DEEP READING · READ 939
How Hidden Fractures Still Threaten the World Economy · Raghuram G. Rajan · 2010
The crisis of 2008 was not produced by a few hundred greedy bankers. It was produced by deep fault lines in politics and the world economy that had been grinding for decades, and it broke where they met — in the financial sector. The worst of those lines runs like this: when incomes stop rising for the bottom half of a country, fixing schools and skills is slow and thankless while extending credit is fast and painless — so the political system chose credit. "Let them eat credit": debt used to fill a gap that only income could actually fill. The gap did not close. It was simply moved onto a bank's balance sheet.
Raghuram Rajan is a professor of finance at Chicago Booth, was chief economist of the IMF from 2003 to 2006, and served as Governor of the Reserve Bank of India from 2013 to 2016. He is best remembered for one afternoon: at the 2005 Jackson Hole conference, in front of a departing Alan Greenspan and a room full of central bankers, he presented a paper arguing that financial innovation had not made the world safer, and that incentives inside the industry were quietly accumulating tail risk. The room turned on him; Larry Summers called the position "slightly Luddite." This 2010 book is that warning worked out at length, and it won the FT/Goldman Sachs Business Book of the Year award.
Pin the metaphor down first, because the book's whole explanatory structure rests on it. In geology a fault is the seam where two plates of crust meet. The plates grind against each other at a few centimetres a year, storing elastic energy while friction keeps the seam locked. When stress exceeds friction, it releases all at once — the earthquake happens at the fault, but the energy is neither generated nor stored there. It comes from decades of plate motion.
That is Rajan's claim about finance. The financial sector is the seam; it genuinely did rupture, and its own distorted incentives made it more brittle. But explaining 2008 as "finance broke" is like explaining an earthquake by saying the rock was weak. The practical value of the metaphor is that it moves the question up a level: not who did wrong, but what force has been pushing for decades while everyone pushing felt entirely justified at every step. The three fault lines:
| Fault line | Where the pressure comes from | How it reaches finance |
| Domestic politics | widening income gaps, stagnant wages, thin safety net | government reaches for credit expansion as an analgesic, above all housing credit |
| Global imbalances | export-led economies suppress domestic consumption, pile up surpluses | global capital floods into the US, depressing rates and fattening credit |
| System mismatch | money from relationship-based systems enters an arm's-length one | two incompatible ways of pricing risk short-circuit; everyone assumes someone else is checking |
All three intersect at the same seam — the American financial system.
This is the sharpest line in the book. It plays on "let them eat cake," but inverts the meaning: this is not aristocratic obliviousness. It is shrewd political arithmetic.
The mechanism runs as follows. Median American wages grew feebly for decades; Rajan's primary explanation is that the supply of skills failed to keep up with technology's demand for them — the education system, especially at the school level, stopped moving enough people into high-skill work, and the gap on either side of a college degree kept widening. Fixing that means fixing schools, community colleges and vocational training: a project with a twenty-year payback and no credit available to whoever starts it. But another road exists. Let people who cannot afford a house borrow enough to buy one. As prices rise they live better and look richer on paper. Nobody has to be taxed, and nobody has to be told out loud that their wages are never going up again.
The policy was concrete: raising the home-ownership rate became a bipartisan achievement to campaign on; the housing agencies Fannie Mae and Freddie Mac were pushed toward ever more aggressive affordable-housing targets; regulators were remarkably relaxed about the expansion of subprime lending. Rajan's point is not that a villain conspired, but that a democracy facing a slow cure and a fast painkiller will predictably reach for the painkiller.
How it changes the way you read things: the next time you see household leverage climbing fast anywhere, do not read it as a financial statistic. An abnormal credit expansion is usually the substitute for a distributional problem nobody wants to name. Asking "what here should have been rising and wasn't?" will usually get you closer to the truth than studying the loan book.
This is the amplifier on the previous fault line, and the most underrated section in the book. A jobless recovery is one where output has resumed growing but employment does not come back. Rajan points to a clear shift in the American pattern: in the early postwar recessions, jobs returned within months of the output trough, whereas after the 1991 and 2001 recessions it took roughly two and nearly three years respectively for employment to regain its pre-recession level.
Why does that matter so much? Because the American safety net is built on the assumption that unemployment is brief: benefits run out quickly and health insurance is attached to a job. Once joblessness lasts two or three years, the net fails to catch people and the political pain becomes unbearable. All of that pressure then lands on the one institution capable of reacting fast: the Federal Reserve.
