"If I had to make my literary Will, and my literary Acknowledgements, I would have to own that I owe more to Macaulay than to any other English writer." -- Winston Churchill to R. V. Jones, as related in Jones's book Most Secret War

"Lord Stanhope once gave me a curious little proof of the accuracy and fulness of Macaulay's memory. Many historians used often to meet at Lord Stanhope's house; and in discussing various subjects, they would sometimes differ from Macaulay, and formerly they often referred to some book to see who was right; but latterly, as Lord Stanhope noticed, no historian ever took this trouble, and whatever Macaulay said was final." -- Charles Darwin, in his autobiography

Those two endorsements were what lead me to check out Thomas Babington Macaulay's History of England. I had thought I had read good history books, but this one is on an another level entirely. It's the only book which I liked so much that I took the Project Gutenberg copy, read through it and corrected some scanning errors, and put it up on this website. Before starting it, Macaulay wrote that he was "acquainted with no history which approaches to [my] notion of what a history ought to be"; and as far as I know there is still no other work of history anywhere near as good, despite many attempts at imitation.

Macaulay chose as his focus the direst crisis that his countrymen had faced since the Dark Ages: the one whose resolution was known as the Glorious Revolution. That is not a popular topic today; people scarcely know the name. It's old history, and even as regards old history the French Revolution is much more frequently studied. But the French Revolution is a story of people new to politics getting power, trying to quickly make a lot of big changes in society, then getting into quarrels among themselves which degenerated into a reign of terror. The French Revolution tested a lot of overly-simplistic theories of politics, and the test was a failure: as Tocqueville documented, what the French had come to consider great innovations of the Revolution (such as centralization) had actually been how the old regime operated under a facade of tradition; the innovation had merely been dropping the facade. The whole episode can serve as an example of the French proverb "La plus ça change, la plus c'est la meme chose" -- usually confusingly translated as "the more things change, the more they stay the same", a pure self-contradiction; a better though too cynical translation would be "the more they yammer about change, the more it's the same old shit". People are creatures of habit; it doesn't work to try to change everything at once, even if the final state would be completely tenable given time to adjust to it.

The French Revolution got rid of old abuses such as government support of the nobility and clergy; but it mainly teaches lessons about what not to do. Napoleon, who put an end to the revolution, is still honored not only for his military genius but for his thorough rewriting of French law, an effort whose influence extends even to the US today, where the law of the state of Louisiana owes much to the Napoleonic Code. Still, his regime was a dictatorship, complete with press censorship and secret police -- repression severer than the old regime had been inflicting, combined with a thirst for glory that plunged all of Europe into war and had to be crushed by foreign armies.

Now, some people still are in desperate need of the lessons of the French Revolution, and particularly of the lesson that overshadows the whole: not to run one's country into bankruptcy. When, in the recent framing of the Iraqi constitution, the framers were prohibited from running for office in the subsequent government, it was a repetition of another mistake the French revolutionaries made. But it is unsatisfying just to study mistakes. The Glorious Revolution, in contrast, furnishes ample material for an advanced degree in what to do in dire crises, as well as what not to. Most of the leaders of that period had long experience in practical politics. Their decades of struggle encompassed wars on land and sea, revolutions (some failed, some successful), a military dictatorship, palace intrigues, parliamentary maneuvering, politicized trials, and monetary troubles. The outcome was the modern system of parliamentary government; it was the origin of such things as freedom of the press and religious toleration under law. It laid the foundations for an ascent to prosperity and power which, for more than a century, made Britain the envy and the terror of the rest of the world.

This book chronicles all that; and in a way it itself is foundational: its influence was extraordinary. It was popular for a long time, was translated into many languages, and there are still lots of copies floating around. Readers will not find any new and unusual arguments in it; they have all passed into common currency. But they probably will find (as I did) arguments that they'd disdained but find they can appreciate in the right context. Talk of oppressed peoples, for instance, has been so overused by the modern left as to have become tiresome, but Macaulay does it well: not exalting the oppressed, and indeed trusting their oppressors' accounts of events more, "for fraud is the resource of the weak". When the newly-liberated horribly misgovern, he doesn't try to pretend that the resulting government is good, but simply regards it as the product of people who are working at a task where knowledge and experience are necessary, and have neither.

Today this book is no longer popular. Even history professors generally don't assign it as reading to their students. It's an entertaining, positive take on Western civilization, and they prefer dreary, negative ones. It's not even the sort of book that can be dished out to students in small doses, with lots of pooh-poohing, to try to immunize them against positive thinking: its ideas are too infectious for that, being presented in an unpretentious way that is easily and naturally understood, and which enters into the head without warning and would be hard to exorcize. Usually when a writer makes sweeping generalizations about historical events, something occurs later that is clearly contrary to his ideas, causing even sympathetic readers to smile at it and think "nice try, but no", and which hostile readers can single out to badmouth the whole thing. In this case it's been a century and a half, yet the ideas in the book still fit eerily well to events that have happened since it was written. Indeed, a history professor today would need exceptional equanimity to have his students read quotes like

"The liberal education of youth passed almost entirely into their hands, and was conducted by them with conspicuous ability. They appear to have discovered the precise point to which intellectual culture can be carried without risk of intellectual emancipation."

It might hit too close to home, especially with "they" being a world-spanning group of ideologues who, "constant only to the fraternity to which they belonged, were in some countries the most dangerous enemy of freedom, and in others the most dangerous enemy of order". He might find his students reading further on their own initiative; and for all the advantages that being a living person with power to issue grades gives a professor, he might find himself losing arguments to this old book in an embarrassing fashion.

In modern terms Macaulay would be something of a libertarian; but modern libertarians have been so far from power for so long that they often think things done by government are abuses when really they are inevitabilities. Macaulay had stood up in front of the House of Commons, had told them "The danger is terrible. The time is short.", and had seen them vote as he urged. He had been the chief architect of a new criminal code for India, the descendant of which India still uses today. He knew the practicalities of power; he wouldn't have dreamed of arguing that quarantines don't work or that armies are unnecessary. He knew that sometimes you have to roll with the punches that public opinion delivers; that it can be better to bend than than to break. His remarks on the Toleration Bill and the difference between the "active statesman" and the "contemplative statesman" bear a strong resemblance to Rickover's letter on the difference between "practical" and "academic" nuclear reactors. Yet being practical strengthens his theoretical positions, rather than weakening them: by admitting that something is a concession to the public mood or to established habits, he avoids the trap of declaring concessions unacceptable, having to make them anyway, then trying to spin some theory which justifies them.

In some respects, indeed, his positions are more extreme than those of any libertarian I've encountered today, who, when writing of laws regulating the press, commonly are weak enough to think that they'll have something resembling their purported effect. Macaulay is having none of it:

"...no person who has studied with attention the political controversies of that time can have failed to perceive that the libels on William's person and government were decidedly less coarse and rancorous during the latter half of his reign than during the earlier half. And the reason evidently is that the press, which had been fettered during the earlier half of his reign, was free during the latter half. While the censorship existed, no tract blaming, even in the most temperate and decorous language, the conduct of any public department, was likely to be printed with the approbation of the licenser. To print such a tract without the approbation of the licenser was illegal. In general, therefore, the respectable and moderate opponents of the Court, not being able to publish in the manner prescribed by law, and not thinking it right or safe to publish in a manner prohibited by law, held their peace, and left the business of criticizing the administration to two classes of men, fanatical nonjurors who sincerely thought that the Prince of Orange was entitled to as little charity or courtesy as the Prince of Darkness, and Grub Street hacks, coarseminded, badhearted and foulmouthed. Thus there was scarcely a single man of judgment, temper and integrity among the many who were in the habit of writing against the government. Indeed the habit of writing against the government had, of itself, an unfavourable effect on the character. For whoever was in the habit of writing against the government was in the habit of breaking the law; and the habit of breaking even an unreasonable law tends to make men altogether lawless. However absurd a tariff may be, a smuggler is but too likely to be a knave and a ruffian. However oppressive a game law may be, the transition is but too easy from a poacher to a murderer. And so, though little indeed can be said in favour of the statutes which imposed restraints on literature, there was much risk that a man who was constantly violating those statutes would not be a man of high honour and rigid uprightness. An author who was determined to print, and could not obtain the sanction of the licenser, must employ the services of needy and desperate outcasts, who, hunted by the peace officers, and forced to assume every week new aliases and new disguises, hid their paper and their types in those dens of vice which are the pest and the shame of great capitals. Such wretches as these he must bribe to keep his secret and to run the chance of having their backs flayed and their ears clipped in his stead. A man stooping to such companions and to such expedients could hardly retain unimpaired the delicacy of his sense of what was right and becoming. The emancipation of the press produced a great and salutary change. The best and wisest men in the ranks of the opposition now assumed an office which had hitherto been abandoned to the unprincipled or the hotheaded. Tracts against the government were written in a style not misbecoming statesmen and gentlemen; and even the compositions of the lower and fiercer class of malecontents became somewhat less brutal and less ribald than in the days of the licensers.

"Some weak men had imagined that religion and morality stood in need of the protection of the licenser. The event signally proved that they were in error. In truth the censorship had scarcely put any restraint on licentiousness or profaneness. The Paradise Lost had narrowly escaped mutilation; for the Paradise Lost was the work of a man whose politics were hateful to the ruling powers. But Etherege's She Would If She Could, Wycherley's Country Wife, Dryden's Translations from the Fourth Book of Lucretius, obtained the Imprimatur without difficulty; for Dryden, Etherege and Wycherley were courtiers. From the day on which the emancipation of our literature was accomplished, the purification of our literature began. That purification was effected, not by the intervention of senates or magistrates, but by the opinion of the great body of educated Englishmen, before whom good and evil were set, and who were left free to make their choice. During a hundred and sixty years the liberty of our press has been constantly becoming more and more entire; and during those hundred and sixty years the restraint imposed on writers by the general feeling of readers has been constantly becoming more and more strict. At length even that class of works in which it was formerly thought that a voluptuous imagination was privileged to disport itself, love songs, comedies, novels, have become more decorous than the sermons of the seventeenth century. At this day foreigners, who dare not print a word reflecting on the government under which they live, are at a loss to understand how it happens that the freest press in Europe is the most prudish."

