How Wind Power Froze Texas

How Wind Power Froze Texas

In the wake of February’s tragic power outages in Texas, during which 4.5 million households suffered service interruptions, partisans on both sides have been quick to interpret the events as confirmation of their preferred energy policies. With news images of helicopters deicing frozen turbines, conservatives lambasted Texas’s increasing reliance on wind power as the villain in the story.

Trying to temper this knee-jerk reaction, columnist Ron Bailey argued that “[m]ost of the shortfall in electric power generation during the current cold snap is the result of natural gas and coal powered plants going offline.” And Paul Krugman for his part declared that it was a “malicious falsehood” to blame wind and solar power for what happened in Texas, as it was primarily a failure of natural gas.

In this article I’ll lay out the basic facts of which power sources stepped up to the plate during the crisis. Contrary to what you would have known from reading Ron Bailey (let alone Paul Krugman), when the Texas freeze hit, electricity from natural gas skyrocketed while wind output fell off a cliff. The people arguing that wind wasn’t to blame mean it in the same way Jimmy Olson wasn’t to blame when General Zod took over: wind is so useless nobody serious ever thought it might help in a crisis.

Krugman on Texas Electricity

In his February 18 column titled “Texas, Land of Wind and Lies,” Krugman declared that

Republican politicians and right-wing media … have coalesced around a malicious falsehood instead: the claim that wind and solar power caused the collapse of the Texas power grid, and that radical environmentalists are somehow responsible for the fact that millions of people are freezing in the dark …

In contrast to this dirty rotten lie from the right-wingers, Krugman instead explains:

A power grid poorly prepared to deal with extreme cold suffered multiple points of failure. The biggest problems appear to have come in the delivery of natural gas, which normally supplies most of the state’s winter electricity, as wellheads and pipelines froze.

A bit later in the article Krugman admits that wind was involved as well, but minimizes its role in this way:

It’s true that the state generates a lot of electricity from wind, although it’s a small fraction of the total. But that’s not because Texas—Texas!—is run by environmental crazies. It’s because these days wind turbines are a cost-effective energy source wherever there’s a lot of wind, and one thing Texas has is a lot of wind.

It’s also true that extreme cold forced some of the state’s insufficiently winterized wind turbines to shut down, but this was happening to Texas energy sources across the board, with the worst problems involving natural gas.

Incidentally, there are literally no numbers in Krugman’s article (except for numerals referring to dates), which is a signal that he’s pulling a fast one on his readers. From his qualitative (not quantitative) description, most people would have assumed that when the unusually cold weather hit Texas last month, electricity generation from various sources was down across the board, but that it mostly fell from natural gas, while the drop in wind was insignificant. As I’ll show in the next section, this is utterly false.

What Really Happened During Texas’s Power Crisis

Had I not seen the analysis from my former colleagues at the Institute for Energy Research (see their articles here and here), I might have believed the spin that the Texas crisis was really a failure of fossil fuels rather than renewables. Yet as we’ll see, the actual numbers tell a much different story from what most Americans probably “learned” from the media discussion.

The simplest way for me to communicate the relevant information is through three infographics, generated from the Energy Information Administration’s handy tool that shows the source mix for daily energy generation by state.

Before showing the numbers, I need to make an important clarification: the demand for electricity soared to unprecedented levels during the freeze. In particular, on February 14, peak demand on the electric grid surpassed sixty-nine gigawatts, breaking the previous winter record of (almost) sixty-six gigawatts set in 2018. It was in the early hours of the following morning (February 15) that the Electric Reliability Council of Texas (ERCOT) implemented rolling blackouts to prevent the entire grid from collapsing. So to be clear, the issue wasn’t that supply in an absolute sense fell, but rather that demand soared. (Texas typically uses more electricity in the summer to keep things cool, rather than in the winter to keep things warm.)

With that context in place, here are the stats for electricity output from various sources on February 15, 2021:

texas power feb 15 2021

Already we see something interesting. Of the total amount of electricity delivered on this first day of blackouts, 65 percent came from natural gas, while only 6 percent came from wind and 2 percent from solar.

But in fairness, maybe what guys like Krugman meant is that this is much lower than what we normally could expect from natural gas. (Remember Krugman had said that natural gas “normally supplies most of the state’s winter electricity.”)

To test this possibility, we can look at the situation one year prior, on February 15, 2020:

texas power feb 15 2020

Now, this is interesting. A year earlier, during a normal mid-February day, natural gas “only” supplied 43 percent of the total electricity, whereas wind accounted for 28 percent and solar was the same at 2 percent. Remember how Krugman said wind was only a “small fraction” of Texas generation? Overall for the year 2020, wind produced 22 percent of Texas’s electricity, a higher share than coal.

Yet besides the proportions, also look at the absolute quantity of electricity generated: on Feb. 15, 2020, natural gas produced 398,130 megawatt hours (compared to 759,708 MWh during the recent freeze), while wind produced 264,024 MWh (compared to 73,395 MWh during the freeze).

To sum up, compared with the same date a year earlier, during the first day of the blackouts in Texas, electricity from natural gas was 91 percent higher, while electricity from wind was 72 percent lower.

To reiterate the clarification I gave earlier, part of the confusion here is that electricity demand in February isn’t normally as high as it was because of the freeze. So to test whether natural gas is the culprit, we can compare the generation from various sources during the freeze to the situation back during the summer. For example, let’s look at how things stood on August 15, 2020:

texas power august 15 2020

As our date occurred in the dog days of summer, total electric demand was higher in mid-August 2020 than on February 15, 2021. Furthermore, output from every source was lower during the freeze when compared with their performance the prior August 15. However, it seems odd to single out natural gas as the culprit, when it experienced the lowest percentage drop, and (on all dates) was the single biggest source.

The following table summarizes electrical output from various sources on the three dates we have analyzed, and shows the change going from the earlier dates to the first day of the recent blackouts:

texas table

As the table indicates, on all three dates natural gas was always the leader in electrical generation. During the freeze, it produced 91 percent more than it had the prior year during a more typical winter day. And although natural gas produced less electricity during the freeze than it had during the peak summer demand, it was only a 7 percent drop.

In contrast, wind power during the freeze was down a whopping 72 percent compared to the previous year, and compared to the summer it was down 47 percent.

Among all sources, the percentage difference between either the previous year or the previous summer was highest for natural gas. That is, the surge in natural gas output year over year was the biggest by far (with coal coming in second with a 54 percent surge), and compared with the summer load its drop was the smallest at 7 percent.

Wind, in contrast, was the worst performer in both cases, if we measure in terms of the difference. That is, wind’s 72 percent drop in the year-over-year column was the biggest one, and its 47 percent drop in the column for summer to winter was also the biggest.

In light of these statistics, it’s a bit odd for commentators to blame the Texas blackouts on natural gas while excusing wind.

What They Mean: Wind Is the Ted Cruz of Electricity

Now, in fairness, what the commentators blaming natural gas have in mind is that ERCOT’s emergency planning assumed that natural gas (and the other “thermal” electricity sources, namely coal and nuclear) could be called upon to fill the gap should there be record demand during a winter storm. If we measure in terms of the total capacity that was temporarily knocked out because of the freeze, then the culprits were thermal sources, rather than wind and solar.

As Jesse Jenkins, an assistant professor at Princeton tweeted out, “Main story continues to be the failure of … natural gas, coal, and nuclear plants … which ERCOT counts on to be there when needed.” He further specified, “Of about 70,000 MW of thermal plants in ERCOT, ~25–30,000 MW have been out since Sunday night. Huge problem.”

And so we see what people mean when they say the Texas blackouts are the fault of natural gas, rather than wind: since no serious official ever expected wind to be any help during a crisis, it can hardly be blamed for not showing up when disaster struck. In effect, Krugman is arguing that wind power is the Ted Cruz of electricity.


When assessing blame for a disaster, it’s hard to know what the relevant counterfactual should be. Yes, had the (relatively) unregulated Texas power providers done a better job in winterizing their natural gas lines, things would have been better last February.

But by the same token, had the federal government never implemented the wind production tax credit (PTC)—which subsidizes wind so heavily that it sometimes sells for a negative price in the Texas wholesale market—then there would have been more fossil fuel-generated capacity in Texas, which the numbers clearly show did better at providing electricity during the deep freeze. Normally the boosters of renewable energy point with pride to Texas, which has the most wind capacity of any state by far in absolute terms, and even has almost 25 percent of its official generating capacity consisting of wind. Yet when wind collapsed during the deep freeze, suddenly even its biggest fans admit that nobody ever thought it could do the same job as natural gas.

This article was originally featured at the Ludwig von Mises Institute and is republished with permission.

Can Libertarians Oppose Short Selling?

Can Libertarians Oppose Short Selling?

Amid the controversy over GameStop, many cynics argued that something sinister was clearly afoot because the hedge funds had shorted 138 percent of the outstanding shares. In this article I’ll review that particular claim, as well as another seemingly dubious practice, so-called naked short selling. My conclusion is that shorting more than the total outstanding shares isn’t perverse or fraudulent, whereas naked short selling—depending on the context—might be.

A Review of Short Selling

Before diving into the specific variants, let me first explain the basics of short selling and why it can be a healthy activity in a free market. (In this section I reproduce material I published in an earlier article.)

In general, those who speculate in the stock market provide a “social” service—if they profit—by steering asset prices to their correct levels, as I explain in this article. If investors believe a particular stock is underpriced, they can buy shares of it and then unload them once the stock has met or surpassed what they view as its “proper” level. In this way, the speculators push up the (initially) underpriced stocks, helping to correct the “error” in the original price structure. Notice that it doesn’t matter whether the investors who notice the initial underpricing own any of the stock at the outset.

However, things are different when an investor believes a stock is overpriced. If the investor happens to own shares of the stock in question, the obvious move is to sell some or all of the position, which both earns a relative gain for the investor and also speeds the downward move in price.

If this were the end of the story, there would be an obvious asymmetry in the market’s ability to rely on the dispersed knowledge of experts in diverse fields. Namely, it would mean that the only people who could act on their belief that a stock is overpriced would be those who already own the stock.

Fortunately, the market allows short selling, where someone who has zero shares of a stock can, in effect, sell shares and end up holding a negative position. In other words, a person “going long” might end up holding a hundred shares of a stock, whereas someone “going short” might end up holding minus a hundred shares.

Suppose stock XYZ is trading at $50, but Sam the Speculator believes tomorrow’s news will contain something very unfavorable for the stock. Further suppose that the rest of the market isn’t seeing things the way Sam is. Sam can borrow, say, a hundred shares of XYZ from Harold the Shareholder, sell them today for $5,000, and then wait for the news to hit. When it does, and the stock sinks to $40, Sam buys back the hundred shares for $4,000, and returns them to Harold (along with a fixed fee). Harold is better off, because he earned the fee; the price drop would have hit him in any case. (If Harold wants to unload his XYZ stock as the price drops but before Sam returns his shares, then Harold himself can short a hundred new shares and end up no worse off than he otherwise would have been.)1For example, suppose the bad news hits (as Sam anticipated but Harold did not), and the price of XYZ begins falling over the course of a few hours from $50 down to $40. If Harold originally would have closed out his position once XYZ fell to (say) $45, then he can still borrow and short a hundred shares once XYZ hits that threshold. Selling the borrowed shares at that price garners him $4,500, and then Harold gives the hundred shares plus the fee to the person from whom he borrowed the shares, once Sam closes out the original short position. When the dust settles, Harold ends up with $4,500, which is exactly what he would have had in an alternate universe where he never met Sam and instructed his broker to sell his shares of XYZ if they ever fell to $45. Meanwhile, Sam has netted $1,000 (minus the fee) from his superior foresight.

