IN PLAIN TERMSEconomists argue about whether the Depression happened to America or was done to it. Friedman and Schwartz counted the dollars, decade by decade, and concluded it was done to it, by the very institution created to prevent it. Nearly every later claim that the Fed caused the Depression traces back to this book.
Before this book the respectable view of 1929-33 was that monetary policy had been tried and had failed: interest rates were on the floor, the authorities had done what could be done, and the collapse therefore demanded a non-monetary explanation. Friedman and Schwartz rebuilt the money stock from bank records and showed that policy had not been easy at all. Measured by the quantity of money rather than the price of credit, it had been the tightest in the institution's history. The consequence was not confined to history. It moved the argument from whether the state should manage demand to whether a monetary authority can be trusted with discretion, and it installed the working assumption every central bank has operated under since, which is that a collapse in the money stock is the authority's fault and is preventable. The Federal Reserve of 2008 was acting on a reading of this book.
The central claim is not that money and income move together, which nobody disputed, but that the causation runs mainly from money. Correlation cannot establish that, and the authors know it. The weight is carried instead by episodes in which the quantity of money moved for reasons plainly unconnected to the state of American business, and income moved afterwards.
Those episodes are the real evidence and are chosen with care: the gold inflow that followed resumption in 1879, the silver agitation and deflationary scare of 1893, the discount rate increases of 1920, and the doubling of reserve requirements across 1936 and 1937. In each, something outside the American business cycle changed the money stock, and nominal income followed in the same direction and roughly in proportion. The statistical apparatus around them, the turning points, the amplitudes, the behaviour of velocity, is supporting material rather than proof, and mistaking it for proof is the commonest way of overstating what the book demonstrates.
The distinction matters because the two kinds of evidence fail differently. Correlations are vulnerable to the objection that banks create deposits when borrowers want them, so a money stock that turns before income may be a banking system anticipating a recovery it can already see. The episodes are immune to that. What they cannot supply is quantity: there are perhaps a dozen of them across ninety-three years, several of those wartime, and a dozen observations will not sustain a stable coefficient linking money to income.
So a weaker proposition is proved very thoroughly and a stronger one by accumulation. The weaker one is that money is sometimes an independent cause of large movements in income. The stronger one, which most readers take away and which the concluding pages encourage, is that it is generally the dominant one. Nothing in the historical record separates the two, and the space between them is where forty years of subsequent argument took place.
Everything else here could stand or fall on its own; this cannot be given up without the thesis going with it. What happened to the money stock across those four years is presented as the depression's principal mechanism rather than its consequence, and the charge against the Federal Reserve is not that it acted wrongly but that it stood still while the banking system did the contracting.
The mechanics offered are precise and are the part that has survived best. Bank failures raised the public's preference for currency over deposits and the surviving banks' preference for reserves over lending, and both ratios multiply against high-powered money in the same direction. High-powered money in fact grew by roughly a sixth across the four years; the collapsing ratios took a third out of the money stock in spite of it, which is why the indictment is one of insufficiency rather than of contraction. Open market purchases at any of three junctures, early 1930, January 1931 after the first banking crisis, or September 1931 when Britain left gold, would have arrested the arithmetic, at sums the System had already managed in the spring of 1932.
Why it did not act is where the account turns from arithmetic to institutional history, and where it is at once most interesting and least verifiable. Benjamin Strong of the New York bank died in 1928, and with him the informal authority that had made the System behave as one; what followed was a Board and regional banks among which each could obstruct and none could lead. Doctrine compounded it: low nominal rates and light borrowing at the discount window were read as proof that money was already easy, when both were symptoms of a collapse in the demand for credit.
The weakness is structural rather than factual. The failure alleged is a failure to act, so the central claim is counterfactual: what collapsed was the deposit multiplier, produced by depositors and bankers, and the assertion is that an intervention never attempted would have reversed it. Two rivals press on that gap. One is that a banking system destroyed as a going concern does direct damage, by wiping out the knowledge of who is creditworthy, which is no monetary channel at all. The other is the gold standard: the contraction was worldwide, and a bank defending a parity is not free to expand, which is what the System said of itself when it raised the discount rate days after Britain left gold. The authors answer that one squarely, holding the American gold reserve ample enough that the constraint never bound, and it is the reply their successors have found least convincing.
The comparison the book keeps returning to is between the decades before 1914 and the decades after. The National Banking system was badly designed, with a currency that could not expand when it was wanted, and it produced recurrent panics. Yet 1907 was contained within months and 1930 to 1933 was not, and the difference between them is the institution that was meant to be the improvement.
The explanation offered is that the Federal Reserve did not merely fail to supply what the older arrangement lacked. It disabled what the older arrangement had. When banks could no longer meet demands for cash, the clearing houses of the previous era restricted the convertibility of deposits into currency, a response that was ugly, of doubtful legality, and effective within weeks. After 1914 that defence was neither used nor permitted, because a lender of last resort now existed and was expected to make it unnecessary. It existed and it did not.
This is the sharpest thing in the book, and the point at which its authors and its most enthusiastic readers part company. The finding is that a monopoly on emergency money creation, once established, destroys the decentralised defences it displaces, and is thereafter only as good as the people running it on the day. A reader arriving from the free banking literature will notice that this is most of an argument against holding such a monopoly at all, and that the authors decline to draw it.
