The tenth anniversary this past September of the collapse of Lehman Brothers inspired a blizzard of commentary, including some deeply misguided observations. One misunderstanding is that Lehman’s demise caused the Great Recession. In fact, the downturn started months earlier in the US, as NBER’s recession dates show. But there’s room for debate on the question of whether the government’s decision to let economic gravity have its way with Lehman turned what might have been a moderate downturn into the deepest contraction since the Great Depression.

Some pundits argue that letting Lehman fail was prudent and necessary to send a signal for inspiring a more responsible style of banking. Yet a new book on financial crises by a pair of experts on the subject reminds us that applying free-market rules to banks is usually short-sighted – especially during a severe banking crisis.

Another bit of confusion about the last crisis is that it was different because the financial instruments were unique relative to financial history. But ultimately, every financial crisis is about short-term debt. They come in varying forms and are constantly evolving through time, ranging from private banknotes to demand deposits, money market funds and asset-backed commercial paper. But the underlying infrastructure is always the same: short-term debt backed by long-term debt — the basis for bank runs under certain conditions.

In any case, “a financial crisis is a systemic event,” Gary Gorton (a finance professor at Yale) and Ellis Tallman (director of research at the Cleveland Fed) write in their recently published book Fighting Financial Crises: Learning from the Past.

In a banking panic all banks are at risk, and the financial system is about to collapse. For example, during the 2007-8 crises [Fed Chairman] Ben Bernanke in his testimony before the US Financial Crisis Inquiry Commission said that of the thirteen most important financial institutions in the United States, “12 were on the verge of failure within a week or two [after Lehman].”

One popular line of thought is that the Lehman’s ill-advised decisions in the years leading up to its death sealed its fate and so it deserved to fail. Maybe, but Gorton and Tallman outline the case that banking is different than other industries and therefore deserves a different treatment with respect to allowing the free market to rule during crises. (Gorton, by the way, reviewed the historical record on this point in an earlier book: Misunderstanding Financial Crises: Why We Don’t See Them Coming) Because banks and their lending activities are deeply interconnected across the financial industry and throughout the economy, repercussions in any one bank can quickly reverberate with dire results. Accordingly, a bank failure can threaten the entire system, even if the institution “deserves” to fail.

The long arc of financial crises in the US offers a solid foundation for studying the role of banks in these events and drawing lessons. The details of every specific crisis is different, but the broad contour follows a well-worn path, which provides clear lessons. Learning those lessons, however, is difficult if not impossible, or so the last crisis suggests.

One of the challenges is recognizing in real time that a crisis has started. Right up until a tipping point, arguing the counterfactual — here’s the risk if we don’t act to save a bank — is challenging because the critical events may be out of the public’s view. Nonetheless, Gorton and Tallman lay out a recurring narrative that’s familiar to students of banking history.

For the Panic of 2007-8, most observers date the start of the panic as the failure of Lehman Brothers in September 2008, although the start was actually the first or second quarter of 2007. The tumultuous event of the Lehman Brothers failure was the catalyst to galvanize a consensus that the events constituted a financial crisis. Only after Lehman’s failure was it widely agreed that events were dire and that congressional action was required to address them… The failure of a large financial firm is typically an effect of a panic, not the cause, but because the failure of a firm is visible, it often becomes a signal that a panic is under way.

More than a decade after Lehman’s collapse, the government’s decision to stand aside and let economic forces take control remains hotly debated. But as Gorton and Tallman explain, and demonstrate by reviewing the grand sweep of history for US financial crises, letting banks die “might make sense long before a crisis hits but are not appropriate during a crisis.”

Some pundits (and policymakers) insist that moral hazard must be avoided at all costs and at all times for all banks. In turn, these folks argue that Lehman was insolvent and so it deserved or even needed to fail. But that’s a short-sighted argument during a financial crisis. As the authors of Fighting Financial Crises explain, the standard definitions of solvency change when a crisis strikes. Even a bank that held only short-term Treasuries might be effectively insolvent in a crisis. The reason isn’t about the assets; rather,

The value of financial assets can become increasingly dependent on the perceived ability of financial institutions to fund positions. In turn, the funding market access of financial market institutions can be impaired by uncertainty about asset values and firm solvency.

In other words, the value of the assets on the books may not matter if a bank’s ability to sell those assets to raise liquidity are blocked by widespread uncertainty about the institution’s future. “During a crisis it is not possible to declare, with full credibility, that a bank is solvent or insolvent,” the authors explain. “Consequently, letting banks fail during a crisis creates uncertainty about whether other banks will also be allowed to fail.”

Whatever the logic of trying to impose market discipline by letting Lehman fail, the timing of the decision was wrong, and the stakes were huge. It’s no accident that financial system began to melt down following Lehman’s collapse.

Our argument is not that banks not face discipline; our point is that the middle of a crisis is not the time for large or interconnected banks to fail… There may be some clearly insolvent banks, but closing only those banks (without a blanket deposit guarantee) causes runs on the remaining banks. The system begins to unravel.

This is old news, at least for anyone who’s studied banking history. Nonetheless, it was a lesson that was overlooked in September 2008. Once again, the lessons seems to be that financial wisdom is cyclical, not cumulative.

But hope springs eternal, which leaves us with the ancient question: Will it be different the next time? Gorton and Tallman think not, predicting that “financial crises are here to stay.”