Now, these passages I've been quoting are not the average passage in the book; it is a history, not a long theoretical tract. Most of it just straightforwardly describes events as they happened. But when Macaulay does make a theoretical point, he does so openly, not by hints and allusions. His opinions of people, likewise, are freely given; there is no pretense of objectivity. When he's describing someone who in modern language is a sociopath, he does so as such:

"He would have been a more sagacious politician if he had sympathized more with those feelings of moral approbation and disapprobation which prevail among the vulgar. For his own indifference to all considerations of justice and mercy was such that, in his schemes, he made no allowance for the consciences and sensibilities of his neighbours. More than once he deliberately recommended wickedness so horrible that wicked men recoiled from it with indignation. But they could not succeed even in making their scruples intelligible to him. To every remonstrance he listened with a cynical sneer, wondering within himself whether those who lectured him were such fools as they professed to be, or were only shamming."

This openness makes the book easier to read than the work of a modern historian who has formed definite opinions but is too delicate to utter them: you don't have to guess at Macaulay's agenda. It doesn't completely immunize him from charges of secretly trying to push another agenda too, but it greatly lowers the probability. When, as in the case of William Penn, most of the incidents mentioned are discreditable to Penn but he is introduced with praise for what he did in other parts of his life (and in particular the founding of Pennsylvania), that too is best taken literally, not as a sly way to lull suspicions so as to better slip the knife in: the man did good things, they were just off-topic for this book. Indeed, though pacifists have their good points, personal friendship with a tyrant is not a situation in which they tend to shine.

Now, many of the events Macaulay mentions involving Penn happened in private; evidence relating to them is meager and questionable, and has indeed been much questioned. Even by Macaulay's telling, Penn didn't do real harm; he just meddled in matters where he couldn't do any good. His embassy to Oxford University is reminiscent of the adventures of the Swedish businessman Birger Dahlerus, who vainly shuttled back and forth between Adolf Hitler and the British government in the last days before war, trying to get them to agree. In such circumstances, mediation is unlikely to be successful, except by getting the wrong side to cave in; it is the sort of thing that is tried by someone who, to use Macaulay's description of Penn, is "a man of eminent virtues", but "not a man of strong sense".

Despite the leading quote from him, Winston Churchill is another who objected, in his case to some of Macaulay's remarks about his famous ancestor John Churchill, a great military commander who eventually was made Duke of Marlborough. As an aside, this book follows a convention on the naming of such people which makes it a bit confusing to read: before being raised to the nobility the man is referred to as Churchill, and after that usually as Marlborough -- and he's far from the only one who undergoes such a name change. (I'll be using only the latter name here, so "Churchill" will only mean Winston.) I suppose people who are used to this know to make a mental note whenever they see the announcement of someone being ennobled: remember that person A might hereafter be referred to as B. It's like having to remember that acetaminophen and Tylenol are the same thing when when reading about medications: the annoying dual naming is designed to enhance the power of people who really are already quite powerful enough.

In any case, reading Churchill's own book about Marlborough did not leave me with a very different impression of the man than I'd gotten from Macaulay; it's more a matter of spin than of factual disagreement. When Macaulay writes of Marlborough's wife that

"...to the last hour of his life, Sarah enjoyed the pleasure and distinction of being the one human being who was able to mislead that farsighted and surefooted judgment, who was fervently loved by that cold heart, and who was servilely feared by that intrepid spirit."

and Churchill writes in response that this is just to say that the two were really in love, it's not a disagreement over facts, just the difference in perspective between Macaulay, who never married, and Churchill, a physical product of the very love being talked about.

Another difference is over Marlborough's betrayal of King James II. Macaulay describes it using epithets like betrayal and treason, which Churchill objects to. But one of the main themes of Macaulay's book is that James was just asking to be betrayed and that by betraying him Marlborough helped save Britain. On the whole, the effect of his using those epithets is to raise the reader's opinion of treason and betrayal, not to lower the reader's opinion of Marlborough.

One objection of Churchill's, though, falls outside the category of spin; it regards the charge that Marlborough betrayed a British raiding expedition on Brest by leaking news of it to James (then in exile), and thus to the French. After about a hundred pages of examination of the available sources, he concludes that the most probable scenario is that Marlborough did leak the news, but only after he knew others had done so (since security was lousy; the expedition was the talk of the local bars), with the intention of writing too late to give the French time to respond, so as to boost his own credit with the exiled king without injuring his country, rather than (as Macaulay would have it) to discredit a fellow general by making his expedition fail. That seems like a fair correction, though of course even wanting to raise his credit that way was rather a treasonable thing for Marlborough to do -- and yes, this was the same James he'd already most severely betrayed, so to correspond with him in the first place was a very strange act, even by the loose standards of the time.

Churchill, by the way, does a much better job on military matters than Macaulay does: as someone who'd served in the military himself and thought about it deeply, he simply understood the business better. Churchill goes into details of tactics, strategy, and logistics; Macaulay talks about heroism, determination, and skill, saying little about what that skill consists of. What he does say is fair enough, but it's clear which of the two would wipe the floor with the other in any trial of military abilities. Yet when it comes to the highest level, that of financing war, Macaulay is again better; Churchill has too much of the military's attitude that support should be taken for granted, whereas Macaulay takes anti-war arguments seriously even when refuting them.

Churchill also wrote a preface to a reprint of the book The New Examen, by John Paget, a severely hostile response to certain parts of Macaulay's history. The book originally appeared near the end of Macaulay's life; he contemptuously refused to respond to it, and it had little influence. Churchill, though admitting that the book's criticism was overblown, recommended that anyone who reads Macaulay should also read it. I've read it, and disagree; it's about two orders of magnitude less information-dense than the book it's criticizing. But it contains two criticisms which strike me as good and strong enough to be worth repeating. One (which is buried in the middle of the book) is the comment that Macaulay has often paraphrased conversations even inside quotes: that is, his quotes are often not exactly what the people were originally reported to have said. Paget gives several examples of this, quoting both the source that Macaulay references and Macaulay's version. He admits that the meaning of the quotes is unchanged; but this does explain why so many of the people in the book seem to be such great speakers, and so few of them halting and incoherent; people in the 1600s were not all actually that eloquent.

The other thing worth mentioning is Macaulay's treatment of John Graham (aka "Claverhouse", then "Dundee"). Whereas the arguments over Penn and Marlborough (both engaged in by Paget) left me with the feeling that that I'd seen the counterargument, that indeed there are likely some errors, but not necessarily and in any case they don't make much difference, in the case of Dundee the accusation of being the bloody butcher of Scotland seems solidly wrong. Now, the man was suppressing a violent insurgency, and one doesn't do that by holding tea parties. Plenty of things done by government soldiers were atrocious. But it seems the incidents that personally implicate Dundee in the worst atrocities didn't happen as related; instead they were rebel propaganda that had survived until Macaulay's era. Paget even managed to dig up a letter from Dundee talking about how he was adopting a mild policy versus the rebels (mild for that era, of course; in today's world people would be screaming bloody murder at it, but life was cheaper then). In any case, when, after Dundee's death at the battle of Killiecrankie, Macaulay remarks that

"it is a strong proof of the extent and vigour of his faculties, that his death seems every where to have been regarded as a complete set off against his victory."

it surprised me: he'd introduced the man as a vicious butcher, not a master statesman. That it had been the latter all along seems most probable. I was at a loss to explain this level of error -- since, as mentioned, in general Macaulay knew how unreliable the claims of the oppressed are -- until I realized that growing up in Scotland, he would have learned these stories during childhood; and it's a common human weakness that the first things one learns can be strangely immune to reason, even in people who reason well about the world in general. (I've even noticed this in myself; it took me quite a while to stop being a Unix partisan. I still like the operating system, but no longer bristle when it's criticized.)

Macaulay aimed to produce a history entertaining enough that people would read and remember it. That always involves a bit of compromise as regards accuracy, as witness the preceding paragraphs. Still, it's pretty obvious when he's getting into sketchy territory for the purposes of adding color; and my guess is that the difference between Macaulay's account and the best version that could be assembled out of modern materials is less than the difference between the latter and the real truth (for some people manage to pull the wool over historians' eyes forever).

The book remains readable, though language usage has changed. As I've gathered from context, where Macaulay writes "rude" (as in "rude huts"), we would say "crude"; where he writes of someone having "good parts", he is referring to brains, rather than arms, legs, etc.; and where he writes of "prescription", as in a "prescriptive right", he's referring to precedent, not to doctors' orders or anyone else's orders. But there aren't many such changes; it's nowhere near as hard as reading Shakespeare.

The main thing that quite reasonably discourages a modern reader is that the events described aren't what people argue about these days. Historical arguments these days are most often about the two world wars. Those too were desperate crises, and are vastly more relevant to today's world -- because they directly shaped it, and perhaps even more importantly, because they took place in a technological environment similar to today's. This book was written before cars, and describes a time before steam engines. It was written before telephones and radios, and describes a time before electricity. It was written before antibiotics, and describes a time before antiseptic technique was accepted. It was written before smokeless gunpowder, and describes a time when swords were sometimes used to win major battles. Macaulay spends the whole of his Chapter III (about 45,000 words) describing the differences between his time and the era his history covers; and he frequently revisits those differences throughout his book. This is of some help to the modern reader, who still, however, has to bridge the gap between 1850 and the present. That is difficult even for people with a good grip on the history of technology. For everyone else, this history must appear to be a sort of fairy tale -- or rather the reverse of a fairy tale. "Any sufficiently advanced technology is indistinguishable from magic", wrote Arthur C. Clarke. This book is about the opposite: people who didn't have the sort of "magic" that almost everyone today takes for granted. It is about people who had to walk where we can literally fly. And yet in a way Macaulay could see what was coming:

"It is now the fashion to place the golden age of England in times when noblemen were destitute of comforts the want of which would be intolerable to a modern footman, when farmers and shopkeepers breakfasted on loaves the very sight of which would raise a riot in a modern workhouse, when to have a clean shirt once a week was a privilege reserved for the higher class of gentry, when men died faster in the purest country air than they now die in the most pestilential lanes of our towns, and when men died faster in the lanes of our towns than they now die on the coast of Guiana. We too shall, in our turn, be outstripped, and in our turn be envied. It may well be, in the twentieth century, that the peasant of Dorsetshire may think himself miserably paid with twenty shillings a week; that the carpenter at Greenwich may receive ten shillings a day; that labouring men may be as little used to dine without meat as they now are to eat rye bread; that sanitary police and medical discoveries may have added several more years to the average length of human life; that numerous comforts and luxuries which are now unknown, or confined to a few, may be within the reach of every diligent and thrifty working man. And yet it may then be the mode to assert that the increase of wealth and the progress of science have benefited the few at the expense of the many, and to talk of the reign of Queen Victoria as the time when England was truly merry England, when all classes were bound together by brotherly sympathy, when the rich did not grind the faces of the poor, and when the poor did not envy the splendour of the rich."