Things get more complicated when the short seller takes longer to return the shares; we have to worry about dividend payments, interest, etc. But the basic principle is simple enough: just as a speculator who wants to go long can borrow money to buy stocks, so too a speculator who wants to go short can borrow stocks to “buy money.” Short selling is no more mysterious than buying stocks on margin.

How Can There Be More Shorts Than Total Shares?

Now that we’ve reviewed the basic framework of short selling, we can answer the question: How can it be possible to short more shares than exist? The answer is pretty straightforward: the same share can be shorted multiple times as it passes hands from one owner to the next.

For example, suppose there are a total of a thousand shares of XYZ stock. Ten different speculators each want to short a hundred shares each. So each of the ten speculators makes arrangements with the original holders of XYZ stock in order to borrow a thousand total shares from them to sell at the original market price.

Now when a speculator shorts a stock, he is selling it. That means there must be a buyer of the shares on the other side of the transaction. When a new person enters the market to buy shares of a stock, there isn’t a Post-it note affixed to some of the shares saying “This share was obtained as part of a short sale, so it is forbidden to be lent out again.” No, all the shares are fungible, and each one just as “shortable” as any other. When a speculator borrows shares of XYZ in order to sell them (with the hopes that the price will go down), the person on the other side of the transaction is engaging in a normal purchase. That new owner acquires full rights to the shares, including the right to lend them out to a new short seller.

Therefore, continuing our scenario, suppose that after the original thousand shares of XYZ have been borrowed as part of ten different short sales the price of XYZ is still too high, in the view of a speculator. This speculator approaches the new owners of the stock, and borrows 380 shares from them to short. At this point, the total (gross) short position is 1,380 shares, even though there are only a thousand total shares in existence. In other words, the gross short position is 138 percent of the outstanding shares.

To understand the situation, consider the gross vs. net positions: there are currently a thousand shares owned outright by various individuals. There are also people who are owed 1,380 shares in total, because they lent out their shares for a short sale; they are “long” the XYZ stock, because they benefit if its price goes up. On the flip side of this, there are the short sellers who collectively owe 1,380 shares to the people from whom they borrowed the stock. (They are clearly “short” the stock, and benefit if its price drops.) Therefore, the gross short position is 1,380 shares, which is 138 percent of the total supply, but the gross long position is 2,380 shares. Thus, the net position is long 2,380 – 1,380 = 1,000 shares, which is 100 percent of the total quantity of shares. Nothing weird here, folks.

Before leaving this section, let me address one potential objection. Isn’t there a sense in which it’s physically impossible for the short sellers to close out their positions? After all, how can they collectively buy back 1,380 shares when there are only a thousand shares total?

The answer is that they don’t have to do this in one fell swoop. (This is also why it’s possible for a community to pay interest plus principal on money loans.) One of the original speculators can buy a hundred shares of XYZ and return them to the original owners, thus reducing the gross short position to 1,280 shares. Then a different speculator could buy another hundred shares and do the same, reducing the shorts to 1,180 shares, and so on. The cumulative short position can be unwound in stages, just as it was originally wound up in stages.

Now it’s true, the speculators might find it expensive to convince the current holders of the XYZ stock to sell some of their shares in order to close out the original short. But this is true regardless of the scale of the operation. In other words, whether the gross short position on XYZ is 138 percent or 1 percent, if the current holders hang on tight to their shares, it might be prohibitively expensive for the shorters to close out their positions. There is nothing peculiar to a gross short position being greater than 100 percent in this regard.

A More Dubious Practice: Naked Short Selling

With “naked” short selling, the transactions are marked up “on paper” without first locating and obtaining actual shares of the stock. To adapt our scenario, if a speculator wants to short a hundred shares of XYZ when it’s selling for $50, perhaps his brokerage will allow him to do so even though it hasn’t gotten possession of a hundred shares. When the speculator closes out his position, the brokerage calculates what his profit/loss would have been if he really had obtained the shares and credits/debits his account accordingly.

The potential problem here occurs if something goes wrong and the speculator can’t close out his original position because XYZ has gone up in price too much. For example, suppose that after shorting the hundred shares at $50 (and thus temporarily earning $5,000), the speculator is horrified to see that XYZ shoots up to $1,000 per share! In order to close out his position, the speculator would need to spend $1,000 x 100 = $100,000 to obtain the hundred shares on the open market and return them to the brokerage. If the speculator doesn’t have that kind of money, he can’t do it.

The victims in this type of scenario are the people who thought they were buying those hundred shares of XYZ back when it was selling for $50. (After all, that’s where the original $5,000 came from.) For small enough dollar amounts, the brokerage itself would cover the speculator’s loss. But in principle, if a brokerage allowed many of its clients to engage in naked short selling, then a surprising spike in stock prices could cause such aggregate losses that even the brokerage itself couldn’t afford to cover them. In that case, investors who thought they were buying shares of stock from the brokerage would discover that even though it had taken their money and told them the transaction went through, they in fact did not own actual shares.

Readers will note the similarity between naked short selling and fractional reserve banking. It’s true in both cases that contracts could carefully spell out the details, and give the brokerage/bank the legal right to refuse redemption, but there is a line in Austrian economics (going from Mises through Rothbard to many of today’s Rothbardians) which argues that legal niceties don’t really matter, it’s commercial practice that does. If bank depositors treat demand deposits as immediately redeemable (and hence equivalent to money proper), then fractional reserve banking causes the boom-bust cycle. Likewise, one might argue as an Austrian that naked short selling is a dubious practice that could foster instability in the financial markets. (Note that my own position on this latter issue has evolved since I previously wrote on naked shorting for


In a free market, short selling is a healthy practice that allows farsighted speculators to push down overpriced stocks. There is nothing inherently dubious or fraudulent should the gross short position exceed the total number of shares. However, the practice of naked short selling could be problematic, if it became large relative to overall trading.

This article was originally featured at the Ludwig von Mises Institute and is republished with permission.

F.A. Hayek’s Conception of Private, Fiat Money

F.A. Hayek’s Conception of Private, Fiat Money

The most famous Austrian economist is 1974 Nobel laureate Friedrich Hayek. Because of his moderate views excusing state interventions in various circumstances, hardcore Rothbardians tend to regard Hayek as less than pure in many areas.

However, one area where Hayek is certainly more radical (though perhaps not correct!) than even Murray Rothbard is monetary institutions, as detailed in his fascinating (1978) pamphlet The Denationalisation of Money.

When it comes to the free market’s handling of money, the typical Austrian argument is over fractional reserve banking (FRB). Some think FRB is perfectly legitimate (so long as the banks do not receive special privileges from the government), while others consider it inherently fraudulent. But both groups agree that fiat money is a horrible creation of the state and that the free market would always settle on a commodity (such as gold) as the underlying base money.

Inasmuch as many of the participants in the FRB debate are far more radical than Hayek on most policy issues, it is quite surprising then that Hayek’s proposal calls for privately issued, competing fiat currencies. That is, Hayek proposes that individual firms issue pieces of paper that are not backed by any production or consumption good. In a sense, Hayek wants to privatize central banking.

As the reader can imagine, this proposal strikes almost everyone—even modern Austrians—as absurd; we will deal with some of the major objections below. But partly because of this near-unanimous rejection, and partly because the analysis in any case is instructive, I will attempt in this article to give Hayek’s case the best possible defense.

Hayek’s Proposal

Hayek argues that, if only government obstacles were removed, the free market would provide the optimal quantity (and variety!) of monetary products. Just as the forces of competition lead to low prices and superior quality in every other line, so too would competition in the “fiat money industry” lead to monies that were infinitely better than their government-produced counterparts. For example, the private monies would be far more stable in their purchasing power, would be harder to counterfeit, and would be available in more convenient denominations.

Although one can imagine an equilibrium situation given that the public is already holding vast quantities of such private currencies, it is difficult to conceive of how they would “get off the ground” in the first place. Here is the most ingenious part of Hayek’s proposal (which naturally I am adapting for a modern exposition):

A private firm could initially print up, say, 1 million pieces of paper (that of course would be difficult for an outsider to reproduce) with a cute picture of Friedrich on them. The firm then contractually pledges to redeem each “Hayek,” at any time, for either $10 or 80 Chinese yuan. Assuming that the firm has substantial assets and that everyone is fully confident of their redeemability, the Hayeks at auction will sell for somewhat more than $10. This is because they will always be worth at least $10, but they might (in the not too distant future) be worth more, if and when the Chinese government lets the yuan appreciate against the dollar. (In that case, investors could redeem each Hayek for ¥80, which would exchange for more than $10.) For the sake of argument, let’s suppose that the firm initially auctions all 1 million Hayeks for $12 each.

Thus far the proposal involves nothing too radical; each Hayek is really just a derivative asset. How, then, would the issuing firm get the public to start treating the Hayeks as money? On the night of the initial auction, after the market price of the Hayeks had been ascertained, the issuing firm would specify a commodity basket (consisting of bread, eggs, milk, and other goods relevant to consumers) that cost, say, $60 at Wal-Mart. Then the firm would announce to the public the following nonbinding pledge: “We will use our firm’s assets to adjust the outstanding supply of Hayeks such that 5 Hayeks will always (insofar as it is humanly possible) have the purchasing power to buy this specified commodity basket.”

Now, as time went on, the US dollar and the Chinese yuan would depreciate vis-à-vis real goods and services. In particular, the dollar price of the specified commodity basket would increase. So long as the Hayeks were still being valued solely because of their tie to dollars and yuan, their value as well would begin to drop; the Hayek price of the commodity basket would start to rise from 5 to 5.05, etc.

At this point the issuing firm would need to prop up the value of its fiat currency. It would need to enter the market and buy back Hayeks from those marginal holders who were most anxious to sell. In this way, the issuing firm could (at least temporarily) maintain the purchasing power of the Hayeks such that 5 Hayeks could still buy the relevant commodity basket at Wal-Mart, even though the dollar price of that basket has risen above $60 (as the US government continued to print new dollars).

Here is where the theory ends and we are stuck with an empirical question: Would the firm eventually buy back all 1 million of the Hayeks? Or, at some point before this happened, would the record of stability of the Hayek (in terms of its purchasing power vis-à-vis the specified commodity basket) allow for a self-fulfilling prophecy in which people begin holding Hayeks not because of the underlying legal redeemability, but because of its expected purchasing power in the future?


Hayek’s proposal was understandably treated with suspicion. Murray Rothbard1 argued that it violated Mises’s “regression theorem,” which demonstrated that all money—even government fiat currency—must ultimately derive its purchasing power from a historical tie to a commodity that was valued in a state of barter. However, this objection overlooks the fact that Hayek’s proposal does contain an initial link to an underlying asset in order to get off the ground.

Rothbard also objects that not all government functions should be privatized, in particular tax collection, torture of prisoners, and the issuance of fiat currency. The point may be conceded, but Hayek’s proposal would certainly be legally permissible in a libertarian society. Even those who consider fractional reserve banking as fraudulent could find no violation of property rights in Hayek’s proposal;2 they would simply have to argue (and a compelling argument it is!) that any firm attempting to circulate its own fiat currency would go bankrupt.