Their own conclusion runs the other way, and the diagnosis explains why. The harm is located in the exercise of discretion rather than in the possession of the power, and every fault the account identifies, the leadership vacuum, the diffusion of authority, the doctrine that mistook cheap credit for easy money, is correctable inside the institution without touching its charter. The constant money growth rule Friedman's name now carries belongs to his separate policy writing of those years, not to this history; what the history supplies is the evidence for binding the monopolist. The finding about the disabled defences is left lying where it fell, and it is the sharpest thing the authors decline to pick up.
The book works because it is written as history rather than as econometrics: a continuous money series built from bank records, set beside what was happening and what the authorities said they were doing, year by year for ninety-three years. Nobody had assembled the series before and it remains in use. The method is also what makes the causal claims hard to pin down.
A narrative can accommodate any outcome, because a reason can always be found for why this episode behaved differently, and a reader cannot easily tell an explanation from an excuse. The standing objection, pressed hardest by Kaldor and by Tobin, is that the direction of causation is precisely what the method cannot settle. If banks lend when firms want to borrow, the money stock will lead income for reasons having nothing to do with the authority, and the leads and lags reported here are what one would expect on either account.
A definitional choice does real work at this joint and is easy to miss. The series the argument rests on counts commercial bank time deposits as money, and the relationships are markedly weaker on the narrower measure. The authors defend the choice on empirical grounds, which is to say on the grounds that it works better, and a critic is entitled to observe that the thing being explained helped select the instrument that explains it. Something similar applies to the money-income relation itself, reported as stable while the intervals between money's turning points and income's are wide enough to accommodate several theories.
None of this has stopped the book winning. Its method was formalised rather than abandoned: identifying policy shocks from the authority's own contemporaneous record, which Romer and Romer later systematised, is this technique with the discretion taken out. And the main rival account of the depression, that the destruction of banks as lending institutions did damage no injection of reserves would have repaired, was framed by Bernanke as a supplement to Friedman and Schwartz rather than a refutation. That is what winning looks like: the opposition now argues inside the frame.
Less than its reputation claims and more than its critics allow, and in a different place from either. That money matters was not new; the quantity theory is centuries old and Friedman had already restated it. What was new was the series: a continuous, defensible estimate of the American money stock back to 1867, assembled where no official figure had ever existed, and a great deal of the argument is simply that series becoming visible for the first time. The second novelty is a method rather than a finding, namely reading the authority's own contemporaneous reasoning to identify moments when policy moved for reasons of its own, which is now standard practice under other names. The famous conclusion about 1929-33 is the least new thing here. Currie and Hawtrey had already said the Federal Reserve failed. What Friedman and Schwartz added was the arithmetic that made the failure impossible to answer with an assertion that policy had been easy.
Both hold that sustained monetary outcomes follow from what the issuing authority did rather than from shocks it absorbed badly — Friedman from the Federal Reserve's own record in 1929-33, Hayek from the issuer's incentives. They part company entirely on the remedy.
SCOPE · on cause, not on cure · Inflation is chosen, not sufferedBoth read monetary outcomes as decisions someone made rather than weather the authorities endured — Friedman from the Federal Reserve's own deliberations, Graeber from who benefits when a debt is enforced or forgiven. What they agree on is that there is an agent to name.
SCOPE · on agency, not on interest · Inflation is chosen, not sufferedFriedman would bind the monopolist with a fixed rule for the money stock. Hayek argues that a rule is a promise an institution makes to itself, withdrawn in exactly the conditions that make it matter, and that only free entry disciplines an issuer.
SCOPE · on rules versus entry · Money without a monopolyFriedman's answer to a badly run monopoly is a rule that binds it. Ammous denies there should be a party capable of being bound at all. The disagreement is not about how much to issue but about whether issuance is a decision anyone should hold.
SCOPE · on whether an issuer is needed · Money without a monopolyNot end to end, and almost nobody has. The stretch covering 1929 to 1933 was reissued as a book of its own for a reason: it is where the evidence, the mechanism and the argument converge. The National Banking era and the postwar decades carry the series rather than the case, answering questions a reader brings to them and making none of their own, which is why they are consulted rather than read through. Anyone wanting the argument without the evidence should be warned that here the argument is the evidence, and that the compressed versions in circulation are exactly what the identification objection is aimed at.
Anyone who believes monetary policy was tried in the Depression and failed, which was the settled view this book unsettled and which returns in a new form after every crisis. Also anyone who reaches for private issue as the answer to central banking and has not read the record of what central banking did with the power, assembled by authors who drew the opposite conclusion from it.
The construction of the money series itself, roughly a third of the volume and the part specialists still argue with; the greenback and resumption period, where the book is at its most technical; and the treatment of velocity, which carries more theoretical load than its low profile suggests. Also the war years, where the identification is at its cleanest and the price and velocity data at their least trustworthy, because rationing and price control were holding the numbers down.
This assessment describes what Milton Friedman & Anna Jacobson Schwartz argues in Princeton paperback edition, Princeton University Press for the National Bureau of Economic Research, 1971, and is derived from that edition. It is a position taken about the argument — not a recommendation from someone who read the book for pleasure and enjoyed it.
https://summarylibrary.com/a-monetary-history-of-the-united-states · summary ref F-1963-01 · updated 2026-08-22 · the summary is not the book.