Still, in today's world Macaulay would be an ignorant man. Late in life, he wrote to a friend "I often regret, and even acutely, my want of a Senior Wrangler's knowledge of physics and mathematics; and I regret still more some habits of mind which a Senior Wrangler is pretty certain to possess." ("Senior Wrangler" was and is still the title given by Cambridge University to the year's top graduate in mathematics.) But to describe the period Macaulay chronicled, his knowledge was sufficient: it was the time when Newton was coming out with his theories, and the world had not yet made much use of them -- though, as Macaulay relates, Newton himself did make very good use of them.

I'd previously read of Newton's time as Warden of the Mint as being a waste of time taking him away from science, or as a government sinecure. Really one might as well describe the Manhattan Project that way. Newton was called in to help address a dire and rapidly-worsening monetary crisis. This was nothing like any of today's financial crises: it was the era of hard money, and it was the monetary unit itself (silver coins) that was melting down -- or to be literal, being clipped and then the clippings being melted down. The coins had been crudely manufactured, and it was easy to clip a small piece off one before passing it on. The worse things got, the faster the deterioration became, since new clipping is harder to detect when most coins are already clipped. Things had gotten to where in some places the average coin was of less than half the nominal weight; even imposing the death penalty for clipping hadn't stopped it.

The fix involved introducing new "milled" coins, which were precisely made and had marked rims so that clipping would be obvious. But this had to be done fast: not only was the problem naturally escalating, but the nation was in the middle of a war with France. Today, anyone who would be given the job of making such machinery work better would have to be familiar with Newton's laws, and at that time who better to give the job to than Newton himself? Macaulay describes the outcome:

"The old officers of the Mint had thought it a great feat to coin silver to the amount of fifteen thousand pounds in a week. When Montague talked of thirty or forty thousand, these men of form and precedent pronounced the thing impracticable. But the energy of the young Chancellor of the Exchequer and of his friend the Warden accomplished far greater wonders. ... The weekly issue increased to sixty thousand pounds, to eighty thousand, to a hundred thousand, and at length to a hundred and twenty thousand. Yet even this issue, though great, not only beyond precedent, but beyond hope, was scanty when compared with the demands of the nation."

With enough patience, it was eventually enough. But if the recoinage had failed it likely would have toppled the government and led to the loss of the war; it might even have ended in the reinstatement of James II (one of France's aims). It's questionable whether science such as Newton's would have been possible in the resulting despotism.

Still, Macaulay's description has something of the air of a tale in which Gandalf is called in and by secret wizardry solves the problem. Modern historians need to do better than that; a modern historian of the second world war, trying to describe the history of radar, would need a solid grounding in math and physics to produce anything better than a superficial description. That wasn't needed to cover the 1600s; the sort of level on which innovation occurred then is exemplified by the designing of a bayonet which was attached to the outside of the musket where it does not block the shot, rather than being stuck into the muzzle -- something which Macaulay could fully appreciate. Likewise, most of the finance and economics of the 1600s was mathematically simple; there is only one place in the book where Macaulay remarks that a piece of economics is a job for "a political economist", not for "a historian"; and that is where, in the description of the history of Britain's national debt, he ventured into his own era. Today, a good historian would need to understand that part of economics, since today it is history.

Now, in saying what modern historians need to know as technical background, I do not mean to imply that they do in fact know it; few of them know much of it, and none of them know all of it. If they did, there'd really be no point in reading this old book: some modern historian would have written something much better. As it is, their lack of knowledge not only damages what they say about technical subjects but bleeds over into errors about human nature.

In this book, there is only one place where a lack of math seems to really be regrettable. It is in the description of the emergence of the first banks in England: the way that the banking system amplifies the quantity of money is misexplained, and its associated risks are laughed at. Today, this part of economics is understood by a large number of people, but I've never seen a great explanation of it. Yet the basic idea is fairly simple. To illustrate it, suppose there are a hundred dollars in gold coin in a bankless economy, and then a bank opens. People deposit that hundred dollars in the bank in exchange for banknotes, and start paying each other in banknotes, each of which represents a gold coin and can be taken back to the bank in exchange for a gold coin from the vault.

If the bank just left all those coins in the vault, this exercise would have only a minor impact on the economy. It would be an exercise in one-for-one substitution of metal with paper, interesting mainly from the point of view of physical security. Banks which operate this way, such as the Bank of Amsterdam as described by Adam Smith in The Wealth of Nations, have to charge fees for storage; they have no other way to make a profit.

But there have been few such banks; even the Bank of Amsterdam eventually started doing what banks generally do these days, which is to make their profits by loaning the money out, keeping a reserve in the vault (say, ten percent of it) to meet immediate demands. This lending introduces risks -- the risk that the reserve will not be enough for a sudden surge of withdrawals, and the risk that the bank will make too many bad loans. It also amplifies the quantity of money in the economy, since people have the hundred dollars in banknotes, representing the deposited gold, and the people who borrowed the 90 dollars have it in gold, which adds up to a total of 190 dollars in circulation. Furthermore, the amplification of money is not yet over: those 90 dollars in gold get deposited again in exchange for more banknotes, mostly lent out again, deposited again, ad infinitum. In the end, all the 100 dollars of gold is in the vault, and in circulation there are 1000 dollars in banknotes. (The number 1000 is not a vague guess, but the result of the relevant formula -- the formula for the sum of an infinite geometric series. Of course in reality the series is not infinite; as usual with formulas in economics, it only approximates reality.)

This scenario was close to what was happening in England, with the first bankers being goldsmiths; as Macaulay describes it:

"...It was at the shops of the goldsmiths of Lombard Street that all the payments in coin were made. Other traders gave and received nothing but paper."

Furthermore, the goldsmiths were lending out the money deposited with them:

"These usurers, it was said, played at hazard with what had been earned by the industry and hoarded by the thrift of other men."

"Usury" was not, in the 1600s, the technical legal term it is now, that means lending at very high interest rates; instead it was a word for charging any interest whatsoever; but then as now it was a term of disapproval. Many Christians back then had the same attitude towards paying interest that strict Moslems do today (it's all usury!), with the odd result that the trade of moneylending was often left to Jews. "Playing at hazard" meant gambling. Banks that take deposits and lend out money always act to expand the quantity of money; and indeed that seems to have been one of the claimed advantages:

"The new system, it was said, saved both labour and money. Two clerks, seated in one counting house, did what, under the old system, must have been done by twenty clerks in twenty different establishments. A goldsmith's note might be transferred ten times in a morning; and thus a hundred guineas, locked in his safe close to the Exchange, did what would formerly have required a thousand guineas, dispersed through many tills, some on Ludgate Hill, some in Austin Friars, and some in Tower Street."

But there the explanation falls down. Certainly a note for a hundred guineas is easier to carry than a hundred gold guineas; and the guineas have to be counted to make sure there are a hundred of them whereas the note can simply be read, which indeed "saves labor". But the thing that "saves money", or in modern language increases the quantity of money, is not that notes for a hundred guineas get traded around ten times as fast as a hundred guineas do; instead it is that the hundred gold guineas get transformed by repeated depositing and lending into notes for a thousand guineas.

Macaulay goes on to treat the element of risk as minor, laughing at an economist of that era who refused to deposit his coin with the goldsmiths; but really the risk is fundamental: when there are only a hundred gold guineas, and all of the holders of the thousand guineas' worth of notes try to withdraw them in gold, only a hundred gold guineas will come out of the vault.

These days, the old role of physical gold or silver is played by paper bills, and the old role of banknotes is played by money in bank accounts, which may be transferred electronically or by writing checks. But the formula for amplification of the quantity of money is the same. In either case there is also money that people keep as cash rather than depositing; that money does not get amplified.

The way the risks manifest themselves is in banking panics. If all the depositors try to withdraw their deposits, not only will most of them fail, but it is the first withdrawers who will succeed. So once it is rumored that the bank is in trouble, people must scramble to be among the first to withdraw -- which immediately breaks the bank, even if it were healthy to begin with.

The result is that a bank may be in business for years, exhibiting no signs of distress, then on rumors of unsoundness it is suddenly ruined. (Historically, usually they've been true rumors, but they've been rumors nevertheless.) There should be little surprise that bankers are known for having no sense of humor about disparaging remarks. It's almost a social duty for them: when a bank is ruined by people making a run on it, huge numbers of people who did nothing wrong are ruined too. One doesn't even need to have personally trusted the bank to be ruined; it can be enough to have trusted someone who trusted the bank. A banking panic is not a discriminating instrument that can "cut out the rot": it usually starts in rot, but can quickly destroy healthy parts of the economy too. So, bankers normally would rather not have people talk about such banking risks -- an attitude which goes far to explain the lack of great explanations: there are plenty of bankers who perfectly understand these issues, but it goes against their social norms to draw attention to them.

To that basic scenario there is a wrinkle of complexity to add: there is usually more than one bank, and if just one bank fails, the damage is minor. The bank may have paid out all the cash in its vault, but it still has assets: its outstanding loans. Commonly in such a situation other banks step in and buy those loans, getting them at bargain prices due to the deal having to be done in haste. In this way, unless the loans had really cratered, depositors would eventually get most of their money back. Indeed, in old accounts of bank runs, one often reads of a failed bank being bought whole by another bank; in that case its depositors would not lose at all. Such purchases are commonly presented as public-spirited acts to restore confidence in the banking system, but they can be done for pure profit. Looking over the failed bank's books, the prospective purchaser might conclude that it hadn't really failed at all -- that the rumors which sparked the run had been false -- and that they would be getting a good deal.

Still, restoring confidence in the banking system was always a goal they had in mind, since if the panic spread enough there wouldn't be any other banks to step in. Then the scenario would play out as if it were all the same bank: they'd all be emptied of cash, they'd all fail, and all the money would be in the hands of the people who got to the banks first -- all resulting in a level of social disorder that can only be imagined, since no society has ever just blithely let it play out to the very end.