A different problem is that, in the world Hayek envisions, there would be no single money, and hence the benefits of a common medium of exchange would be curtailed. To this I would respond that it is possible that even under a 100 percent commodity standard, some groups use gold, others use silver, and others use cows as a medium of exchange.

Yes, there would be forces tending to promote the emergence of a single money throughout the entire world, but this would not be instantaneous, as conditions differ greatly from region to region. So long as each of the local monies could be freely exchanged against one another, modern currency markets (aided by computers) would significantly reduce the transactions costs involved. By the same token, we cannot say that the benefits of a single money outweigh all other considerations and that therefore Hayek’s system must be rejected.

Another objection (raised by Selgin and White) is that a private “central bank” would, just as its government counterpart, always find it most profitable to hyperinflate. It is true that this would cause the public to abandon the currency, but so what? If 5 Hayeks currently exchange for so many eggs, milk, etc., why not print up 2 billion of them and buy as many real goods as possible? Surely this one-shot move will earn more than the present discounted value of responsible management of the supply of Hayeks.

This fear overlooks the fact that the Hayeks (in our example) are always legally redeemable for $10 or ¥80. That places a floor below which their value cannot sink (without draining the reserves of the issuing firm).

Pete Canning acknowledges this fact and refines the objection by pointing out that, eventually, the government currencies will have depreciated so much that this check will soon be impotent. Ironically, here is where another of the alleged deficiencies—namely the multiplicity of currencies—comes to the rescue. Precisely because each issuing firm will only provide the money held by a fraction of the public, one firm’s decision to hyperinflate would not be nearly as disastrous as when a monopoly government does so.

Moreover, if a major firm ever did decide to hyperinflate, the public would demand measures to prevent a recurrence. For example, in addition to pledging to redeem Hayeks at any time for $10 or ¥80, our hypothetical firm might also legally pledge “We will never increase the supply of Hayeks by more than 100 percent per year.”3


Let me close by pointing out some of the overlooked benefits of Hayek’s scheme. First, in principle privately issued fiat currencies could prove more stable than even commodity metals in terms of their purchasing power. The whole job of the firm issuing Hayeks (in our example) is to closely monitor the financial markets to fine tune the exchange value of the Hayeks, such that five of them always purchase the specified commodity basket at a major grocery store. This is not true when it comes to gold; the exchange rate between gold and the commodity basket would be far more volatile (though of course much more stable than the exchange rate between government currencies and the basket).

Another benefit is that the firms could change the composition of the commodity basket to reflect the preferences of the holders of their monies. For example, some people may not care about the price of eggs and bread, and would prefer a money that had stable purchasing power in terms of a basket of aluminum, platinum, etc. A firm could fill that niche.

Another interesting feature of Hayek’s system is that holders of money would themselves reap the advantages of inflation of the currency, rather than the issuing firm. Consider: if the public ever did accept Hayeks (and Lachmanns etc.) as media of exchange, over time the market would increase the production of eggs, butter, etc., and hence there would be a tendency for their Hayek price to fall. Therefore, in order to maintain the stated purchasing power, the issuing firm would need to print up and distribute additional Hayeks periodically.

Now if the firm were a monopoly, naturally its owners would spend the new Hayeks themselves. But because of competition, the firm can only keep the public using Hayeks if, in addition to the incredibly stable purchasing power, holders of Hayeks receive new units in proportion to their holdings. That is, the firm would have to periodically increase the supply of Hayeks at large in order to maintain a constant purchasing power, but it would need to give the new units to its customers. (An easy way to achieve this would be for the firm to also act as banker and pay dividends on deposits.)

Finally—and I admit this is quite fanciful—suppose that in the distant future, humans develop the Star Trek capacity to reproduce (within limits) any type of physical item. In that case, no commodity could serve as a useful medium of exchange, because people would simply mass produce it at virtually no cost. In such a world, money would probably become mere numbers on computers.

Yes, if governments were expected to responsibly run such a system, all would be lost. But it is at least worth exploring whether a system based on Hayek’s ideas could provide sound media of exchange in that futuristic environment.

Robert P. Murphy is a Senior Fellow with the Mises Institute. He is the author of many books. His latest is Contra Krugman: Smashing the Errors of America’s Most Famous Keynesian. His other works include Chaos TheoryLessons for the Young Economist, and Choice: Cooperation, Enterprise, and Human Action (Independent Institute, 2015) which is a modern distillation of the essentials of Mises’s thought for the layperson. Murphy is cohost, with Tom Woods, of the popular podcast Contra Krugman, which is a weekly refutation of Paul Krugman’s New York Times column. He is also host of The Bob Murphy Show. This article was originally featured at the Ludwig von Mises Institute and is republished with permission.

A Rebuke of ‘Modern Monetary Theory’

A Rebuke of ‘Modern Monetary Theory’

[Review of Stephanie Kelton, The Deficit Myth: Modern Monetary Theory and the Birth of the People’s Economy (New York: PublicAffairs, 2020).]

I’ve got good news and bad news. The good news is that Stephanie Kelton—economics professor at Stony Brook and advisor to the 2016 Bernie Sanders campaign—has written a book on modern monetary theory (MMT) that is very readable and will strike many readers as persuasive and clever. The bad news is that Stephanie Kelton has written a book on MMT that is very readable and will strike many readers as persuasive and clever.

To illustrate the flavor of the book, we can review Kelton’s reminiscences of serving as chief economist for the Democratic staff on the US Senate Budget Committee. When she was first selected, journalists reported that Senator Sanders had hired a “deficit owl”—a new term Kelton had coined. Unlike a deficit hawk or a deficit dove, Kelton’s deficit owl was “a good mascot for MMT because people associate owls with wisdom and also because owls’ ability to rotate their heads nearly 360 degrees would allow them to look at deficits from a different perspective” (Kelton 2020, p. 76).

Soon after joining the Budget Committee, Kelton the deficit owl played a game with the staffers. She would first ask if they would wave a magic wand that had the power to eliminate the national debt. They all said yes. Then Kelton would ask, “Suppose that wand had the power to rid the world of US Treasuries. Would you wave it?” This question—even though it was equivalent to asking to wipe out the national debt—“drew puzzled looks, furrowed brows, and pensive expressions. Eventually, everyone would decide against waving the wand” (Kelton 2020, p. 77).

Such is the spirit of Kelton’s book, The Deficit Myth. She takes the reader down trains of thought that turn conventional wisdom about federal budget deficits on its head. Kelton makes absurd claims that the reader will think surely can’t be true…but then she seems to justify them by appealing to accounting tautologies. And because she uses apt analogies and relevant anecdotes, Kelton is able to keep the book moving despite its dry subject matter. She promises the reader that MMT opens up grand new possibilities for the federal government to help the unemployed, the uninsured, and even the planet itself…if we would only open our minds to a paradigm shift.

So why is this bad news? Because Kelton’s concrete policy proposals would be an absolute disaster. Her message can be boiled down to two sentences (and these are my words, not an exact quotation): Because the Federal Reserve has the legal ability to print an unlimited number of dollars, we should stop worrying about how the government will “pay for” the various spending programs the public desiresIf they print too much money we will experience high inflation, but Uncle Sam doesn’t need to worry about “finding the money” the same way a household or business does.

This is an incredibly dangerous message to be injecting into the American discourse. If it were mere inflationism, we could hope that enough of the public and the policy wonks would rely on their common sense to reject it. Yet because Kelton dresses up her message with equations and thought experiments, she may end up convincing an alarming number of readers that MMT really can turn unaffordable government boondoggles into sensible investments, just by changing the way we think about them.

Precisely because Kelton’s book is so unexpectedly impressive, I would urge longstanding critics of MMT to resist the urge to dismiss it with ridicule. Although it’s fun to lambaste “magical monetary theory” on social media and to ask, “Why don’t you move to Zimbabwe?” such moves will only serve to enhance the credibility of MMT in the eyes of those who are receptive to it. Consequently, in this review I will craft a lengthy critique that takes Kelton quite seriously in order to show the readers just how wrong her message actually is, despite its apparent sophistication and even charm.

Monetary Sovereignty

In her introductory chapter, Kelton lures the reader with the promise of MMT and also sheds light on her book title:

[W]hat if the federal budget is fundamentally different than your household budget? What if I showed you that the deficit bogeyman isn’t real? What if I could convince you that we can have an economy that puts people and planet first? That finding the money to do this is not the problem? (Kelton 2020, p. 2, bold added)

The first chapter of the book makes the fundamental distinction for MMT, between currency issuers and currency users. Our political discourse is plagued, according to Kelton, with the fallacy of treating currency issuers like Uncle Sam as if they were mere currency users, like you, me, and Walmart.

We mere currency users have to worry about financing our spending; we need to come up with the money—and this includes borrowing from others—before we can buy something. In complete contrast, a currency issuer has no such constraints, and needn’t worry about revenue when deciding which projects to fund.

Actually, the situation is a bit more nuanced. To truly reap the advantages unlocked by MMT, a government must enjoy monetary sovereignty. For this, being a currency issuer is a necessary but insufficient condition. There are two other conditions, as Kelton explains:

To take full advantage of the special powers that accrue to the currency issuer, countries need to do more than just grant themselves the exclusive right to issue the currency. It’s also important that they don’t promise to convert their currency into something they could run out of (e.g. gold or some other country’s currency). And they need to refrain from borrowing…in a currency that isn’t their own. When a country issues its own nonconvertible (fiat) currency and only borrows in its own currency, that country has attained monetary sovereignty. Countries with monetary sovereignty, then, don’t have to manage their budgets as a household would. They can use their currency-issuing capacity to pursue policies aimed a maintaining a full employment economy. (Kelton 2020, pp. 18–19, bold added)

Countries with a “high degree of monetary sovereignty” include “the US, Japan, the UK, Australia, Canada, and many more” (Kelton 2020, p. 19). (And notice that even these countries weren’t “sovereign” back in the days of the gold standard, because they had to be careful in issuing currency lest they run out of gold.) In contrast, countries like Greece and France today are not monetarily sovereign, because they no longer issue the drachma and franc but instead adopted the euro as their currency.

The insistence on countries issuing debt in their own currency helps to explain away awkward cases such as Venezuela, which is suffering from hyperinflation and yet has the ability to issue its own currency. The answer (from an MMT perspective) is that Venezuela had a large proportion of its foreign-held debt denominated in US dollars, rather than the bolivar, and hence the Venezuelan government couldn’t simply print its way out of the hole. In contrast, goes the MMT argument, the US government owes its debts in US dollars, and so never need worry about a fiscal crisis.

Yes, Kelton Knows about Inflation

At this stage of the argument, the obvious retort for any postpubescent reader will be, “But what about inflation?!” And here’s where the critic of MMT needs to be careful. Kelton repeatedly stresses throughout her book—and I’ve seen her do it in interviews and even on Twitter—that printing money is not a source of unlimited real wealth. She (and Warren Mosler too, as he explained when I interviewed him on my podcast) understands and warns her readers that if the federal government prints too many dollars in a vain attempt to fund too many programs, then the economy will hit its genuine resource constraint, resulting in rapidly rising prices. As Kelton puts it:

Can we just print our way to prosperity? Absolutely not! MMT is not a free lunch. There are very real limits, and failing to identify—and respect—those limits could bring great harm. MMT is about distinguishing the real limits from the self-imposed constraints that we have the power to change. (Kelton 2020, p. 37, bold added)

In other words, when someone like Alexandria Ocasio-Cortez proposes a Green New Deal, from an MMT perspective the relevant questions are not, “Can the Congress afford such an expensive project? Will it drown us in red ink? Are we saddling our grandchildren with a huge credit card bill?” Rather, the relevant questions are, “Is there enough slack in the economy to implement a Green New Deal without reducing other types of output? If we approve this spending, will the new demand largely absorb workers from the ranks of the unemployed? Or will it siphon workers away from existing jobs by bidding up wages?”