The result is sustained ultra-low rates, and an expectation built on top of them — the Greenspan put. A put option pays you when an asset falls; the phrase means that investors came to believe the Fed would never deliberately deflate a bubble but would always flood the system after a crash, which amounts to free downside insurance. If the downside is socialised and the upside is yours, leverage is the rational choice. Notice how counter-intuitive the chain is: a country's unemployment insurance being too thin ends up expressing itself as its asset bubbles being too large.
The second plate lies outside America. Germany, Japan, and later much of East Asia including China, all travelled the same road: hold down domestic consumption — through high savings, weak social insurance, managed exchange rates and restrained wages — and aim capacity at foreign markets. During a catch-up phase this works extremely well, concentrating resources in tradable sectors and routing around a small domestic market.
But it has an accounting consequence nobody can escape: current account balances must sum to zero across the world. If someone sells more than they buy, someone else must buy more than they sell. The Asian crisis of 1997–98 magnified this. Emerging markets that had borrowed abroad were brutally punished, and afterwards they collectively changed strategy: stop being net borrowers, accumulate foreign reserves as self-insurance instead. The result was a strange spectacle — poor countries exporting capital to rich ones.
That money needed somewhere to go. Much of it bought US Treasuries and dollar assets, pushing down American long rates — in the one country that, for domestic political reasons, badly wanted cheap credit to keep voters content. The two plates met the seam precisely: one side desperate to borrow and spend, the other desperate to lend. America became the consumer of last resort. Rajan's insistence is that this is neither unilateral American profligacy nor unilateral surplus-country dumping, but two domestic political predicaments feeding each other across an ocean.
The third fault line is the most technical and the easiest to skip, but it explains why the money, once it arrived, blew up the way it did.
An arm's-length financial system — the American and British model — has strangers transacting through public information, standardised contracts, price signals, courts and ratings. Capital can move quickly to new opportunities and abandon failures just as quickly. A relationship-based system — postwar Japan, Germany, most of East Asia — runs on long-term bonds between banks and firms, on inside information and implicit commitments (the Japanese main bank, the German Hausbank). It is very good at mobilising patient capital during catch-up growth, bad at admitting failure, and systematically suppresses both depositor returns and household consumption.
The trouble arrives at the junction. The enormous savings generated by relationship-based systems entered an arm's-length one carried by people without the capabilities that system assumes. They were used to someone watching on their behalf; this system's premise is that everyone watches for themselves. So a role quietly went vacant: foreign central banks and institutions bought American assets trusting the AAA label; the rating agencies' models rested on historical data in which house prices had never fallen nationwide at once; the investment banks structuring the securities needed only for them to sell. Every link in the chain assumed the risk was being watched by the link above or below. No link was watching.
This was the core of the 2005 Jackson Hole paper, and it is Rajan's mechanism for how clever people collectively walk into something stupid.
Tail risk means a very low probability of a catastrophic loss. The crucial property is this: before the tail event arrives, a strategy loaded with tail risk is indistinguishable from genuine skill. Hence the image of picking up nickels in front of a steamroller — you collect money every single day until the day you fail to step aside. The sharper analogy is writing earthquake insurance: you book the premiums as profit year after year, your returns look steady and enviable, and then in one year you pay back twenty years of profits with interest.
Layer the pay structure on top and it becomes lethal. Bonuses settle annually; tail events arrive once a decade or less. The bonus has been banked long before the loss lands on the firm — and on the taxpayer. So the individually optimal move is to take on as much currently invisible risk as possible. Citigroup's Chuck Prince in 2007, in the most quoted sentence of the era (paraphrased) — as long as the music is playing, you have to get up and dance — was not being arrogant. He was describing the incentive structure accurately: if you sit out, you underperform your peers and you are replaced before the music stops.
How it changes the way you read things: any unusually smooth, low-volatility, above-peer return deserves one question — is this alpha, or is it an insurance premium in disguise? And this is not only about finance. Any organisation that judges long-horizon risk by short-horizon metrics — hospitals, airlines, engineering firms, content platforms — is mass-producing the same error.
The prescriptions follow the diagnosis: if the disease is not in finance, reforming finance alone cannot cure it. His order of priority is human capital first — early childhood education, community colleges, affordable retraining — because it is the only thing that genuinely raises wages. Then rebuild the safety net in forms that do not punish work: detach health coverage from the employer, prefer wage insurance to simply extending unemployment benefits. Internationally, push surplus countries toward domestic demand. Within finance, his emphasis falls not on more rules but on fixing incentives: deferred bonuses, clawbacks, and debt that converts to equity in a crisis, which dismantles the implicit insurance policy.