Besides its risks, the chief importance of the amplification of the quantity of money is in inflation and deflation: if there is ten times as much money in circulation, all else being the same, then each piece of money is ten times less valuable -- that is to say, everything costs ten times as much. Inflation is good for debtors: in effect it enriches them by robbing creditors. Conversely, deflation enriches creditors by robbing debtors. If you have money, you are pleased to see its value go up; if you owe money, you are pleased to see its value go down -- though both parties normally wish that the movement not be great enough to damage the economy in general, as that would injure them in other ways.

In the last century there have been numerous instances of hyperinflation, each one produced by a government that kept on printing more and more paper money. Minor blips of inflation or deflation can have other causes, but when inflation is so severe that bills get used as toilet paper it's never an accident. Typically the reason a government resorts to printing more and more money is that it itself is deeply in debt: it prints money to survive, and even then often doesn't survive long.

Deflation has been almost entirely absent from the world since the Second World War, so to see its evils we need to look farther back. Farmers are usually debtors: they own land against which they can borrow money, and they often need to borrow money in bad years when crops fail (and sometimes even in years when crops succeed too well and there is a surplus, driving prices down). It was largely farmers to whom William Jennings Bryan appealed in 1896 with his speech about not "crucifying mankind on a cross of gold". His goal was to stop the deflation which had been occurring for decades due to the combination of economic growth and a lack of new gold discoveries; his proposed remedy was "free coinage" of silver.

Bryan is commonly disparaged, which in a way is appropriate: anybody who starts a political speech, as Bryan did that one, by admitting that his personal abilities are inferior to his opponent's abilities, is to be taken at his word, even if he doesn't mean you to. Indeed, in a way the proposal was quite modest. "Bimetallism", the use of both gold and silver as currency standards, was the law of the land at that time. But bimetallism has always been an uneasy arrangement: if a dollar is to be equal at the same time to a weight X of silver and to a weight Y of gold, and the ratio of the value of gold to the value of silver changes, what is the dollar to do? It's not like those values can be dictated by the government; they're values set by the world market -- by the interaction of supply and demand all over the globe. The government can peg the dollar to the value of one metal, or at least can try to, but trying to peg it to two values simultaneously is not mathematically possible.

At that time the world price of silver was low enough that a silver dollar, as minted by the government, had less than half a dollar's worth of silver in it. So in practice, though both gold and silver were officially definitions of the dollar, it was gold that really defined it. Silver dollars had value mostly because the government said they had value, as with modern currency. And as with modern currency, the government minted only a limited number of silver coins. "Free coinage of silver" would have made that unlimited: anyone could present any amount of silver metal to the government to be coined into dollars. That would indeed have produced the desired inflationary effect, and also would have exiled gold: with people being able to import silver to the US, get higher prices than world prices for it, and export gold at world prices, the US's stock of gold coins and bullion would have been exported, so the US would have moved from effectively being on the gold standard to effectively being on the silver standard. The huge trade of gold for silver would itself have shifted world prices somewhat, lessening the profits that the exporters of gold would make on the trade, but at that time the US held only a modest share of the world's precious metals, so it wouldn't have lessened them all that much.

Losing money trading gold for silver on those terms would also have made us the chumps of the world, which in itself is enough to explain Wall Street's resistance to "free silver": Wall Street hates to be the chump. But there were other reasons, too, for the resistance. Being on a silver standard would have made commerce with gold-standard countries trickier -- and those were most of the countries of the world, including France, Britain, and Germany. And of course the big Eastern money men were net creditors, the ones who got the benefits of squeezing farmers.

Bryan's effort is also commonly regarded as a failure; but while he did indeed lose the election and in 1900 the nation officially moved to the gold standard, in previous decades the pro-silver movement had succeeded in getting the government to coin quite a lot of silver, to where in 1896 there was almost as much silver money circulating as gold. That silver stayed in circulation even after 1900, with a silver dollar being worth as much as any other dollar; only much later did it become an object worth much more than a dollar for the silver in it and thus disappear from circulation. There was also paper currency circulating at that time, a dollar of which was likewise worth as much as any other dollar; indeed, the increased silver had mostly displaced paper currency rather than increasing the total money supply. By 1896, silver's advocates had gotten tired of the failure of modest silver coinage to stop deflation and were ready for the extreme "free silver" position.

Yet at the time Bryan made his speech, the deflation was already starting to be reversed, partly by new discoveries of gold (the Yukon gold rush being one such), and partly by the invention of the cyanide process for extracting gold from its ore. These things produced a period of "gold inflation", though the increase in prices was only about 2% a year, what the Federal Reserve System tries for today; the rest of the increase in gold supply went into the expansion of the economy.

Friedman and Schwartz's A Monetary History of the United States, from which I am getting the above history of the gold supply, remarks on the pre-1900 era of deflation as showing that deflation doesn't cripple economic growth. At the same time, though, it is famous as the "Gilded Age", the era of the robber barons, the time when big money became bigger and the poor were ground underfoot. A remarkable thing about today's debates over rising inequality is how minor the increase in wealth inequality has been compared to where it was in 1910. This does not mean it is of no concern today, but in 1910 it was stratospheric: the top 1% had nearly 90% of the wealth. This was not because of their larger incomes: today's income inequality is at a similar level to 1910. These weren't the new rich; this was old money -- big money that had indeed become bigger. It was the era of monopolistic trusts -- of Rockefeller the First taking over almost the entire nation's oil refining, plus much of the rest of the oil business. Yet despite much talk to the contrary, the amount of skulduggery used in Rockefeller's takeover was minimal: for the most part he accumulated his empire simply by buying out competitors. Contrary to legend, he rarely tried to squeeze people out of the business by underpricing them; instead he bought them out. From his monopoly profits he could afford it; and once he had bought up the existing competition, new competition found it hard to get started in that tight-money environment, where in addition to the nominal interest rate that borrowers had to pay, the dollars that they returned were worth more than the dollars they borrowed; the real interest rate, instead of being less than the nominal interest rate (as is usual these days), was greater than it.

It is remarkable how respected those corporate empires were at the time. In today's accounts of the Ludlow Massacre, for instance, the company (a mining enterprise, part of the Rockefeller empire) is the villain for hiring thugs who machine-gunned and burned a camp of striking workers and their families. But at the time the public sympathy was with the company, and against the largely-immigrant strikers, "these foreigners who do not intend to make America their home, and who live like rats in order to save money" -- an accusation that has a strange sound to modern ears: there are many things said against immigrants these days, but accusing them of self-denial to save money is not one of them, nor is accusing them of actually wanting to leave. But back then popular sentiment was very different. The investigations into the massacre didn't lead to anyone on the company's side being punished; only strikers were. That in itself is perhaps almost normal, but the popular sympathy went far beyond that:

"In Colorado, Andrews notes, revulsion at the strike violence encouraged political paranoia about immigrants and race, laying the groundwork for the surprising rise of the Ku Klux Klan, which took control of the Statehouse and the governor’s office for a grim time in the 1920s. “Ludlow really brought about a rightward turn,” he says. “It set the stage for the rise of the Klan.”

Now, the strikers had been violent, and became more so after the massacre; that wasn't an entirely baseless revulsion. Still, the reflexive casting of blame on the strikers shows how different that era was; today that sort of worship of power is usually in the service of big government, but back then big corporations were the thing. That's what decades of deflation had done: not only concentrate money, but make people feel that the concentration was natural and proper. Indeed, the big corporations' money let them buy skill and knowledge; they were not contemptible organizations, and people could quite reasonably look up to them even while resenting their dominance.

The backlash from that era gave us the Sherman Antitrust Act, Communism, union laws, and an immense number of regulations on business operations which never previously seemed like they needed regulating, but which in that environment people despaired of ever correcting without government intervention -- and which have often, contrary to their popular image, protected large businesses rather than hindering them, by serving as moats to protect them from new competition. Economists call this "regulatory capture", implying that the regulatory impulse started out good and pure but was then captured by corporations; but Gabriel Kolko looked into it and reported that the regulation of the early 1900s was, from the start, mostly shaped by the regulated businesses, and titled his book The Triumph of Conservatism. Indeed, one might think of it not as part of the backlash but as a continuation of the problem: corporations having grown so powerful they could suborn government to ensure their dominance when tight money stopped being a defense.

Today it's easy to avoid deflation: just print more money. Back in the 1600s it was harder. One way, which Spain used, was to conquer most of the Americas and take their gold and silver. Another way was to introduce banking and let it expand the quantity of money. But that can only go so far: the more the expansion, the greater the risk. Beyond a factor of ten, the risk becomes extreme; and even a factor of ten is questionable.

How best to manage that species of risk is still an open question -- a trillion-dollar open question. The current try at an answer is to federally guarantee bank deposits: there's no spread of panic because depositors never need to panic (unless their deposit account holds more than $250,000; amounts in excess of that are not insured). But this has had the cost of making banks effectively wards of the government. It wouldn't do for someone to be able to set up as a bank, take deposits, make "loans" to his friends and relatives, then fold and let the feds eat all the losses. Instead, the flip side of getting FDIC insurance, for the bank, is tight control over what sorts of loans it can make. Where that has been relaxed, as with "Savings and Loans" banks in the 1980s, there have indeed been "loans" to friends and such -- though that was just the tip of the iceberg of the S&L crisis: most of the $160 billion loss was on investments that were merely risky, not criminal. Still, when it's the feds' money, it's the feds' job to protect it, and they did so, tightening up the rules again -- back to investments which were "as safe as houses". That was the next big bailout: finding out that even housing wasn't necessarily a safe bet -- though it wasn't the bank insurers but the mortgage insurers, "Fannie and Freddie", that took it on the chin that time. Still, as in the original complaint from the 1600s, bankers are being allowed to gamble with other people's money; it's just that it's the feds' money now, and they strictly control what can be gambled on.

That strict control has its downsides. Walter Bagehot, in Lombard Street, boasted that the British banking system was so good at getting money to where it was needed that "We have entirely lost the idea that any undertaking likely to pay, and seen to be likely, can perish for lack of money". Today the banking system is prevented from financing much of the economic activity that goes on. In response, people have turned to other funding sources. Venture capitalists try to satisfy funding needs of new, innovative companies. Leveraged buyout specialists try to satisfy funding needs for revamping old companies. But they charge high for their services. Ordinary bank deposits, the cheapest source of money, aren't free to flow into these areas. There isn't an airtight wall separating them -- banks are allowed a small fraction of risky investments, and money may also flow through respected intermediate companies -- but such routes are long and narrow, and push up the price. Bagehot's world, in which a "new man" could borrow working capital at 5% and displace an established capitalist who used only his own money and insisted on a 10% return on it, is not the world we live in. It's common to think of banking regulations as being technicalities of interest only to specialists; but really the question of what is to be invested in is the question of what direction civilization is to go in, which is of interest to everyone.