The Fundamental Problem with MMT

Now that we’ve set the table, we can succinctly state the fundamental problem with Kelton’s vision: regardless of what happens to the “price level,” monetary inflation transfers real resources away from the private sector and into the hands of political officials. If a government project is deemed unaffordable according to conventional accounting, then it should also be denied funding via the printing press.

What makes MMT “cool” is that it’s (allegedly) based on a fresh insight showing how all of the mainstream economists and bean counters are locked in old habits of thought. Why, these fuddy-duddies keep treating Uncle Sam like a giant corporation, which has to make ends meet and always has to satisfy the bottom line. In contrast, the MMTers understand that the feds can print as many dollars as they want. It’s not revenue but (price) inflation that limits the government’s spending capacity.

I hate to break it to Kelton and the other MMT gurus, but economists—particularly those in the free market tradition—have been teaching this for decades (and perhaps centuries). For example, here’s Murray Rothbard in his 1962 treatise, Man, Economy, and State:

At this time, let us emphasize the important point that government cannot be in any way a fountain of resources; all that it spends, all that it distributes in largesse, it must first acquire in revenue, i.e., it must first extract from the “private sector.” The great bulk of the revenues of government, the very nub of its power and its essence, is taxation, to which we turn in the next section. Another method is inflation, the creation of new money, which we shall discuss further below. A third method is borrowing from the public. (Rothbard 1962, pp. 913–14, bold added)

To repeat, this is standard fare in the lore of free market economics. After explaining that government spending programs merely return resources to the private sector that had previously been taken from it, the economist will inform the public that there are three methods by which this taking occurs: taxation, borrowing, and inflation. The economist will often add that government borrowing can be considered merely deferred taxation, while inflation is merely hidden taxation.

And it’s not merely that inflation is equivalent to taxation. Because it’s harder for the public to understand what’s happening when government money printing makes them poorer, there is a definite sense in which standard taxation is “honest” whereas inflation is insidious. This is why Ludwig von Mises considered inflationary finance to be “essentially antidemocratic”: the printing press allows the government to get away with spending that the public would never agree to explicitly pay for through straightforward tax hikes.

Kelton and other MMT theorists argue that inflation isn’t a problem right now in the US and other advanced economies and so we don’t need to be shy about cranking up the printing press. But whether or not the Consumer Price Index is rising at an “unacceptably” high rate, it is a simple fact that when the government prints an extra $1 million to finance spending, then prices (quoted in US dollars) are higher than they otherwise would have been, and people holding dollar-denominated assets are poorer than they otherwise would have been. Suppose that prices would have fallen in the absence of government money printing. In this case, everybody holding dollar assets would have seen their real wealth go up because of the price deflation. If the government merely prints enough new dollars to keep prices stable, it’s still the case that those original dollar holders end up poorer relative to what otherwise would have happened.

Now to be sure, Kelton and other MMT theorists would object at this point in my argument. They claim that if there is still some “slack” in the economy, in the sense of unemployed workers and factories operating below capacity, then a burst of monetary inflation can put those idle resources to work. Even though the rising prices lead to redistribution, if total output is higher, then per capita output must be higher too. So, on average, the people still benefit from the inflation, right?

On this score, we simply have a disagreement about how the economy works, and in this dispute I think the Austrians are right while the MMTers are wrong. According to Mises’s theory of the business cycle, “idle capacity” in the economy doesn’t just fall out of the sky, but is instead the result of the malinvestments made during the preceding boom. So if we follow Kelton’s advice and crank up the printing press in an attempt to put those unemployed resources back to work, it will simply set in motion another unsustainable boom/bust cycle. In any event, in the real world, government projects financed by inflation won’t merely draw on resources that are currently idle, but will also siphon at least some workers and raw materials out of other, private sector outlets, as I elaborate in this article.

In summary, the fundamental “insight” of MMT—namely, that governments issuing fiat currencies need only fear price inflation, not insolvency—is something that other economists have acknowledged for decades. Where the MMTers do say something different is when they claim that printing money only carries an opportunity cost when the economy is at full employment. But on this point, the MMTers—like their more orthodox cousins the Keynesians—are simply wrong.

Tough Questions for MMT

A standard rhetorical move is for proponents to claim that MMT isn’t ideological, but merely describes how a financial system based on fiat money actually works. (For example, this was the lead argument Mike Norman used when he and I were dueling with YouTube videos.) Yet since so much hinges on whether a government has “monetary sovereignty,” it’s amazing that the MMTers never seem to ask why some governments enjoy this status while others don’t.

For her part, Kelton criticizes certain nonmonetarily sovereign governments for particular actions, such as joining a currency union (Kelton 2020, p. 145), but she doesn’t ask the basic question: Once an MMT economist explains its benefits, why doesn’t every government on earth follow the criteria for becoming a monetary sovereign? Indeed, why don’t all of us as individuals issue our own paper notes—in my case, I’d print RPMs, which has a nice ring to it—and furthermore only borrow from lenders in our own personal currencies? That way, if you fell behind in your mortgage payments, you could simply print up more of your own personal notes to get current with the bank.

Posed in this way, these questions have obvious answers. The reason Greece adopted the euro, and why Venezuela borrows so much in US dollar–denominated debt, and the reason I use dollars rather than conducting transactions in RPMs, is that the rest of the financial community is very leery of the Greek drachma, the Venezuelan bolivar, or the Murphyian RPM note. Consequently, the Greek and Venezualan governments, as well as me personally, all subordinated our technical freedom to be “monetary sovereigns” and violated one or more of Kelton’s criteria.

In short, the reason most governments (including state governments in the US) in the world aren’t “monetary sovereigns” is that members of the financial community are worried that they would abuse a printing press. The Greek government knew its economy would receive more investment, and that it would be able to borrow on cheaper terms, if it abandoned the drachma and adopted the euro. The Venezuelan government knew it could obtain much larger “real” loans if they were denominated in a relatively hard currency like the USD rather than the Venezuelan currency, which could so readily be debased (as history has shown). And I personally can’t interest anybody in financial transactions involving my authentic RPM notes, and so, reluctantly, I have to join the dollar zone.

Now that we’ve covered this basic terrain, I have a follow-up question for the MMT camp: What would it take for a government to lose its monetary sovereignty? In other words, of those governments that are currently monetary sovereigns, what would have to happen in order for the governments to start borrowing in foreign currencies, or for them to tie their own currency to a redemption pledge, or even abandon their own currency and embrace one issued by a foreign entity?

Here again the answer is clear: a government that engaged too recklessly in monetary inflation—thus leading investors to shun that particular “sovereign” currency—would be forced to pursue one or more of these concessions in order to remain part of the global financial community. Ironically, current monetary sovereigns would run the risk of forfeiting their coveted status if they actually followed Stephanie Kelton’s policy advice.

MMT Is Actually Wrong about Money

For a framework that prides itself on neutrally describing the actual operation of money and banking since the world abandoned the gold standard, it’s awkward that MMT is simply wrong about money. In this section I’ll summarize three of the main errors Kelton makes about money.

Money Mistake #1: Contrary to MMT, the Treasury Needs Revenue before It Can Spend

A bedrock claim of the MMT camp is that unlike you, me, and Walmart, the US Treasury doesn’t need to have money before spending it. Here’s an example of Kelton laying out the MMT description of government financing:

Take military spending. In 2019, the House and Senate passed legislation that increased the military budget, approving $716 billion…There was no debate about how to pay for the spending…Instead, Congress committed to spending money it did not have. It can do that because of its special power over the US dollar. Once Congress authorizes the spending, agencies like the Department of Defense are given permission to enter into contracts with companies like Boeing, Lockheed Martin, and so on. To provision itself with F-35 fighters, the US Treasury instructs its bank, the Federal Reserve, to carry out the payment on its behalf. The Fed does this by marking up the numbers in Lockheed’s bank account. Congress doesn’t need to “find the money” to spend it. It needs to find the votes! Once it has the votes, it can authorize the spending. The rest is just accounting. As the checks go out, the Federal Reserve clears the payments by crediting the sellers’ account with the appropriate number of digital dollars, known as bank reserves. That’s why MMT sometimes describes the Fed as the scorekeeper for the dollar. The scorekeeper can’t run out of points. (Kelton 2020, p. 29, bold added)

For a more rigorous, technical treatment, the advanced readers can consult Kelton’s peer-reviewed journal article from the late 1990s on the same issues. Yet whether we rely on Kelton’s pop book or her technical article, the problem for the MMTers is still there: nothing in their description is unique to the US Treasury.

For example, when I write a personal check for $100 to Jim Smith, who also uses my bank, we could explain what happens like this: “Murphy instructed Bank of America to simply add 100 digital dollars to the account of Jim Smith.” Notice that this description is exactly the same thing that Kelton said about the Treasury buying military hardware in the block quotation above. (It’s true that Bank of America can’t create legal tender base money the way the Fed can; I plug that hole in the analogy a bit below with my Goldman Sachs example.)

Now of course, I can’t spend an unlimited amount of dollars, since I’m a currency user, not a monetary sovereign. In particular, if I “instruct” Bank of America to mark up Jim Smith’s checking account balance by more dollars than I have in my own checking account, the bank may ignore my instructions. Or, if my overdraft isn’t too large, the bank might go ahead and honor the transaction but then show that I have a negative balance (and charge me an insufficient funds fee on top of it).

The only difference between my situation and the US Treasury’s is that I actually have bounced checks and online payments before, whereas the US Treasury hasn’t. Indeed, Kelton’s own journal article shows that the Treasury consistently maintained (as of the time of her research) a checking account balance of around $5 billion and that the daily closing amount never dipped much below this level (Kelton 1998, p. 11, figure 4).

Indeed, the Treasury itself sure acts as if it needed revenue before it can spend. That’s why the Treasury secretary engages in all sorts of fancy maneuvers—such as postponing contributions to government employees’ retirement plans—whenever there’s a debt ceiling standoff and Uncle Sam hits a cash crunch.

The MMTers take it for granted that if the Treasury ever actually tried to spend more than it contained in its Fed checking account balance, the Fed would honor the request. Maybe it would, and maybe it wouldn’t; CNBC’s John Carney (who moderated the debate at Columbia University between MMT godfather Warren Mosler and yours truly) thinks it’s an open question in terms of the actual legal requirements, though Carney believes that in practice the Fed would go ahead and cash the check.

Yet, to reiterate, so far the Treasury has never tried to spend money that it didn’t already have sitting in its checking account. The MMT camp would have you believe that there is something special occurring day in and day out when it comes to Treasury spending, but they are simply mistaken: so far, at least, the Treasury has never dared the Fed by overdrawing its account.