What he leaves out should be said just as plainly. The book pins wage stagnation mainly on an inadequate supply of skills, and gives comparatively little space to the alternatives: the collapse of labour's bargaining power, rising market concentration, and the geographically concentrated damage of trade shocks. That tilts the prescription to the supply side — the people who fell behind are asked to go back to school, rather than the rules of distribution being changed. It is the book's most durable controversy, and the next section takes it seriously.
Compressed to a single line: pressure → political shortcut → capital inflow → distorted incentives → rupture.
Inside America, widening income gaps and a thin safety net produced a politically intolerable kind of pain. The real cure — education and skills — was too slow, so credit expansion was chosen as the painkiller, with housing finance as the delivery mechanism. Meanwhile a group of countries staking growth on exports needed someone to consume their output, and after the Asian crisis they turned collectively into net exporters of capital; that ocean of savings flowed into the United States, pushed rates down and widened the credit channel further. The money then passed through an arm's-length system it did not understand, whose internal pay structures were rewarding precisely those who hid tail risk, with an implicit government guarantee absorbing their downside. Three forces stacked on one seam. The moment house prices stopped rising, the seam opened.
So what the book establishes is a single proposition: a financial crisis is the financial symptom of a political-economic disease. Its method is to run three independently evidenced causal chains — inequality and politics, global imbalances, incentives and tail risk — into the same endpoint. That is the insight, and it is also the exposure: joining them elegantly does not make every segment equally strong.
Misreading 1: "It's a book attacking bankers." The opposite — it explicitly refuses the villain story. Blaming greed reduces a structural problem to a moral one, and the practical result is that you replace some people, leave the incentives untouched, and wait for the next time.
Misreading 2: "Fault lines means flaws in the financial system." That inverts the metaphor. The fault is where the break happens, not where the energy comes from, and the book's entire effort goes into arguing about the plates.
Misreading 3: "He predicted the subprime crisis." Be precise. His 2005 warning was that incentives inside finance were accumulating tail risk, and directionally it was uncannily right — but that is not the same as predicting when or how house prices would collapse. Calling it a prophecy both overstates the paper and undersells what it actually did, which was identify a structure rather than a date.
The criticism that lands:
1. Earthquakes happen at the fault, but the energy comes from decades of plate motion. Wall Street was the fault, not the plate — prosecuting bankers is standing at the crack asking why it cracked.
2. The hardest line in the book is "let them eat credit": wages won't rise, fixing schools takes twenty years and earns no one any credit, while loosening lending works immediately, taxes nobody, and requires telling nobody that their income has stopped growing.
3. So an abnormal credit expansion is usually the substitute for a distributional problem nobody wants to name. When household leverage spikes, don't start with the loan book — start with "what here should have been rising and wasn't?"
4. A chain worth memorising for how counter-intuitive it is: unemployment insurance that is too thin ends up expressing itself as asset bubbles that are too large, because when joblessness lasts years and the net fails, the pressure lands on the only fast-moving institution — the central bank.
5. The Greenspan put: markets came to believe the Fed would never deflate a bubble but would always flood the system afterwards. Socialised downside plus private upside makes leverage the rational choice.
6. The accounting is merciless: if someone sells more than they buy, someone else must buy more than they sell. Export-led success requires a debtor, and after 1998 emerging markets became net capital exporters — poor countries sending money to rich ones.
7. When relationship-based money enters an arm's-length system, the people carrying it expect someone to be watching for them, while the system's whole premise is that everyone watches for themselves — central banks trusted ratings, ratings trusted historical house prices, banks only needed the paper to sell.
8. A strategy stuffed with tail risk looks exactly like skill right up until the tail event — like booking earthquake premiums as profit and never reserving. Bonuses settle annually, catastrophes arrive once a decade, so hiding risk is individually optimal.
9. Hence the question to ask of any unusually smooth excess return: is this alpha, or an insurance premium in disguise? Every organisation that judges long-horizon risk on short-horizon metrics is mass-producing the same error.
10. The line to take, and to discount: a financial crisis is the financial symptom of a political-economic disease. The three fault lines are joined elegantly, but not equally strongly — the role of Fannie and Freddie, and the inequality-to-credit mechanism, remain the two most contested joints in the whole structure.