The risk squeezed out of the banking system has also migrated to less-regulated parts of finance. In 2008, besides propping up sections of the financial system that were not explicitly guaranteed (as with Fannie and Freddie, where the government had disclaimed responsibility for backstopping them in the event that they went bankrupt, but then did it anyway), the government also propped up sections where there had not been even an implicit guarantee. In the program known as TARP, large investment banks were bailed out. The government later stepped in to guarantee "money market" funds (the business of short-term lending to corporations) after one fund "broke the buck" (had negative returns). They also bailed out the insurer AIG, which had been bankrupted by selling credit default swaps on mortgage-backed securities: in effect AIG had been betting that mortgages would not fail. Unlike in normal banking, where to take the risk of a loan you actually have to make the loan, the invention of credit default swaps let AIG take the risk without putting up much money; it was only one small division of the insurance giant which took the risk, but they lost enough to sink the whole.

After spending several years unwinding the investments they made in the AIG bailout, the government claims to have made money rather than lost it. That claim is also made for other bailouts that were made during the crisis, to the point where one might wonder who actually did lose money. The answers seem to be:

  1. There actually was a liquidity crunch: the initial price shock was an overreaction. It wasn't just prices going down to where they deserved to be; they dropped farther than that. The actual losses on mortgages were (very roughly) $200 billion borne by Fannie and Freddie and perhaps as much borne by others; the swings in market prices were much larger (in the tens of trillions). When market prices are way down, a company to which money is lent in a bailout can easily make a profit. Even if it were headed for bankruptcy before the bailout, by buying things cheap it can profit enough to repay both its old debts and its new bailout loan.

  2. The reason why financiers had been comfortable giving mortgages to people who wouldn't pay them back was that with rising house prices, the mortgage pays off in full even in a foreclosure. As housing prices recovered, this started to be true again. People who were foreclosed-on still lost money, but that was their problem, not the financial system's. Even when the financial system did lose money on a foreclosure, it was less of a loss.

  3. Some bailouts were partly paid off by other bailouts: the initial TARP (thrown over the problem, as it were) seems to have been paid off largely by subsequent aid to buy up the "toxic assets" whose ownership big banks were suffering from. (TALF doesn't actually stand for "Toxic Asset Liberation Front", but that's more to the point than its actual name.) TALF, too, is said to have made a profit; but still, the use of one bailout to help pay off another makes the whole easier to explain.

  4. There were a wide variety of losers. Stockholders in Lehman Brothers lost their entire investment. Stockholders in AIG had their collective share of the company reduced to 20% by dilution of stock, with the government taking the new shares for the other 80% -- which, on the whole, was a very large, viable insurance company (and still is). On top of that, large subdivisions of AIG were sold off to pay off the government bailout loan.

It is common for pundits to talk of the 2008 crisis as being a result of insufficient regulation, as if regulation were something that could be bought by the ton and it was just a question of how much more of it to dump on the financial industry to make it behave. The financial industry, in turn, commonly reacts as the rabbit reacts in the Uncle Remus story: "Please, Br'er Bear, please don't throw me in the briar patch of regulation". It is rare, having been thrown in, that they voice the followup to that: "Br'er Rabbit: born and bred in a briar patch"; but it occasionally happens. Ken Lay, the CEO of Enron, once taunted California regulators, saying that no matter how they modified their rules, he "had people working for him that could figure out a way to make money" by gaming those rules. Tangled thickets of rules were where Enron was born and bred; simple and clear rules, carefully thought out as part of a rigorous system, might have posed an actual threat, but not the complicated briar patch of modern regulation, where any little rule that really does get in the way can in time be chewed through (with patient lobbying), the rest of the thicket serving to conceal the damage.

But as regards carefully thinking out what the rules should be, some modesty is indicated: it seems doubtful that there can be any set of rules that would absolutely eliminate financial crises. Panic is part of human nature; one can no more hope to eliminate panic than to eliminate war. Whenever people are gathered together for any sort of collective effort, if enough of them panic it ruins the effort even for the rest. No complicated economy is needed to have a housing panic: if two people are out in the wilderness building a house together, and one of them suddenly gets into his head that the way they're building it isn't going to work, then the whole project is lost -- provided, of course, that it really does need both of them: perhaps two people are needed to lift the beams, or just one person can't finish the house before winter.

Yet maybe the house really wouldn't work; maybe continuing really is a waste of time. One should not look at the situation from a distance and tell the two, paraphrasing Franklin Delano Roosevelt, that "the only thing you have to fear is fear itself": they also have to fear the house falling on their heads.

It's not just in economic affairs that collective action can be undone by panic. Many an army has fled to escape what would otherwise have been complete destruction, but many others have fled because their enemy put on a good bluff. Even soldiers who saw through the bluff have had to flee: when your comrades are running away, it's useless to try to stand there alone. A well-disciplined military retreats in good order, not in a wild rout, but still may retreat unnecessarily.

Banking panics, likewise, can be well-grounded or not. In the Great Depression, those words of Roosevelt's were apt: the banking panic hadn't been brought on by any big mistake in any particular industry. One does not read, for instance, in histories of cars, that the cars of 1929 were particularly deficient. In 2008, in contrast, a high proportion of the mortgages written were severely deficient, having been issued to people who had little chance of paying them back. And in the early-2020 financial panic the coronavirus lockdowns were damaging not just one sort of business but many sorts.

But whatever their origins, financial panics spread. Some of this is inescapable: less income to any group of people means they buy less from other sectors of the economy. But that's a mild, diffusive sort of spread, somewhat like the spread of heat through an object, where the farther away from the source of heat, the slower that part of the object heats up and the lower the temperature it reaches. To really blow up an economy the spread has to be self-reinforcing, like the progress of a detonation wave through an explosive.

That was the state of the banking system before it became a ward of the government: its combination of deposits that could be withdrawn at any time and long-term loans meant that panic fed on itself. That combination, indeed, seems crazy at first glance. On a small scale it would be utterly unworkable: someone would have to be mad to take deposits from three of his neighbors and loan the money to a fourth for a thirty-year term, when he could expect at any moment to have any of the three demand their money back and when that would bankrupt him. What makes it work at large scale is the law of large numbers: with ten thousand depositors rather than three of them, the flow of money averages out, provided that those depositors do not act in sync but are depositing or withdrawing money for their own personal reasons -- that, in statistical language, deposits and withdrawals from different people are "independent" of each other. Of course they're never entirely independent: banks see more withdrawals for Christmas and more deposits on paydays; but still the flow is smooth enough to be predictable. But when people start doubting the bank's soundness, the assumption of statistical independence goes out the window: people withdraw money not for their own personal reasons but for a reason they all have in common: the bank.

When something bad is self-reinforcing, the earlier and more decisively one moves to put a stop to it the better. A little intervention early on can save having to do a much bigger intervention later. The attitude of Andrew Mellon, Herbert Hoover's Secretary of the Treasury, that the panic was a good way to shake out rotten businesses from the economy was... well, I have made no particular study of Mellon, but there is a stereotypical attitude that "old money" people, like Mellon, have toward "new money" people: they're rude, pushy, and got their money by questionable means -- and if they were to fail, it'd probably be well-deserved. And the 1920s were a time when several decades of tight money, keeping the club exclusive, had been followed by a couple of decades of looser money, letting newcomers in. Someone with the "old money" attitude is inclined to think many more businesses are rotten than actually are, and is not inclined to take the sorts of measures Bagehot praised in Lombard Street:

"The way in which the panic of 1825 was stopped by advancing money has been described in so broad and graphic a way that the passage has become classical. 'We lent it,' said Mr. Harman, on behalf of the Bank of England, 'by every possible means and in modes we had never adopted before; we took in stock on security, we purchased Exchequer bills, we made advances on Exchequer bills, we not only discounted outright, but we made advances on the deposit of bills of exchange to an immense amount, in short, by every possible means consistent with the safety of the Bank, and we were not on some occasions over-nice. Seeing the dreadful state in which the public were, we rendered every assistance in our power.'"

Even with such interventions, the economy still takes a hit. That does indeed expose some rotten businesses -- the 2008 crisis exposed Bernie Madoff -- but it does so even when the intervention comes quickly; it's not a reason to deliberately wait while the banking system continues to implode.

Before the Federal Reserve System was created in 1913, bankers had their own interventions for financial crises. Sometimes banks would gather together and find money to bail out failing firms, as when J. P. Morgan famously gathered together other top bankers and locked them in a room until they came to agreement as to how they would pay for the bailout. (That was in 1907; they liked that so little that they later secretly conspired to have the government take over such duties, which is how we got the Federal Reserve System.) Sometimes things would go beyond where a bailout could work, and there would be "suspension of payment". In its mildest form, banks would get together and announce that while payments could still be made from one bank account to another, cash could not be withdrawn from any bank (except perhaps in small amounts). Suspension did not solve anything in itself, but gave them time to determine which banks had actually failed; then the failed ones would go into bankruptcy while the rest resumed normal business. Other times banks that were under pressure would simply close their doors and not let anyone in. Such expedients often violated normal banking laws, but in emergencies people are excused for such violations -- especially people as reputable as bankers.

There is one sort of law, indeed, which is worth waiving before thinking of any sort of bailout: laws establishing reserve requirements or capital requirements. Even before banks became wards of the state, US banking law had reserve requirements, whereby banks were forbidden from lending out some fraction of their deposits, typically something like 10% or 20%. Those started, indeed, not as matters of law but as conditions banks imposed on each other before they would honor each others' banknotes. When written into law there is a problem with them: the reserves can't actually be used in an emergency, because the law forbids it. A bank has to keep additional reserves, over and above the legally required reserves, to meet its actual requirements for suddenly disgorging cash. That problem is easily solved: in emergencies the authorities can waive the reserve requirements. Indeed, in the early-2020 coronavirus crisis the Federal Reserve for the first time did so: reserve requirements are now at 0%. To take something set aside for use in an emergency and to declare "this is now that emergency" is a completely natural act, but it's taken until now for us to get around to doing it.

