a legitimate feature of the road.

You already know which one drives better in the real world.

Perhaps the strongest conclusion from fifty years of experience is this: humans are prediction machines, but they predict with limited fuel. They extrapolate, herd, panic, recover, and learn. The economy is simply the output of that gigantic, messy, predictive organism. Rational expectations supplies the boundaries of what’s believable; business cycles supply the chaos inside those boundaries.

Smart macro analysts do not ask whether rational expectations is true or false, not in some binary sense. They ask where it serves as a useful fiction. Then they watch for the moments where that fiction slips, because the most powerful investment opportunities of the next decade will come from places where everyone trusts the model and the model fails them.

Question the neat forecast. Question the confident equilibrium. And never, ever walk past a twenty-dollar bill.