Indeed, Kelton herself in her technical article from the late 1990s implicitly gives away the game when she defends the MMT worldview in this fashion:

[S]ince the government’s balance sheet can be considered on a consolidated basis, given by the sum of the Treasury’s and Federal Reserve’s balance sheets with offsetting assets and liabilities simply canceling one another out…the sale of bonds by the Treasury to the Fed is simply an internal accounting operation, providing the government with a self-constructed spendable balance. Although self-imposed constraints may prevent the Treasury from creating all of its deposits in this way, there is no real limit on its ability to do so. (Kelton 1998, p. 16, italics in original)

What Kelton writes here is true, but by the same token we can consider the Federal Reserve and Goldman Sachs balance sheets on a consolidated basis. If we do that, then Goldman Sachs can now spend an infinite amount of money. Sure, its accountants might still construct profit and loss statements and warn about bad investments, but these are self-imposed constraints; so long as the Fed in practice will honor any check Goldman Sachs writes, then all overdrafts are automatically covered by an internal loan from the Fed to the investment bank. The only reason this wouldn’t work is if the Fed actually stood up to Goldman and said no. But that’s exactly what the situation is with respect to the Treasury too.

Whenever I argue the merits of MMT, I debate whether or not to bring up this particular quibble, because wondering whether the Fed would actually cover a Treasury overdraft doesn’t get to the essence of what’s wrong with MMT. I’m actually sympathetic to the MMT claims that the Fed would be obligated to backstop the Treasury in all circumstances; it would be very naïve to think that the Fed actually enjoys “independence” from the federal government that grants the central bank its power. Furthermore, I believe that the various rounds of quantitative easing (QE) during the Obama years weren’t merely driven by a desire to minimize the output gap, but instead were necessary to help monetize the boatloads of new federal debt being issued. (Of course Trump and Powell are performing a similar dance.)

Even so, I think it’s important for the public to realize that the heroes of MMT are misleading them when they claim there is something unique to Uncle Sam in the way he interacts with his banker. So far, this is technically not the case. Even when the Fed has clearly been monetizing new debt issuance—such as during the world wars—all of the players involved technically have gone through the motions of having the Treasury first float bonds in order to fill its coffers with borrowed funds and only then spending the money. The innocent reader wouldn’t know this if he or she relied on the standard MMT accounts of how the world works.

Money Mistake #2: Contrary to MMT, Taxes Don’t Prop Up (Most) Currencies

Another central mistake in the MMT approach is its theory of the origin and value of money. (If you want to see the Austrian view, consult my article on the contributions of Menger and Mises.) To set the stage, here is Kelton explaining how Warren Mosler stumbled upon the worldview that would eventually be dubbed modern monetary theory:

Mosler is considered the father of MMT because he brought these ideas to a handful of us in the 1990s. He says…it just struck him after his years of experience working in financial markets. He was used to thinking in terms of debits and credits because he had been trading financial instruments and watching funds transfer between bank accounts. One day, he started to think about where all those dollars must have originally come from. It occurred to him that before the government could subtract (debit) any dollars away from us, it must first add (credit) them. He reasoned that spending must have come first, otherwise where would anyone have gotten the dollars they needed to pay the tax? (Kelton 2020, p. 24)

This MMT understanding ties in with its view of the origin of money and how taxes give money its value. Kelton explains by continuing to summarize what she learned from Mosler:

[A] currency-issuing government wants something real, not something monetary. It’s not our tax money the government wants. It’s our time. To get us to produce things for the state, the government invents taxes…This isn’t the explanation you’ll find in most economics textbooks, where a superficial story about money being invented to overcome the inefficiencies associated with bartering…is preferred. In that story, money is just a convenient device that sprang up organically as a way to make trade more efficient. Although students are taught that barter was once omnipresent, a sort of natural state of being, scholars of the ancient world have found little evidence that societies were ever organized around barter exchange.

MMT rejects the ahistorical barter narrative, drawing instead on an extensive body of scholarship known as chartalism, which shows that taxes were the vehicle that allowed ancient rulers and early nation-states to introduce their own currencies, which only later circulated as a medium of exchange among private individuals. From inception, the tax liability creates people looking for paid work…in the government’s currency. The government…then spends its currency into existence, giving people access to the tokens they need to settle their obligations to the state. Obviously, no one can pay the tax until the government first supplies its tokens. As a simple point of logic, Mosler explained that most of us had the sequencing wrong. Taxpayers weren’t funding the government; the government was funding the taxpayers. (Kelton 2020, pp. 26–27, bold added)

I have included these lengthy quotations to be sure the reader understands the superficial appeal of MMT. Isn’t that intriguing—Mosler argues that the government funds the taxpayers! And when you think through his simple point about debits and credits, it seems that he isn’t just probably correct, but that he must be correct.

Again, it’s a tidy little demonstration. The only problem is that it’s demonstrably false. It is simply not true that dollars were invented when some autocratic ruler out of the blue imposed taxes on a subject population, payable only in this new unit called “dollar.” The MMT explanation of where money comes from doesn’t apply to the dollar, the euro, the yen, the pound…Come to think of it, I don’t believe the MMT explanation applies even to a single currency issued by a monetary sovereign. All of the countries that currently enjoy monetary sovereignty have built their economic strength and goodwill with investors by relying on a history of hard money.

In a review of Kelton’s book, I’m not going to delve into the problems with the alleged anthropological evidence that purportedly shows that ancient civilizations used money that was invented by political fiat, rather than money that emerged spontaneously from trade in commodities. For that topic, I refer the interested reader to my review of David Graeber’s book.

Yet let me mention before leaving this subsection that the MMT story at best only explains why a currency has a nonzero value; it doesn’t explain the actual amount of its purchasing power. For example, if the IRS declares that every US citizen must pay $1,000 in a poll tax each year, then it’s true, US citizens will need to obtain the requisite number of dollars. But they could do so whether the average wage rate were $10 per hour or $10,000 per hour, and whether a loaf of bread cost $1 or $1,000.

Furthermore, other things equal, if the government lowers tax rates, then it strengthens the currency. That’s surely part of the reason that the US dollar rose some 50 percent against other currencies after the tax rate reductions in the early Reagan years. So the MMT claim that taxes are necessary, not to raise revenue (we have a printing press for that), but to prop up the value of the currency, is at best seriously misleading.

Money Mistake #3: MMT Confuses Debt with Money

Amazingly, even though their system claims to explain how money works, the MMTers apparently don’t know the simple difference between money and debt. Here’s Kelton trying to defuse hysteria over the national debt:

The truth is, we’re fine. The debt clock on West 43rd Street simply displays a historical record of how many dollars the federal government has added to people’s pockets without subtracting (taxing) them away. Those dollars are being saved in the form of US Treasuries. If you’re lucky enough to own some, congratulations! They’re part of your wealth. While others may refer to it as a debt clock, it’s really a US dollar savings clock. (Kelton 2020, pp. 78–79, bold added)

The part I’ve put in bold in the quotation above is simply wrong. And I don’t mean, “It’s wrong according to Austrian economics but right according to MMT.” No, even in the MMT framework, Kelton’s claim about the national debt is wrong. The outstanding federal debt would only correspond to “how many dollars [have been] added to people’s pockets without subtracting…them away” to the extent that the Federal Reserve had monetized the debt by taking the Treasury securities onto its own balance sheet. But to the extent that some of the outstanding Treasury debt is currently held by individuals and entities that aren’t the Federal Reserve, Kelton’s statement is simply wrong.

In the MMT framework, federal government spending creates new dollars, while taxing destroys them. Since a federal budget deficit refers to a situation where Uncle Sam spends more than he taxes, it’s understandable why Kelton concluded that the federal debt—which reflects the cumulative history of the net budget deficits and surpluses over time—is equal to the net number of dollars that Uncle Sam “spent into existence.”

But to repeat, this is wrong. Kelton forgot that when the Treasury floats new bonds, that action (in the MMT framework) also destroys dollars by removing them from the hands of the public. So if all of the outstanding Treasury debt were held by the public (or foreign central banks), then the cumulative federal budget deficits wouldn’t correspond to any net dollar creation, even in the MMT framework.

Stay with me; we have one more step: in the MMT framework (and the Austrian framework too, for that matter), when the Federal Reserve buys outstanding Treasury securities in the secondary market and takes them onto its balance sheet, this creates new dollars. Therefore, to the extent that the outstanding Treasury securities are sitting on the Fed’s balance sheet, then that portion of the national debt would correspond to “how many dollars [have been] added to people’s pockets without subtracting…them away.”

Does the reader see how cumbersome the MMT framework is? It led its chief proponent to make an elementary mistake in her attempt to explain the basics to the public. In contrast, coming from an orthodox background, I immediately knew Kelton’s claim was wrong, because borrowing money per se doesn’t create money. This is true whether corporations do it or whether Uncle Sam does it. (Just imagine $1 billion in actual currency and that the Treasury keeps issuing new $1 billion bonds to keep borrowing that same pile of green pieces of paper to continually respend them. This procedure would run up the national debt as much as we want, but at any moment there would still be the same $1 billion in currency.)

To drive home just how confused Kelton is on the difference between US Treasurys and US dollars, later in the book she writes, “Heck, I don’t even think we should be referring to the sale of US Treasuries as borrowing or labeling the securities themselves as the national debt. It just confuses the issue and causes unnecessary grief” (Kelton 2020, p. 81).

Here’s another way I can demonstrate that Kelton’s discussion is obviously missing something: if Kelton were right and the US national debt were a tally of how many dollars on net the government has “spent into existence,” then when Andrew Jackson paid off the national debt, the American people would have had no money—the last dollar would have been destroyed. And yet even Kelton doesn’t claim that dollars were temporarily banished from planet Earth, she merely claims that Jackson’s policy caused a depression. (For the Austrian take on this historical episode, see this article.)

For an even starker illustration of the MMT confusion between debt and money, consider Kelton’s approving quotations of a thought experiment from Eric Lonergan, who asked, “What if Japan monetized 100% of outstanding JGBs [Japanese government bonds]?” That is, What if the Bank of Japan issued new money in order to buy up every last Japanese government bond on earth? Lonergan argues that “nothing would change,” because the private sector’s wealth would be the same; the BOJ will have engaged in a mere asset swap. In fact, because their interest income would now be lower while their wealth would be the same, people in the private sector would spend less after the total debt monetization (!), according to Lonergan.

In response to these claims, I make a simple point: you can’t spend Japanese government bonds in the grocery store. That’s why money and debt are different things. If Lonergan were correct, then we could also go the other way: specifically, if the Japanese government issued enough bonds to absorb every last yen on planet Earth, then apparently Lonergan would have to say that aggregate demand measured in yen would go through the roof. Yet how could it, if nobody held any yen anymore? Remember, you can’t pay your rent or buy groceries with government bonds.

Do Government Deficits = Private Savings?

In chapter 4, Kelton lays out the MMT case that government deficits, far from “crowding out” private sector saving, actually are the sole source of net private assets. Using simple accounting tautologies, Kelton seems to demonstrate that the only way the nongovernment sector can run a fiscal surplus is if the government sector runs a fiscal deficit.

Going the other way, when the government is “responsible” by running a budget surplus and starts paying down its debt, by sheer accounting we see that this must be reducing net financial assets held by the private sector. (This is why it should come as no surprise, Kelton argues, that every major government surplus led to a bad recession [Kelton 2020, p. 96].)