Besides reserve requirements, banks and other financial institutions are also under capital requirements. The two requirements cover different risks: reserve requirements cover the risk of a sudden cash drain, while capital requirements cover the risk of loans going bad. To make a $100 million loan, a bank might have to set aside $5 million of its own money on top of $10 million of depositor money. The same logic applies, though: the setaside is to cover emergencies, and when an emergency actually does arise the bank should be permitted to use it. Of course there is the risk of wasting reserves on a minor emergency and then getting hit by a major one, so one shouldn't be too quick on the trigger; but there should be a trigger.

If waiving requirements isn't enough, it's time for active intervention, meaning printing money and injecting it into the economy. (Though I'm using the traditional phrase "printing money", of course in this digital age no actual ink needs to touch any actual paper.) The question is whom the new money is to be given to. I tend to agree with previous writers on the subject: it should be given to me. The only difficulty is that I'm a different "me" than any of the others have been -- and, well, they've usually said "us" rather than "me", shared greed being politically practical whereas individual greed of that magnitude is a joke. Federal Reserve interventions are largely designed to pump money into banks, as befits an entity designed by bankers. Now, they're loans, not gifts; whoever gets the money is expected to repay it. But he gets the money in the depths of a crisis when prices of investments have tanked, so this still gives him an easy profit, provided he is shrewd enough to not sink the money into things that have irredeemably failed.

An excuse for pumping money into banks is that really they are just illiquid, not insolvent: they will eventually be able to repay, but just can't scrape together the cash at the moment. But even an individual can think of himself that way: if, say, someone would give him a loan for a car, he could get a job and pay off the car loan with money left over; there's nothing fundamentally wrong with his finances, he's just temporarily short of money. That could be true or false, but either way the Fed isn't going to give him a loan; he's too small an entity for them to notice.

Even while regretting the favoritism involved in pumping money into a chosen institution, there is a lazy sort of logic to choosing banks. Meat packers, when pumping a curing solution into a ham, often do so by taking a big artery and pumping into it; and banks are the big arteries of finance: they are accustomed to handling vast quantities of money and supplying the rest of the economy with it. And of course (to switch analogies) the banking system was the explosive part of the system, so stopping the explosion was most directly done there. But it's easy to get into "moral hazard" territory, where you're rewarding the people who caused the problem.

Perhaps the most neutral sort of injection of newly-printed money has been one that got much use in the coronavirus financial crisis: buying up government bonds. It's not a subsidy to large banks or any other institution; it just injects cash into the economy in general, via the people who formerly owned those bonds going out in search of other investments. Even to those most jealous of government favoritism, the government buying government bonds should be a "give unto Caesar what is Caesar's" moment.

But there's no free lunch: the availability of this method depends on there being a large national debt, which is not a good thing. In the coronavirus response, the Fed bought more than a trillion dollars of Treasury bonds. For that to be a reasonable project, many more than a trillion dollars' worth must exist, since not all Treasury bonds are on the market at any given moment. The rest can be dragged onto the market by paying a premium, but that starts to get into favoritism again: the people who hold out longer will be rewarded more.

Macaulay, after describing the origins of the national debt in Britain, remarks on it:

"It may be desirable to add a few words touching the way in which the system of funding has affected the interests of the great commonwealth of nations. If it be true that whatever gives to intelligence an advantage over brute force and to honesty an advantage over dishonesty has a tendency to promote the happiness and virtue of our race, it can scarcely be denied that, in the largest view, the effect of this system has been salutary. For it is manifest that all credit depends on two things, on the power of a debtor to pay debts, and on his inclination to pay them. The power of a society to pay debts is proportioned to the progress which that society has made in industry, in commerce, and in all the arts and sciences which flourish under the benignant influence of freedom and of equal law. The inclination of a society to pay debts is proportioned to the degree in which that society respects the obligations of plighted faith. Of the strength which consists in extent of territory and in number of fighting men, a rude despot who knows no law but his own childish fancies and headstrong passions, or a convention of socialists which proclaims all property to be robbery, may have more than falls to the lot of the best and wisest government. But the strength which is derived from the confidence of capitalists such a despot, such a convention, never can possess. That strength,-- and it is a strength which has decided the event of more than one great conflict,--flies, by the law of its nature, from barbarism and fraud, from tyranny and anarchy, to follow civilisation and virtue, liberty and order."

Though there is nothing to disagree with in that, it says nothing in favor of running a national debt in times of peace. Indeed, if a country already has a big national debt built up in peacetime, that impairs its ability to go deeper into debt if a war arises.

Macaulay also writes favorably of government debt as being something that gives people an easy way to stash money where it will earn interest. The flip side of that, though, is that it's the taxpayer who has to pay that interest. Normally plutocrats who want a good interest rate have to work for it by finding something worthwhile to invest in; unless the government itself is doing something just as worthwhile with the money (as it might in a war), it shouldn't purchase their support.

The "Keynesian" sort of intervention for a financial crisis is usually explained in entirely different terms: the economy is bad because people are sort of stuck in a funk; but if the government spends money on something, paying the salaries of one group of people, those people go out and spend that money on other things, which pays the salaries of other groups, and so forth, all of which adds up to a "Keynesian multiplier" that can be multiplied by the government's outlay to calculate the boost to the economy. But while that sequence of consequences is fair enough, it begs the question of where the money the government spends is to come from. If the government increases taxes to get the money, it's pumping money into one group of people while taking it from another group, so the benefit of having one group spend more is counterbalanced by the cost of another group spending less, both groups' spending to be multiplied by more or less the same multiplier. One could argue for taking money away from scrimpers and savers and giving it to spendthrifts, so that the two multipliers would be different, but it's not even clear that would be effective (let alone wise): scrimpers and savers tend to put their money in banks, which lend it out to people who spend it. If instead of taxing the government borrows the money, that borrowing takes money away from other loans that would put the money into circulation. It's only if the money is newly printed that the effect is a pure economic boost. So the Keynesian sort of intervention still depends on printing money; it is just that the new money is spent by the government on -- well, hopefully on something that actually makes sense, rather than on the common economists' example of paying one group to dig holes and another to fill them in.

If the government has nothing that is worth spending more money on, it can use the newly-printed money to give rebates on taxes. In recent crises something like this has been done, in the form of "stimulus payments". This gets outside of what is called Keynesian, but it's a reasonable option for injecting newly-printed money. (Keynes came out with his "General Theory" in 1936, a time when pandering to big government was a ticket to success -- a ticket he rode far on. Giving back money to taxpayers isn't part of the big-government agenda.)

In normal times printing money would be inflationary, but financial panics are deflationary. This is most clearly seen in an old-fashioned banking panic, where the banks have created most of the money in circulation and when they fail that money goes away; but it seems also to extend to modern panics. Perhaps the ideal in a panic is to print just enough money to stabilize the money supply. But even if they hit that target exactly, it still leaves the question of what the newly-created money will do later, after the crisis is over: how and when to suck it back in to avoid inflation. Perhaps the cleanest scenario is if, in the crisis response, banks themselves were the only ones allowed to add to the money supply -- that is, if the government intervention consisted solely of lowering reserve (or capital) requirements. In that case the obvious reversal procedure is to raise the requirements again. Likewise, an intervention by purchasing government debt can be cleanly reversed by selling it.

If an intervention has to go beyond that point, though, and gets into bailouts, the reversal stops being something that can be done cleanly. If the bailouts pay off, the money returns to government coffers and can be destroyed; but if they don't pay off the money stays out there. The government could still fend off inflation by raising banks' reserve requirements above where they had been before the crisis. Indeed, this is a natural response if the panic blew through reserves: that indicates that the old reserve levels aren't sufficient and should be raised. Raising reserve requirements was indeed done in the recovery from the Depression, in 1937 -- which seems to have been too early and to have made things worse again. (Banking regulators spotted that banks were starting to accumulate "excess reserves", but didn't realize that banks, having been bitten hard, considered those "excess" reserves essential, since they never had been allowed to dip into their legally-required reserves.) In the modern world, with federally guaranteed deposits, bank reserve levels don't determine the tendency to panic, but they still could be raised to lower the quantity of money. Alternatively the level of caution could be increased somewhere else in the system, ideally in the place whose lack of caution caused the crisis.

In any case, different crises have different causes, and the details of any bailout must depend on the cause, as should the method of cleaning up after the bailout. A wide-ranging knowledge and a deep understanding of the dynamics of crises are the main tools, not any fixed recipe. It is common and reasonable to distrust politicians and central bankers, but it really is the role of someone with responsibility for the whole economy to act in such situations: to decide what parts of the economy are to be defended and what parts (if any) are to be left to their fate.

One phrase that has been used for the part of the financial system that caused the 2008 crisis is "the shadow banking system". What exactly does and does not constitute "shadow banking" is not well defined, but the central mystery seems to be a form of lending that goes by the name of "repo". The name has nothing to do with "repo" as the word is used outside the financial industry, meaning repossession of cars whose owners have defaulted on their loans, as celebrated in the movie Repo Man. Instead it's a system of "repurchase agreements". ("Repu" perhaps would sound too Indian for financiers' tastes, or perhaps just wouldn't be confusing enough.)

It is easy to find definitions of repo: one party sells something (commonly Treasury bonds) to the other party, with an agreement to buy them back soon (commonly the next day) for a little bit more than the initial purchase price. Although the agreement is just for one day, it's commonly repeated the next day. Usually left mysterious, though, is why any two parties would engage in such a transaction. It can be regarded as a collateralized loan, with the extra money repaid being the interest on the loan; but ordinarily someone who takes out a loan does so because he needs money. Here he has money, in the form of Treasury bonds; he could simply sell them and then buy them back again when done using the money. The market for Treasury bonds is one of the most liquid markets in existence, so why the need for a prearranged repurchase?

The main answer seems to be that the repo borrower is a bank (or another big financial institution), and this is a service to their big corporate customers, who want to deposit money as an average customer would, so as to use it in everyday transactions, but want some sort of insurance against the bank failing. To a big corporation, the $250,000 limit on federal deposit insurance is nothing; they might want the account to hold $50 million. So they demand collateral from the bank; a repo agreement gives it to them.

For its part, the bank's reason for engaging in repo transactions is that the transaction is set up so that although the bank nominally has "sold" the bond for the day, it still gets all the bond's interest payments. From those it pays the repo interest, which is normally lower; so it earns a profit on the difference. In exchange it takes the risk of the bond rising or falling in value -- as bonds do when interest rates change, even if repayment of the bond is guaranteed.

