Market Failures

One of the best-kept secrets in economics is that there is no theoretical proof that markets reach equilibrium. The brightest minds in the field spent a century trying. Arrow and Debreu proved in 1954 that some set of equilibrium prices exists — but nobody ever showed that any mechanism would actually get you there. In the 1970s, a series of seminal papers, one authored by Debreu himself, eliminated what Franklin Fisher calls "any last forlorn hope" of proving market stability.¹

An engineering analogy helps. The invisible hand metaphor imagines economies as passenger planes — buffeted by turbulence, they settle back into a slightly different path. General equilibrium theory suggests they're more like fighter jets, which would spin out of control without fly-by-wire guidance systems continually redirecting them. The word "equilibrium" is deeply misleading; it describes a hope, not a mechanism.

This isn't just academic hair-splitting. The Federal Reserve's pre-2008 models literally assumed markets are always in instantaneous equilibrium. How could they predict a crisis that was, by assumption, impossible? After the crisis, the Fed dropped those models and responded with brute force — which is what you do when your fighter jet's guidance system fails.

The Taxonomy of Failure

Scott Alexander's Non-Libertarian FAQ provides one of the clearest catalogs of how markets actually fail, organized around four mechanisms: externalities, coordination problems, irrational choices, and information failures.²

Externalities are the most familiar. I sell my house to a wasp farmer; our trade benefits us both, but my neighbors didn't consent to the wasps. A factory spews carcinogenic chemicals; the widget buyer and the factory both profit, but the people breathing the air weren't consulted. The libertarian solution — voluntary contracts to ban wasp farming — runs into a practical wall: you'd need every single person in town to sign, and a single holdout destroys it. The proposed fix — a neighborhood association that collects dues and enforces rules — is just the reinvention of government.

Coordination problems are subtler. Imagine a thousand fish farms on a lake, each producing waste that lowers everyone's productivity. Each farm could install a $300 filter. But if you install yours and your neighbor doesn't, you've paid $300 and the lake is still polluted. The rational individual choice is to free-ride. The collectively rational outcome (everyone installs filters) requires coordination that the market doesn't naturally provide.

These aren't edge cases. They're the structure of many of the most important problems we face — climate change, antibiotic resistance, arms races, housing supply. The mathematical structure is always the same: what's individually rational produces collectively insane outcomes, and no amount of voluntary trading solves it.

Information failures deserve special attention because they explain some of the most puzzling market outcomes. George Akerlof's "market for lemons" showed that when buyers can't distinguish quality, only low-quality goods survive in the market. The art market is an extreme case: value is subjective, intrinsic worth is just paint and canvas, and the signaling process that determines what's "good" is controlled by a handful of galleries who manipulate prices in ways that would be illegal in any other industry.³ Galleries keep sales prices secret, choose who they'll sell to, and will bid on their own artists' work at auction to control the market price. An entire young collector refused millions for a painting because reselling without the gallery's permission would have blackballed her from the industry. The system is a cartel with a conscience — or at least, a cartel that believes it has one.

The Monopoly Problem

The market failure that's gotten worse fastest is concentration. Platform Monopolies covers this in detail — the short version is that Google, Facebook, and Amazon have achieved market positions that classical economics says should attract competition but don't, because network effects and data advantages create self-reinforcing moats that no amount of venture capital can breach. The antitrust framework built for Standard Oil can't handle platforms that price below cost to capture markets, or whose products are "free" to consumers while the real costs are paid in data extraction and the hollowing out of the creative middle class. Lina Khan's Amazon's Antitrust Paradox laid out why the consumer welfare standard breaks down; the deeper question is whether the platforms are competing within markets or have become the markets themselves.⁴⁵

Superstar Firms and the Labor Share

The concentration problem connects to one of the most important macroeconomic trends of the past forty years: the declining share of income going to workers. It was long Keynesian dogma that two-thirds of economic output goes to labor. This has come undone. Corporate profit — which in a perfectly competitive economy would be zero — has gone from just over 2% of GDP in the early 1980s to more than 15% today.⁵

The "superstar firms" hypothesis says this is driven by a shift toward winner-take-most markets. A few hyper-productive companies capture most of the surplus in their industry. They don't need proportionally large workforces, so the labor share within each firm stays constant, but the industry-level labor share falls as the superstar captures more revenue with relatively fewer workers.

This matters because it means the standard policy responses — education, retraining, making workers more "competitive" — are aimed at the wrong target. The problem isn't that workers are failing to be productive enough. It's that the structure of markets increasingly allows a few firms to capture the gains from productivity while distributing less to the people who do the work. It's a structural problem, not an effort problem.

Kenneth Arrow, perhaps the greatest economist of the 20th century, put it bluntly: "In a system where virtually all resources are available for a price, economic power can be translated into political power by channels too obvious for mention. In a capitalist society, economic power is very unequally distributed, and hence democratic government is inevitably something of a sham."⁵

Unregulated Markets as Natural Experiments

Cryptocurrency markets offer the clearest natural experiment in what happens when market failure mechanisms operate without regulation. A detailed anatomy of the July 2021 Bitcoin short squeeze reveals a playbook that combines media manipulation with order-book exploitation. First, a planted story — an unverified "insider" claim that Amazon would accept Bitcoin payments — was traded up the chain from a low-tier publication to mainstream outlets. Then, in the minutes before the liquidation event, spoofed orders were placed and rapidly cancelled to trigger momentum ignition, followed by the simultaneous shutdown of market-making bots to create the thinnest possible liquidity environment. The result: cascading forced liquidations of leveraged traders, with hundreds of millions in losses.⁶

The forensics are damning. Blockchain analysis traced $290 million in USDT transfers to Binance in the hours before the event, most likely from Alameda Research (later notorious for the FTX collapse). The manipulation wasn't illegal — cryptocurrency markets are, as the author noted, "a lawless wasteland." But they're also a demonstration of what information failures and market power produce in the absence of institutional guardrails. The spoofed orders, the planted media stories, the coordinated timing — these are the same mechanisms that operate in regulated markets, just with the regulatory friction removed. The SEC's repeated citation of price manipulation as the reason for rejecting Bitcoin ETF applications isn't paranoia. It's pattern recognition.

What Equilibrium Actually Means

I keep coming back to the invisible hand result, or rather its absence. Most working economists treat market stability as obvious — Milton Friedman once said he saw no point in studying general equilibrium stability because the economy is "obviously stable," and if it isn't, "we are all wasting our time." Franklin Fisher quipped that the bit about economists wasting their time was perceptive.¹

But the fact that economies muddle along most of the time doesn't mean they're self-correcting in the way the invisible hand metaphor implies. It might mean that something outside markets — social norms, economic regulation, central banks in their happier moments — usually averts disaster. The invisible hand sees an aerodynamically stable plane. The theory sees a fighter jet held in the air by constant computational intervention. The difference doesn't matter much in calm weather, but it matters enormously when turbulence hits.

The practical upshot isn't "markets are bad" or "government is good." It's that the question "should this be left to the market?" has no general answer. You have to look at the specific case: the externalities, the coordination structure, the information asymmetries, the potential for concentration. Sometimes markets work brilliantly. Sometimes they produce the 2008 financial crisis, or RealPage's algorithmic rent cartel, or a world where YouTube makes $9 billion in revenue while musicians earn more from vinyl than streaming.

The failure to prove general equilibrium isn't a curiosity. It's the most important negative result in economics, and most economists have never reckoned with it.