In the present review, I won’t carefully review and critique this particular argument, as I’ve done so in this article. Suffice it to say that you could replace “government” in the MMT argument with any other entity and achieve the same outcome. For example, if Google borrows $10 million by issuing corporate bonds and then it spends the money, then the net financial assets held by The World Except Google go up by precisely $10 million. (Or rather, the way you define terms in order to make these claims true is the same way Kelton gets the MMT claims about Uncle Sam to go through.) So did I just prove something really important about Google’s finances?

Obviously something is screwy here. Using standard definitions, people in the private sector can save, and even accumulate net financial wealth, without considering the government sector at all. (I spell all of this out in this article.) For example, Robinson Crusoe on his deserted island can “save” out of his coconut income in order to finance his investment of future labor hours into a boat and net. Even if we insist on a modern financial context, individuals can acquire shares of equity in new corporations, thus acquiring assets that don’t correspond to a “debit” of anyone else.

It is a contrived and seriously misleading use of terminology when MMT proponents argue that government deficits are a source of financial wealth for the private sector. Forget the accounting and look at the big picture: even if the central bank creates a new $10 million and simply hands it to Jim Smith for free, it hasn’t made the community $10 million richer—except in the nominal sense in which we could all be “millionaires” with this practice. Mere money creation doesn’t make any more houses or cars or acres of arable farmland available. Printing new money doesn’t make the community richer. At best it’s a wash with redistribution, and in fact in practice it makes the community poorer by distorting the ability of prices to guide economic decisions.

The MMT Job Guarantee

The last item I wish to discuss is the MMT job guarantee. Strictly speaking, this proposal is distinct from the general MMT framework, but in practice I believe every major MMT theorist endorses some version of it.

Under Kelton’s proposal, the federal government would have a standing offer to employ any worker at $15 per hour (Kelton 2020, p. 68). This would set a floor against all other jobs; Kelton likens it to the Federal Reserve setting the federal funds rate, which then becomes the base rate for every other interest rate in the economy.

Kelton argues that her proposal would eliminate the unnecessary slack in our economic system, where millions of workers languish in involuntary unemployment. Furthermore, she claims that her job guarantee would raise the long-term productivity of the workforce and even help people find better private sector job placement. This is because currently “Employers just don’t want to take a chance on hiring someone who has no recent employment record” (ibid., p. 68).

There are several problems with this proposal. First of all, why does Kelton assume that it would only draw workers out of the ranks of the unemployed? For example, suppose Kelton set the pay at $100 per hour. Surely even she could see the problem here, right? Workers would be siphoned out of productive private sector employment and into the government realm, providing dubious service at best at the direction of political officials.

Second, why would employers be keen on hiring someone who has spent, say, the last three years working in the guaranteed job sector? These would be, by design, the cushiest jobs in America. Kelton admits this when she says that the base wage rate would be the floor for all other jobs.

Looking at it another way, it’s not really a job guarantee if it’s difficult to maintain the position. In other words, if the people running the federal jobs program are allowed to fire employees who show up drunk or who are simply awful workers, then it’s no longer a guarantee.


Stephanie Kelton’s new book The Deficit Myth does a very good job explaining MMT to new readers. I must admit that I was pleasantly surprised at how many different topics Kelton could discuss from a new view, in a manner that was simultaneously absurd and yet apparently compelling.

The problem is that Kelton’s fun book is utterly wrong. The boring suits with their standard accounting are correct: it actually costs something when the government spends money. The fact that since 1971 we have had an unfettered printing press doesn’t give us more options, it merely gives the Fed greater license to cause boom/bust cycles and redistribute wealth to politically connected insiders.

Robert P. Murphy is a Senior Fellow with the Mises Institute. He is the author of many books. His latest is Contra Krugman: Smashing the Errors of America’s Most Famous Keynesian. His other works include Chaos TheoryLessons for the Young Economist, and Choice: Cooperation, Enterprise, and Human Action (Independent Institute, 2015) which is a modern distillation of the essentials of Mises’s thought for the layperson. Murphy is cohost, with Tom Woods, of the popular podcast Contra Krugman, which is a weekly refutation of Paul Krugman’s New York Times column. He is also host of The Bob Murphy Show. This article was originally featured at the Ludwig von Mises Institute and is republished with permission. 

The Fed’s Massive Injection of ‘Liquidity’ Also Benefits Uncle Sam

The Fed’s Massive Injection of ‘Liquidity’ Also Benefits Uncle Sam

There’s a lot to be said regarding the Fed’s surprise announcements—including its Sunday surprise of $700 billion in renewed QE and the complete elimination of all reserve requirements for banks—but here let me just focus on one element: the tendency for Fed officials and all the pundits to treat injections of “liquidity” as if they don’t count as much when distorting the economy. I’ve seen some analysts literally call the Fed’s repo operations “free” as opposed to fiscal policy, which they agree actually costs something.

These distinctions are phony. The Fed’s $1.5 trillion was a “handout” in the same way that a Pentagon fighter jet contract is a handout to a defense contractor. However, the defenders of the Fed are correct that financial institutions per se are not reaping extraordinary gains from the new policy. No, the primary beneficiaries of the Fed’s recent announcements are the holders of US Treasury debt (which includesinvestment banks, of course), as well as the US Treasury, which, after all, is the institution that issues new Treasury debt. The Fed’s massive wave of intended dollar-creation is designed to keep the liquidity of US government-issued debt close to par with US government-issued money.

Just think slowly through what the repo market is: it’s a market where firms sell their Treasury (or other very safe collateral) securities in exchange for actual dollars but also agree contractually to “repurchase” (hence the name) these Treasurys after a certain time. Now, there is an inbuilt (slight) difference in the sell/buy price, allowing the implicit lender of money to earn a return on it.

Especially for overnight loans of cash, the short-term repo market seems to be very secure lending. After all, the party advancing cash gets to hold on to the Treasury securities as collateral. Specifically, if the other party that needed the cash ends up being unable to repurchase the Treasurys, then the party that lent the money at least gets to keep them as compensation for the contractual default. So, the lender either (a) gets his cash back with interest or (b) gets to keep the Treasury securities.

Now what the Fed announced last week is that it will itself enter the repo market and be prepared to offer up to $1.5 trillion in (newly created) US dollars in order to allow institutions to pledge their Treasurys as collateral and borrow such vast sums. But these transactions won’t be overnight loans; instead, the $1.5 trillion consists of $500 billion bursts of financing in the one-month and three-month repo contracts.

The whole point of this Fed intervention was to keep the implicit interest rate in the Treasury repo market down to acceptable levels. In other words, if the Fed had not intervened, then repo rates would have soared. Remember, last September the repo rate suddenly jumped from about 2.2 percent to 6 percent in two days. That was deemed a crisis at the time, justifying the Fed’s large (and recently expanded) ongoing intervention in the repo markets.

It’s true that when fear grips the world, investors do look to US government debt as a “safe haven.” That’s why US government bond yields collapsed to record lows recently and stock markets are falling: many portfolio managers are switching from equity to fixed-income assets.

But what the spikes in the repo market reveal is that in the very short term, such as a period of 1–90 days, actual cash is king. Right now, asset managers do not at all view a Treasury security “as almost the same thing” as US dollars issued by the Federal Reserve. One way the market communicates such a change in risk appetites is a “skyrocketing” implicit interest rate in the Treasury repo market. People who control actual US cash right now are not as willing to see that transformed into an “equivalent” amount of Treasurys, and so they demand a higher compensation (interest return) to make the asset swap. This is the market process that the Fed is trying desperately to hammer away.

By announcing that it is willing to throw up to $1.5 trillion in electronically created money in order to give three-month loans to those institutions that have bought Treasury debt, the Fed is bailing out not only the holders of Treasury debt, but also the Treasury itself. The Fed is ensuring a healthy demand—now and in the perceived future—for Treasurys, since now private bond dealers won’t worry about a sudden change in liquidity for their asset.

If the Fed bought $1.5 trillion in military hardware, everybody would instantly recognize it as a gift to the defense contractors and their main suppliers, including assembly workers (who might live in the state of a politician who voted for the spending bill). Likewise, when the Fed announces $1.5 trillion in new financing available for a certain portion of the financial sector, it is a gift to those institutions and their suppliers, such as the Treasury Department which has run up the debt by an additional $1.1 trillion in the last twelve months.

Reprinted from the Independent Institute.

The Theory and Brief History of Money and Banking

The Theory and Brief History of Money and Banking

The ultimate purpose of this booklet is to give the reader a solid grasp of how money works in today’s world. Yet before diving into the particulars of central banks, repo markets, and LIBOR—all topics that will be covered in future chapters—we should first provide a general framework giving the basic theory or “economic logic” of money and banking.

In short: why do we have money in the first place? Where does it come from, and what determines its form (livestock, metal ingots, coins, paper notes, electronic ledger entries, etc.)? What qualities make for a good money? What role do banks play—is it something other than what money itself does for us?

In this chapter, we’ll answer these elementary yet essential questions. To be clear, we are not here offering an actual history lesson, though we do mention some important historical episodes and illustrative examples. Rather we are providing a mental framework for understanding everything else that follows in the booklet.

The Limits of Direct Exchange

To understand the importance of money, let’s first imagine a society without money. In a world limited to barter, or what economists more precisely call direct exchange, there would still be private property and people would still benefit from voluntary trade. Because economic value is subjective—the “utility” of a good is in the eye (or mind) of the beholder—we can have win-win exchanges, in which both parties walk away correctly believing that they got the better end of the deal.

However, if society were limited to direct exchange—in which individuals only accept items in trade that they plan on using personally—then people would miss out on many advantageous transactions. Let’s consider a simplistic example. Suppose there are three individuals: a farmer, a butcher, and a cobbler. The farmer starts out with some eggs that he’s just taken from his hens. He would like to trade his eggs in order to get his tattered shoes repaired. The problem, though, is that the cobbler doesn’t want any eggs—but he would be willing to repair the shoes for bacon.

Unfortunately, the farmer doesn’t currently have bacon. However, his neighbor the butcher does have bacon. Yet the butcher doesn’t want to trade with the cobbler, because the butcher’s shoes are just fine. What the butcher would really like are some eggs. Yet, the farmer himself doesn’t like the taste of bacon, and would rather eat his own eggs.

In a world limited to direct exchange, these men are at an impasse, because no single transaction would benefit any pair of them. Yet all of them could improve their situation with a rearrangement of the goods.

The solution is to introduce indirect exchange, in which at least one person accepts an item in trade that he doesn’t plan on using himself but holds merely to trade away again in the future. In our example, suppose that the farmer has an epiphany: Even though he personally dislikes its taste, he trades his eggs to the butcher to obtain the bacon. Then he takes the bacon to the cobbler, who accepts it as payment for fixing his tattered shoes.

direct exchange diagram

After these two trades, all three individuals are better off than they were originally. Remember, though, that the solution relied on the farmer accepting an item in trade—in this case the bacon—that he didn’t plan on using himself. Economists call such a good a medium of exchange. Just as air is a “medium” through which sound waves travel, the bacon served as a medium through which the farmer’s ultimate exchange was effected—namely giving up his eggs in order to receive shoe-repair services.

Media of Exchange and the Origin of Money

As our fable illustrated, individuals can often improve their position by trading away goods that are less marketable and accepting goods that are more marketable, even if they don’t personally plan on using the items. As the founder of the Austrian school, Carl Menger, demonstrated in an 1892 essay1 (though earlier economists had anticipated some of the explanation), this principle is all we need to explain the emergence of money.