Unamplified, that is not a major risk; but the perhaps 1% difference between those two interest rates is not a return on capital that banks are happy with. So they "lever up": they amplify the risk by repeating the process. They take the money lent to them via repo, buy more bonds, and use those as collateral to borrow more money via repo. Do this ten times, and an unacceptable 1% return turns into an acceptable 10%. But the risk is also multiplied by a factor of ten.

This finally gets us into the sort of situation that's risky enough to deserve a name like "shadow banking", and from a similar cause: money repeatedly going into the bank and then back out of it. Indeed, the parallels are deeper. Taking repo loans is a way of taking deposits, and buying bonds to use in repo is like making loans to the companies that issued those bonds, which might be in the "real economy", using the proceeds of their bond sales for building factories or telecom networks or mines or such. The system doesn't serve all of the "real economy", since a company has to be large to issue bonds; and since the bonds already exist when the repo loan is made, the money doesn't go directly to that company but rather to a middleman in the bond market. But the whole process still, as in conventional banking, funnels money from depositors to real users, with the banking system assuming the risks of those users defaulting.

There is also a parallel to the reserve ratio, which goes by the name of the "haircut": especially if the bond is a risky one, the value of the bond may be greater than the money exchanged for it, the difference between the two being the haircut. The same infinite-geometric-series formula as applies to reserve ratios comes into play: if the bank starts with $100 in bonds, with a 10% haircut the bank only gets $90 in cash. Then when it uses that to buy more bonds and repeat the deal, it is left with $81 in cash, and so forth. So the haircut percentage goes into the formula in the same way the reserve ratio does, and in both cases the output of the formula is the amount of lending / bond buying the bank can do (and also the quantity of money the bank creates).

In 2008, banks were using not just Treasury bonds as repo collateral but also mortgage-backed securities which at the time were regarded as nearly risk-free. When those lost much of their value, banks that had levered-up on repo not only lost money directly (as the bonds' owners), but also had to de-lever quickly. The haircut percentage that repo lenders demanded for mortgage-backed securities rose from hardly anything to in some cases more than 50% -- representing a fear not just that the bonds' values might decline by 50%, but a fear their value might decline by 50% in a single day. Even without that decline in value actually happening, the increased haircuts drastically cut banks' leverage, forcing banks to lower their repo borrowing and sell bonds that had been used as collateral -- in effect contracting the money supply as in a traditional bank run.

From there, though, the parallels start to diverge. In a traditional bank run, the bank has great difficulty selling its loans: they are not at all standardized; determining their value is a matter of individually inspecting them in detail. Bonds are standardized, have something like a market price, and can be sold quickly. But still, when banks are forced to sell them it drives down their price. So in a crisis the suspicions of repo lenders are not focused on one bank or one group of banks, but on one sort of bond or one general class of bonds. It's those bond prices that bear the pressure of a sort of leveraged suspicion where panic feeds on itself. The bank's soundness is is not entirely irrelevant -- it is supposed to honor the repo agreement even if it loses money repurchasing the bond -- but the bond's quality is the main focus.

An even bigger difference to traditional bank runs comes when a bank actually fails; repo lenders to it are not left empty-handed but with bonds which they are entitled to sell to recoup their money. If the bonds' prices were driven down enough, they might take a loss on this -- though they could also choose to hold onto the bonds in hopes that their price recovers. They are not supposed to get a windfall from it, though: if the bond sells for more than they were to originally receive from the repo, they are supposed to return the extra money -- though according to Matt Levine, "in practice we just sue each other forever".

So the overall story is of risk forced out of the traditional banking system popping up again in a different guise. The parallel is strong enough that it even suggests regulatory measures: just as reserve ratios were originally required only by private counterparties, not by the government, repo haircuts (today just a matter of private negotiation) might be mandated by the government in future, in the interest of the stability of the system as a whole.

The repo market does seem rather delicate at present. In late 2019, the Fed intervened in it after repo rates spiked, jumping in as a repo lender (temporary buyer of bonds and provider of money). People argue over the cause of the spike, but the possibilities advanced are relatively minor, which goes to show how little it takes to derange the repo system. In the intervention the Fed was a picky lender, accepting only Treasury bonds or government-guaranteed mortgage-backed securities, so it took on no real risk: for other people to trust the government can represent real risk, but for one part of the government to trust another is a no-op. Its intervention was a parallel to its traditional role of lending money to banks against collateral -- though since banks themselves could just make use of that traditional role and borrow directly from the Fed, the main beneficiaries of this repo lending intervention were hedge funds and other large non-bank financial institutions.

For a repo crisis, it is not necessary to have, as in 2008, some category of bonds whose repayment suddenly looks quite unlikely. Even Treasury bonds could wreak havoc in the repo market if inflation rises, interest rates are raised to match, and values of bonds maturing far in the future fall proportionately.

One thing that was done in response to the 2008 crisis was to raise capital requirements. Capital requirements have the advantage of directly addressing solvency; but that is also their disadvantage: solvency is not easily judged. If the regulating agency has one opinion about whether a loan will go bad, and the bank has another opinion, who is to say which of them is correct? One might think that power will win out -- that the regulating agency's opinion will prevail -- but that ignores the social dynamics. Both regulators and regulated are part of the same social environment: they come from the same schools, read the same news sources, and share the same general worldview. In that environment, regulators are low on the pecking order: they are paid less than the regulated and outnumbered by them. They don't have the collective brainpower to convince the regulated, and are normally unwilling to rule with an iron hand. If one of them does get into his head to throw his weight around, a call to a friendly congressman will usually put him in his place. In the aftermath of a crisis, when the industry is ashamed of itself, they can wield more power, but that eventually dwindles to where their opinions count for little; the only rules they can enforce well are those which can be defined unambiguously.

One much-used way to try to make them unambiguous is to dictate that financial institutions are to "mark to market" their assets, i.e. use market prices to value them. But many loans and other financial assets are not traded on any market, so no market price can be looked up for them. Even with assets that are listed on a market, the trading may be so thin that the market price does not represent their real value. And even when there is a liquid market, market prices may go on a roller coaster ride in a crisis, as they did in 2008, when mark-to-market rules meant that institutions suddenly found themselves having to raise more capital at exactly the worst time to do so. That last problem can be solved by waiving capital requirements in a crisis. But still, they are at best another layer of caution, never a panacea.

Reserve requirements are not just for crises: even in normal times, the reserve ratio is quite a powerful tool for adjusting the quantity of money. Entering a reserve ratio of zero into the formula for an infinite geometric series results in division by zero, meaning the quantity of money becomes infinite. Of course really what happens is that the assumptions break down: setting the minimum legal reserve ratio to zero doesn't actually mean banks will keep zero reserves, and the process of money being deposited and then lent out again doesn't actually continue ad infinitum. But though the formula is at best approximate, it still shows how big a difference a small change in the reserve ratio can make.

The Fed usually doesn't fiddle with reserve ratios to control the quantity of money; instead they use the the other two tools they have for that: setting interest rates (in particular the "discount rate" at which they lend money to banks) and buying and selling Treasury bonds ("open market operations"). In essence, when they buy bonds they are printing money to do so and thus increasing the money supply; conversely, selling bonds decreases it. Likewise, when they lend money to banks, that money is newly created; and the lower the discount rate the more gets lent. (The word "discount" does not refer to anybody getting a bargain, but rather to the process of "discounting", meaning lending money against collateral consisting of other loans.) They have learned how to control inflation quite adeptly using those tools: there is argument over whether the definition of inflation that they use (the Consumer Price Index) is the right one, but certainly they have good control over it and would have equal control over any reasonable alternative definition.

Besides controlling inflation, the Fed is charged with using their control of the money supply to promote economic growth. That order they have essentially ignored (paying only lip service to it), and for good reason. In mathematical terms, they have one variable to control (inflation) and one variable to control it with (the money supply, however altered). That means the solution is completely determined: after controlling inflation there are zero degrees of freedom left with which to control anything else. They can only goose the economy by compromising their inflation goals -- and "goose" is the right word: pumping in more money produces a brief growth spurt, as the recipients of that money buy more; but soon people get tired of working harder and producing more, and raise their prices to get demand back to normal. Then when the money spigot is eventually tightened to stop the inflation, there is a decrease in growth which more or less cancels out the original spurt. Despite complicated economic theories that attempt to deny it, the obvious truth is that the way the economy really grows durably is via the hard work of finding better ways of doing things, accomplished by people all over the nation in a variety of roles. Monetary trickery can't do it.

But to return to the question of how to lower the risks of the banking system, a reserve ratio of 100% would certainly work. That would mean banks couldn't lend out deposits at all. They could still fail: money could go missing via robbery, embezzlement, or just excessive operating expenses. There could still be rumors of covered-up embezzlement, resulting in bank runs, but those couldn't cause healthy banks to fail, only insolvent ones: a healthy bank with 100% reserves would just pay them all back (and probably see them deposited again the next day). With banks not being part of a big interconnected financial system, a panic about one bank wouldn't cause panic about the banking system in general.

There actually was a recent attempt to establish such a bank, to be known as "The Narrow Bank": rather than doing any lending it would deposit all its customers' funds at the Federal Reserve -- which, these days, pays interest on deposits, so The Narrow Bank could even pay its depositors interest. But the Fed quite unsportingly rejected their application for deposit privileges, for reasons that seem to amount to mere jealousy from the usual bankers. (Paying interest to those bankers already is favoritism: it's not like the Fed gets any income from the money so as to justify paying interest.)

Of course eliminating bank loans would be even more restrictive than the current environment, where banks can make government-approved sorts of loans. But if 100% reserves is safe and 10% is unsafe, somewhere in between is "safe enough" -- perhaps 50%. Banks would still fail and depositors would still lose money, but the goal is not to avoid all failures; it is to avoid systemic panic. To avoid an explosion, one doesn't need to remove every atom of explosive; one just needs to dilute it enough with stable atoms that a detonation wave doesn't propagate. Of course there are scenarios in which even 50% would not be enough: global thermonuclear war, for one. But when the world as we know it really is ending, it's hard for the banking system to be the exception. The idea wouldn't be to absolutely eradicate systemic panic, just to turn it into perhaps a once-in-a-century thing rather than the once-in-a-decade thing that it used to be in the 1800s.

Banks would of course protest a 50% reserve requirement, saying that they are really safer than that, and that a lower percentage would let them pay more interest to depositors and serve the needs of borrowers better. But half of the nation's demand deposits would still amount to an enormous source of funding. Directed as banks thought best, it could easily do more good than a larger percentage limited to only government-approved avenues.