As individuals in the community seek to trade away their less marketable (or less liquid) goods in exchange for more marketable (or more liquid) goods, a snowball process is set in motion: those goods that started out with a wide appeal based on their intrinsic qualities see a boost in their popularity simply because they are so popular. (For a more modern example, the prisoners in a World War II POW camp would gladly trade away their rations in exchange for cigarettes even if they were nonsmokers, because enough of the other prisoners were smokers.2) Eventually, one or two commodities become so popular that just about everyone in the community would be willing to accept them in trade. At that point, money has been born.

A formal definition for money is that it’s a universally accepted medium of exchange. Menger’s explanation showed how such a commodity could emerge from its peers merely through voluntary transactions and without any individual seeing the big picture or trying to “invent” money. (See the endnotes for recent anthropological criticism of Mengerian-type explanations of the origin of money.3)

The Qualities of a Good (Commodity) Money

Money that emerged in the process we’ve described would necessarily be commodity money, in which the monetary good itself is also a regular commodity. (In Chapter 3 we will discuss fiat money, in which the monetary good serves no other function than to be the money.) Historically, many types of commodities have served as money in various regions, including livestock, shells, tobacco, and of course the precious metals gold and silver.

What would make a community gravitate towards some commodities but not others? Besides having a wide marketability, an individual would want a medium of exchange to possess the following qualities: ease of transport, durability, divisibility, homogeneity, and convenient size and weight for the intended transactions.

In our fable above, although bacon served as the medium of exchange, it would be ill-suited to serve this purpose generally, as bacon is perishable. Likewise, a shotgun might be very valuable in certain communities, but it’s not divisible; you can’t cut it in half to “make change.” Diamonds might seem like a great candidate for a medium of exchange, but they aren’t homogeneous: one giant diamond is more valuable than five smaller diamonds that (combined) weigh the same amount.

These types of considerations help explain why eventually gold and silver emerged as the market’s commodity monies of choice. These precious metals satisfied all of the criteria of what makes a convenient medium of exchange, and once the community generally agreed, they were money.

Monetary Calculation

The emergence of money meant that a single commodity was on one side of every transaction. This greatly reduced the calculations required to navigate the marketplace. For example, consider a merchant whose business required him to closely follow twenty different goods. In a world of pure barter—where each good traded directly against every other good—in principle he would have to keep track of 190 separate barter “prices”4 (meaning the ratios at which one good traded for another). But if one of those twenty goods also serves as the monetary good—maybe it’s silver—then the merchant only needs to keep track of nineteen different prices (all quoted in silver), because each of the other goods is always being bought and sold against silver.

Moving from a state of barter to a monetary economy allows for economic decisions to be appraised in terms of a standard unit. With the use of money, business owners can engage in accounting, where they can easily calculate whether they had a profitable year. Trying to compare revenues to expenses would be much more difficult in a pure barter system. A factory owner could know that her operation used up certain quantities of hundreds of input commodities (including labor hours), in order to produce certain quantities of dozens of outputs, but without being able to reckon these physically distinct commodities in terms of money prices, she would face the same type of problem plaguing socialist central planners.5

The Function of Monetary Coins (and Tokens)

We have seen how a commodity money can emerge spontaneously from a prior state of barter, facilitating exchanges and profit/loss calculations. However, even though a community benefits tremendously from the existence of money, there would still be limitations if the money remained in its “raw” form. It would hamper trade if shopkeepers had to perform metallurgical tests on hunks of metal that customers presented for payment to verify that the hunks were indeed silver (or gold, etc.) of the claimed weight.

The solution to this problem is to coin the raw hunks of metal into recognizable disks of a uniform size and purity (or “fineness”). We should emphasize that a full-bodied coin was not money because of the stamping process; the markings on the coin merely indicated to the community that the hunk of metal in question did indeed contain the specified weight in the underlying commodity that served as money.

In addition to striking full-bodied coins, another possible solution is for reputable outlets to issue token coins, which represent redemption claims on the issuer for a specified amount of the actual money commodity. Note that to perform their function well, even token coins would need to be recognizable in the community and difficult to counterfeit. For a modern example, consider the plastic chips issued by casinos: A Las Vegas casino needs to have chips that are distinctive and “authentic”-looking, and which can’t be easy for outsiders to replicate. Because such chips will be instantly redeemed by the casino, within its walls (and even perhaps in the surrounding neighborhood) they are “as good as money.” But a gambler who travels back home wouldn’t be able to buy groceries with chips issued from a Las Vegas casino.

Just as the money itself can arise without the intervention of political authorities, so too can the private sector handle the operations of turning the commodity money into coins. Indeed, numismatists agree that some of the highest-quality coins (and tokens) ever produced originated in eighteenth-century Britain from private mints.

The full story is too long to tell here,6 but the quick version is that the British Royal Mint had utterly failed to provide the common people with coins that could serve their needs for everyday commerce, and regulations prohibited banks from issuing notes in small denominations. As a result, employers resorted to various inconvenient remedies, including paying their workers in waves (so that, say, the first third of the employees would spend their new wages in town, after which the employers could then collect the coins in order to pay the second third of their workers, etc.) and making arrangements with the local tavern owners so that the workers’ beer tabs would effectively reduce the wages they were owed. The shortage of government-produced coinage was so severe that even obviously counterfeit coins were tolerated because bad money was better than no money at all.

In this intolerable situation, Thomas Williams, the principal owner of the giant Parys copper mine, hit upon the bright idea of installing a commercial-scale mint on the premises. He then struck (token) coins out of the copper with instructions on where they could be redeemed for money, and paid his workers—the ones actually mining the copper—with these token coins. Soon afterwards Matthew Boulton, famous for his collaboration with James Watt in the refinement of the modern steam engine, followed suit with the privately owned Soho Mint, where he was the first to implement a process of using steam power to mass-produce exquisite coinage. The following photos exhibit the remarkable craftsmanship of the privately struck coins and tokens from this era.7

Peck 1075 Coin
 A Penny from a Soho Mint 1797 pattern striking. Photo Credit: Bill McKivor, The Copper Corner.
Promissory Half Penny 1791
 A 1791 token promising a half-penny to the bearer. Photo Credit: Bill McKivor, The Copper Corner.

The Function and Origin of Banks

Even in a community with a commodity money stamped into high-quality coins, there would still be limitations on commerce. For example, wealthy individuals would be nervous about holding vast sums of gold or silver in their homes where they would be vulnerable to theft, and it would be inconvenient to transport large amounts of coin or bullion for every transaction involving a significant purchase price.

bank solves these problems by providing a secure location where members of the community can store their excess supplies of money. (The other main function of banks is to serve as credit intermediaries, which act as a conduit between borrowers and savers.) The goldsmith was a logical person to also act as banker, because his business already involved storing stockpiles of gold. It was easy enough for members of the community to deposit coins with the goldsmith in exchange for an official receipt indicating how much of the money commodity they (the depositors) had stored with him.

Exchange Gold for Promissory Note Diagram

The reason a booklet on the mechanics of money must also cover banking is that—to put it bluntly—banks enjoy the legal ability to create money. In Chapter 5 we will explain this process in much greater detail, but for now let us quote the Chicago Federal Reserve on the historical origins (at least in England) of this practice:

[B]anks can build up deposits by increasing loans and investments so long as they keep enough currency on hand to redeem whatever amounts the holders of deposits want to convert into currency. This unique attribute of the banking business was discovered many centuries ago.

It started with goldsmiths. As early bankers, they initially provided safekeeping services, making a profit from vault storage fees for gold and coins deposited with them. People would redeem their “deposit receipts” whenever they needed gold or coins to purchase something, and physically take the gold or coins to the seller who, in turn, would deposit them for safekeeping, often with the same banker. Everyone soon found that it was a lot easier simply to use the deposit receipts directly as a means of payment. These receipts, which became known as notes, were acceptable as money since whoever held them could go to the banker and exchange them for metallic money.

Then, bankers discovered that they could make loans merely by giving their promises to pay, or bank notes, to borrowers. In this way, banks began to create money. More notes could be issued than the gold and coin on hand because only a portion of the notes outstanding would be presented for payment at any one time. Enough metallic money had to be kept on hand, of course, to redeem whatever volume of notes was presented for payment. [Emphasis added.]8

Once the banker (such as the goldsmith) realized that his deposit receipts (“notes”) were treated by at least some members of the community as being “as good as money,” he could lend out some of the coins that his customers had deposited with him, even though the customers still held paper receipts entitling them to immediate redemption. The whole operation was viable so long as the banker always had enough coins on hand to satisfy whoever might show up to demand their deposits back.

Exchange Gold for Promissory Note Many Diagram

This booklet will focus on the mechanics and economic implications of the fact that banks have the legal ability to create money, but we’ll wrap up our historical sketch here with a note on the judicial treatment. If someone hands over an item for safekeeping in which the specific article is important—such as a college student placing her furniture in a storage unit for the summer, or a diner checking his coat when entering a restaurant—this is handled under bailment law. In such a situation, the person acting as a warehouser obtains physical possession but not legal ownership of the items in question, and is obligated to act as their custodian until the actual owner wishes to retrieve them. It would be a breach of contract for the manager of a storage facility to rent out the student’s couch, even if he had it safely back in her storage unit when she returned from summer break.

However, when the deposited items are fungible goods, such as wheat or oil, then the relationship is more nuanced. With such an “irregular deposit,” the depositor isn’t entitled to the specific physical items that were handed over for safekeeping, but instead merely expects to receive comparable items back. In the typical scenario, this is the type of deposit applicable to money; the people handing over coins to the goldsmith didn’t care about receiving back those particular coins, they merely wanted to be assured of obtaining the same number of comparable coins when they redeemed their deposit receipts (i.e., banknotes).

As a result of various court rulings, it is now standard to treat the deposit of money with a bank as a loan, so that the depositor becomes a creditor of the bank and the actual ownership of the money transfers to the banker, even for “demand deposits,” which are payable upon notice. Rightly or wrongly,9 it is this legal treatment that allowed the proverbial goldsmith to lend out some of the coins that his depositors had placed with him for safekeeping, and which allows modern banks to engage in “fractional reserve banking.” To reiterate, it is this practice by which banks can create (and destroy) money—a process that we will fully explain in Chapter 5.

We will close this chapter with an excerpt from an opinion issued by Lord Cottenham in the 1848 case Foley v. Hill and Others:

The money placed in the custody of a banker is, to all intents and purposes, the money of the banker, to do with as he pleases; he is guilty of no breach of trust in employing it; he is not answerable to the principal if he puts it into jeopardy, if he engages in a hazardous speculation; he is not bound to keep it or deal with it as the property of his principal; but he is, of course, answerable for the amount, because he has contracted.10

Judge Reading Ruling

Reprinted from the Mises Institute.

“Rules of Origin” Show Why Trade Agreements Aren’t Free Trade

“Rules of Origin” Show Why Trade Agreements Aren’t Free Trade

Ludwig von Mises famously argued that people must choose between outright socialism and unfettered capitalism, because there is no coherent “middle ground” between the two. The allegedly reasonable compromise of a highly interventionist state — where the authorities retain nominal private property but issue edicts regulating how legal owners may use their property — is unstable. Mises argued that one round of interventionism invites consequences that are even worse than the original problem, leading to yet more interventionism.