Another thing that might be done towards panic-proofing is to eliminate incentives for companies to be in debt. At present, in the US, companies pay corporate tax on profits delivered to shareholders as dividends, but not on interest payments delivered to creditors: those count as business expenses rather than profits. This encourages companies to go deeply into debt, putting them at risk in downturns. Taxing both sorts of payments on an equal footing would eliminate this. When the economy is levered up to its eyeballs, to not bail it out in a crisis is to risk collapse; but being in that state in the first place isn't something to encourage via tax policy.

(There's another tax dodge that companies have thought up lately that is even better: delivering money to shareholders via buybacks of their own stock shares, which increase stock prices. Buybacks are classified as expenses, not as profits, so escape corporate tax, and even at the individual level the gains from increased stock prices are classified as "capital gains", which is taxed at a lower rate than other income. As more companies realize this, it may become universal as a replacement for dividends.)

Much is said in favor of the gold standard as a source of stability -- not in mainstream economic circles, but still. On the gold standard governments can't print money, or at least can't print vast amounts of it. But there can still be inflation: the Spanish discovered that the vast amounts of gold and silver that they'd taken from the Americas didn't buy nearly as much as they'd thought it would, because prices rose. If mining asteroids for precious metals ever really gets going, there may be a comparable episode in the world's future. As for banking panics, as we've seen, those got their start back when not only were precious metals the basis of money but most people personally carried around gold and silver coins. The panics didn't hit the people who actually did keep personal custody of their own gold and silver, but it shows that just saying "gold" or "silver" is not enough: as soon as someone entrusts his gold or silver to someone else's care he's become reliant on promises, so the question has become of the relative advantages and disadvantages of different promises. It's not paper versus metal, but paper versus paper: printed currency versus vault receipts. (Or, these days, electrons versus electrons.)

A hundred years ago, the "gold standard" meant that although paper money was used in daily life along with gold and silver coins, the government would, on request, give anyone gold in exchange for paper money at a fixed exchange rate. But the government could at any moment choose to stop honoring that promise. If someone is going to trust the government with that promise, he can just as well trust it with the promise to not blow out the currency by printing money. Or if he's not going to trust the government, he should keep his gold and silver in his own hands or in the hands of people bound to him by something stronger than just a contract, since contracts are enforced by the government and can be suborned by it. This is not mere theory; in 1933 FDR banned the private ownership of gold coin, gold bullion, and gold certificates, and nullified contracts providing for payment in gold.

The reason for that ban was to devalue the currency, which they proceeded to do after they'd collected the nation's stock of monetary gold. (Gold in jewelry and in industrial use was not covered by the confiscation order.) That was by no means the first time that a precious-metals currency was devalued. One might naively think, for instance, that the British "pound sterling", back in the days when it was defined as a quantity of silver, would refer to a physical pound (the weight measure) of the alloy known as sterling silver (92.5% silver by weight). It indeed started out as a pound, back in Anglo-Saxon times, but through successive debasements ended up as a quarter-pound.

Of course in any such debasement people who personally held on to the old silver coins lost no value. Likewise, in 1933 an estimated $287 million in gold coins were "lost" rather than being surrendered as ordered. That was was about half of the gold coins in circulation at the time. People had a pretty shrewd idea of what was coming, and knew that if they held onto actual physical gold they would find it worth more. Indeed, after the revaluation a $20 "double eagle" coin was worth about $34 -- provided one could manage to sell it without attracting too much attention; the ban on ownership of monetary gold continued even after the devaluation, and was repealed only in 1974. (The ban did, however, allow each person to own up to $100 in face value of gold coins; it was only large stashes that could get someone in trouble, though prosecutions were not frequent even for those.)

Yet even with private gold ownership outlawed, the nation nominally stayed on the gold standard, and really was so for the purposes of foreign commerce. The government would sell gold to foreigners at the new price of $35 per troy ounce, despite not doing so to citizens. It would also buy any gold offered to it at that price from foreigners (or from domestic sources if newly mined). There was much propaganda against private "gold hoarding", as being an economically destructive tying-up of gold that should be in motion -- but of course the way to govern is not by exhorting people to avoid something while giving them incentives to do exactly that. For that matter, "hoarding" gold was what the government itself was doing on an international level: the new $35 price caused an outflow of gold from other nations, worsening the stress on their economic systems until they too were forced to devalue.

The idea of devaluing the dollar was not entirely ridiculous: one of the main problems of the Depression was deflation, and devaluation is a remedy for that. But the nation's money supply at the time was mostly money created by banks' lending: at the start of the Depression, more than nine tenths of it were. Stopping the banking collapse was much more to the point -- and was also done, albeit in old pre-Federal-Reserve style: declaring a national "bank holiday" during which no banking was to be done, quickly examining banks for soundness and (after about a week) letting the obviously-sound ones open while others remained under scrutiny -- the latter amounting to about a third of the total number of banks, holding a seventh of the nation's bank deposits.

In 2002, Ben Bernanke said in a speech

"Let me end my talk by abusing slightly my status as an official representative of the Federal Reserve. I would like to say to Milton and Anna: Regarding the Great Depression. You're right, we did it. We're very sorry. But thanks to you, we won't do it again."

That's right: "we" (the Fed) caused the Great Depression. Reading the full speech, which summarizes arguments made in Friedman and Schwartz's book, leaves no doubt that he meant exactly that: the Fed first sparked the crisis by raising interest rates, causing the stock market crash of 1929, then kept them high, leading eventually to successive banking panics: three years during which they stood idly by rather than using their ample powers to stop the panics, with the interventions they did do being too small and too late.

The Federal Reserve System had been established for the purpose of dealing with banking crises, and this was their maiden effort, the depression of 1920 not having been severe enough to produce a banking crisis. Internally they were busy with a power struggle between the Federal Reserve Board in Washington DC and the Federal Reserve Bank of New York, after the head of the latter (Benjamin Strong) died. But primarily the problem was, as Bernanke put it in the speech, "the generally low level of central banking expertise in the Federal Reserve System" -- and, he might have added, outside it as well; although the Fed's charter somewhat protects it from political meddling, they are somewhat responsive to it and perhaps even more responsive to educated financial opinion. It was educated financial opinion as a whole, not just the Fed, that didn't appreciate that when pushed beyond a certain point the system became inherently unstable and that the only way to restore stability was by prompt, vigorous action.

It's a sort of environment that can easily be appreciated by the sorts of engineers who deal with stability issues. The Wright Brothers' first aircraft were unstable: if left to their own devices, they would quickly try to swap ends and fly backwards. The sort of structure that most aircraft have in the tail, they had in the front. People could still fly those aircraft; it just meant a higher workload for the pilot, who had to constantly keep putting in small corrections on the controls. If he let up for a moment, larger corrections were needed, and if he spaced out for even ten seconds the aicraft might be lost. The Wright Brothers mistakenly thought that instability was necessary for the plane to be steerable. Though they were wrong on that, there was a grain of truth in it; fighter planes have often been designed to be unstable or nearly so, as that gives more agility.

What was being called on from the leaders of the Federal Reserve in the early 1930s was that they shift from being steady, methodical central bankers, carefully scrutinizing all their transactions and taking as little risk as possible, into people with the quickness and aggression of a fighter pilot (albeit on a timescale of days rather than seconds). Only careful forethought and premeditation can enable people to make a shift like that. It has been argued that Strong would have done so, but Strong was dead.

As it was, the banking crisis went on through several waves of failures. The stock market crash of 1929 was only the beginning; though it exposed a lot of speculative excesses (and punished many of the speculators), in itself it was not particularly damaging: that a company's stock price has crashed does not interfere with its normal daily operations. In 1987 there was also a large crash in the stock market (about a 30% decline in the Dow Jones Industrial Average, as compared to about 50% in the crash of 1929), with hardly any economic decline. The main damage in the Depression was done by the banking crises, which only started happening more than a year later. So many banks failed that the money supply was seriously reduced, yielding a rapid deflation: from 1929 to 1933 prices dropped by about one third.

As mentioned, a slow, steady deflation over decades is the sort of thing that concentrates wealth, making big money bigger and crushing upstarts. Sudden deflation is more severe. A deflation of one-third means an increase of 50% in the real value of every debt, of every salary, and of every other promise to pay. (No, not a one-third increase; 2/3 * 3/2 = 1.) Every price has to be changed; every salary and every contract has to be renegotiated -- if it even can be. That number (50%) only became known later; people at the time had to guess at it -- and the vast majority didn't even know, at first, that there was a number that needed guessing; they thought they could continue with business as usual. There had been a similar level of deflation in 1920; but that had just reversed the previous year's inflation, so people had not had time to adjust to the higher price levels before they reverted. But between then and the Depression there had been no inflation; the deflation of the Depression was an unexpected burden, not a welcome correction. The Depression is commonly blamed on companies getting overconfident and overproducing until their warehouses were filled with unsold goods. But businessmen are constantly trying to find each others' flaws and blind spots, in order to outcompete each other; it is most implausible that all of them, not just in one industry but in all sorts of different industries, got overconfident at once. Really the problem must have been the shortage of money, not a surplus of goods for sale.

Many of the other features of the Depression also bear the marks of deflation. Farmers with mortgages foreclosed in huge numbers; many businesses that operated with debt went bankrupt. (Businesses that had avoided debt, like General Motors, often did fine: GM made a profit all through the Depression, though it had to lay off many workers.) These bankruptcies and foreclosures -- from the banks' point of view, bad loans -- further increased the banks' problems, making still more of them fail and increasing the public distrust of them still more.

That was not the end of it: the deflation spread internationally via the gold standard, which acted as an international common currency. For the first two years there were massive inflows of gold into the US, lessening the deflation here by spreading it abroad. Britain saw the train wreck coming, and went off the gold standard in 1931. But before that, and elsewhere even after that, deflation abroad led to bank failures abroad, leading to still more deflation -- so much so that gold began going abroad again. The misery greased Hitler's way to power: there was indeed money mischief going on, and in the absence of a good explanation it was easy for him to blame "the Jews", with their long history of being moneylenders (and of being scapegoats). The total damage done is thus incalculable.

So it's regrettable that such an influential book as this did not give a good explanation of banking risks. By any normal standard of judging historical writing, of course, it is no criticism at all to object that a history book did not help in a crisis that occurred eight decades after it was written. But Macaulay was aiming for a much higher standard, and generally hit it; this was a rare miss.