During the debates over Obamacare, I pointed out just how relevant Mises’s lesson was: we couldn’t get the “good parts” of Obamacare (such as universal coverage) without the “bad parts” (such as the individual mandate and massive tax hikes). When it comes to today’s controversies over trade with China, once again Mises’s insights are valuable. You can’t levy punitive tariffs on China but leave other trading routes relatively free, because then the Chinese will simply ship their exports via a more circuitous route. China hawks need to decide if they are going to abandon attempts to coercively manage trade, or if they are prepared for even more extensive top-down planning of global commerce.

Mises and Milk

Mises’s standard example for the phenomenon of one intervention leading to another was a price control on milk. Suppose the government wants to make milk more affordable for poor families. It can enact strict price controls on milk. But if this isolated price ceiling is imposed in the context of an otherwise free market economy, the immediate result will be a shortage of milk. Now, rather than poor families struggling to afford milk for their children, the stores won’t carry any milk, period. At this point, the government can either admit its error and retreat back to pure laissez-faire, or it can impose further price controls, this time on cattle feed etc. in order to coax dairy farmers into once again supplying the market with milk. Yet this second round of intervention leads to even more undesirable consequences, and so on.

Mises’s Lesson Applied to International Trade

In the debate over free trade, we see a similar phenomenon. The Trump administration has been engaged in a low-level trade war with China, levying targeted tariffs on its imports in an effort to bring Beijing to the bargaining table. Yet the remaining pockets of free(r) trade are stymying the effect, because of the phenomenon of “transshipment” — in which China exports its goods to a third country, from which they can be sold to the United States without penalty. As a recent article in the WSJ, entitled “American Tariffs on China Are Being Blunted by Trade Cheats,” pointed out,

Billions of dollars’ worth of China-made goods subject to tariffs by the Trump administration in its trade fight with Beijing are dodging the China levies by entering the U.S. via other countries in Asia, especially Vietnam, according to trade data and overseas officials.

And thus we see the relevance of Mises’s warning. The goal of the initial intervention — the levying of tariffs on Chinese imports — was to hurt Chinese exporters and thereby convince Chinese government officials to concede to American demands. But much of the intended effect has been muted because of transshipment.

At this point, American officials can admit that their approach was ill-advised, and stop using taxes as a way to make America great again. Or, they can expand the trade war with China to include an extensive monitoring of the content of goods coming from every other country on Earth.

“Rules of Origin” in Trade Agreements

This isn’t hyperbole on my part. As Ryan McMaken explained on this site over the summer, special trade agreements — such as the US pact with Central America — have clauses signifying that only qualified goods can escape duties. McMaken linked to this relevant passage from an FAQ on the CAFTA-DR (Central America-Dominican Republic-United States-Free Trade Agreement):

How can my product qualify to take advantage of the CAFTA-DR?

The product must qualify as an “originating” good under the terms of the Agreement. This means that the product must have sufficient U.S., Nicaraguan, Guatemalan, Honduran, Salvadoran, Costa Rican, and/or Dominican content or processing to meet the criteria of the Agreement. If goods contain only U.S. or Central American or Dominican Republic inputs, they qualify. If they contain some inputs from other countries, they still might qualify if they meet specific criteria set out in the Rules of Origin of the Agreement. Each product has a unique rule, based on its tariff classification. Most of the rules require either that the non-originating inputs undergo a specified transformation through processing in the United States or one or more of the other signatory countries (tariff shift method) and/or that they have a sufficient level of originating content as determined by a formula (regional value content method).

And now we see why a “free trade agreement” in practice isn’t simply an index card declaring, “Tariffs on Country X are 0 percent, three cheers for Bastiat!” These are managed trade agreements, with hundreds of pages devoted to detailed regulations that smack of top-down Soviet planning.


As Mises stressed time and again, people must decide whether to embrace capitalism or socialism. There is no third way, where we can enjoy the dynamism of markets while avoiding their “excesses” through strategic interventions. In the case of tariffs, particularly when the goal isn’t a broad-based revenue source but rather the achievement of a bargaining position with a particular country, a simple policy will soon break down, because the targeted country can simply ship its exports via other channels. (This same problem occurs in the case of “carbon tariffs” levied on countries that don’t punish greenhouse gas emitters to the same extent as the original country.)

The only logical end point is a country having to keep track of the entire network of trade flows, and levying the appropriate tariffs accordingly. Rather than this byzantine nightmare, trade hawks would be wiser to throw in the towel and try another strategy to achieve their goals. Unilateral free trade would make Americans richer, and our example might eventually inspire other governments to allow their own people more economic freedom as well.

Reprinted from the Mises Institute

Understanding Elizabeth Warren’s Radical Wealth Tax

Understanding Elizabeth Warren’s Radical Wealth Tax

Democratic presidential candidate Elizabeth Warren has had a long-standing call for a 2% wealth tax on any individuals with a net worth exceeding $50 million, and a 3% tax on wealth exceeding $1 billion. Yet when pressed on how to pay for her “Medicare for All” plan, Warren upped the ante to a 6% wealth tax for those fortunes exceeding $1 billion. (As I noted at the time of the announcement: If Warren doubles her wealth tax during the campaign, imagine how fast it will rise if she’s actually elected.)

Naturally, many conservative and libertarian analysts recoiled from such an economically destructive proposal. One of the ways critics used to illustrate the severity of Warren’s idea was to translate a wealth tax into an “equivalent” income tax on dividends, interest, and capital gains. But other economists pointed out problems with that line of attack, because after all, wealth and income are different things, and so taxes on them affect behavior differently. Wealth taxes are inefficient, no doubt about it, but not because “it’s the same thing as a huge income tax.” In the present piece I’ll try to referee the disputes and present the reader with an intuitive understanding of the issues involved.

How Wealth Taxes Can Correspond to Very High Income Taxes

In order to show that Warren’s seemingly modest 6% wealth tax was in fact quite radical, Richard Rubin at the Wall Street Journal warned that “Warren has unveiled sweeping tax proposals that would push federal tax rates on some billionaires and multimillionaires above 100%.” Likewise, Columbia University economist Wojciech Kopczuk—while commenting on the more academic proposal coming from economists Gabriel Zucman and Emmanual Saez—argued that“If you consider a safe rate of return of, say, 3%, a 3% wealth tax is a 103% tax on the corresponding capital income and a 6% tax rate is a 206% tax.”

Before proceeding, let’s illustrate Kopczuk’s argument with a numerical example. (Note that in the rest of this article, in my examples I’m going to use small amounts of wealth, such as $1,000, to keep the math simple. Warren’s actual proposals of course only apply to wealth exceeding $50 million and $1 billion—at least so far!)

Now then, suppose someone starts with $1,000 in wealth. If he consumes it, then he faces no wealth tax nor income tax. (Kopczuk adopts the convention that any wealth taxes are assessed on wealth at the end of the period, while income taxes are based on income generated during the period.)

However, suppose the individual invests the wealth in fairly safe bonds that yield a return of 3%. At the end of the period, the individual will have the original $1,000 plus the $30 in gross interest income, for a new level of wealth of $1,030. If that wealth is then taxed at 3%, the individual owes the IRS ($1,030 x 3%) = $31 (with rounding). Yikes! The income generated by that wealth was only $30 during the period, and so if the individual had the same liability from a income tax (applied to interest), then the “equivalent” tax rate would be 103%!

In General, a Wealth Tax Is Not Equivalent to an Income Tax

Although such calculations may be useful to wake up the average American to just how economically destructive even a “low” wealth tax may be, strictly speaking it is incorrect to argue that a wealth tax of x% is “equivalent to” or “the same thing as” a tax on capital income of y%. Over at EconLog, economists Scott Sumner and David R. Henderson both laid out some of the problems.

For our purposes, let me focus on Henderson’s commentary, where he showed the problem with Kopczuk’s analysis. Note, however, that in his actual example, Henderson ran the numbers for a 2% wealth tax. I’m going to change the calculations to make his same point, but using a 3% wealth tax, because I think that’s easier for the reader and also to be consistent with my commentary above:

The way to see what the marginal tax rate on capital income is[,] is to think on the margin: change the income from capital and see how much extra tax is paid.

So, for example, start with $1,000 at the start of the year that earns what Kopczuk calls the riskless rate of return, 3%. With a wealth tax, $1,030 at year’s end is taxed at [3%], leaving the owner with [97%] of $1,030, which is [$999.10].

Now raise the rate of return to 4%. With a wealth tax, $1,040 is taxed at [3%], leaving the owner with [97%] of $1,040, which is [$1,008.80].

How much more did the owner of capital net from the investment at 4% rather than at 3%? [$1,008.80] minus [$999.10], which is [$9.70]. In other words, for an extra income from capital of $10, the owner kept [$9.70]. The wealth tax amounted to a [3%] tax on the income from capital. [David R. Henderson, bold added, with bracketed numbers reflecting Murphy’s tweaking of the size of the wealth tax.]

As Henderson’s example shows, in general you can’t take a given wealth tax and then translate it into the “equivalent” income tax. In his example, if an investor has the choice between Investment A that is relatively safe and carries a return of 3%, and Investment B that is riskier but promises the higher expected return of 4%, then the wealth tax of 3% provides different incentives than a tax on capital income of 103%.

Specifically, under a 3% wealth tax, the investor who takes on the extra risk by switching to Investment B—trying to boost his gross rate of return from 3% to 4%, and hence his gross income on the investment from $30 to $40—will be able to keep 97% of that extra $10 in expected return on the investment.

In utter contrast, if the investor faces not a wealth tax, but instead a tax on capital income of 103%, then even if the riskier investment pays off as expected, the investor ends up worse off! Specifically, if he goes with Investment A our investor ends up with $1,030 gross on which he must pay ($30 x 103%) = $30.90 in income tax, leaving him with $999.10 after the dust settles. But if he goes with the riskier Investment B andeven if it pays off as he’d hoped, the investor ends up with $1,040 gross on which he must pay ($40 x 103%) = $41.20 in income tax, leaving him with $998.80 when the dust settles.

In summary, David R. Henderson has come up with a specific example to show why it’s wrong to argue that a wealth tax of 3% is “equivalent to” a capital income tax of 103%. If we assume an investor has the option of putting his wealth into a riskier investment with a higher rate of return, then the 3% wealth tax only distorts the decision by 3% (loosely speaking). If the risker investment pays out, then the investor’s upside is only clipped by the modest 3% tax on the extra wealth he now holds. In contrast, under a 103% income tax, then it would be insane for the individual to even consider the riskier asset. Perversely, the more it pays out, the worse off the investor ends up, because the government assesses a tax that is proportional to, but bigger than, any gains.

At this point, we see that wealth and income taxes can have very different effects on investor behavior. Generally speaking, if we are considering long-term deployments of financial capital, and comparing it to a no-tax baseline, a modest wealth tax will lead investors to seek riskier assets earning higher (expected) rates of return, while a very high capital income tax will lead investors to tread water, putting their wealth into safe assets that earn very low rates of return. Both types of taxes distort financial decisions, but they do so in different ways. They aren’t “equivalent” in general.

Still Not the Full Story

My apologies dear reader, but we’re not done yet: Henderson’s analysis isn’t the full story, either. Strictly speaking, what he showed is that under a wealth tax of 3%, an investor who consumes all of his wealth at the end of the period only faces a marginal income tax rate of 3%. Yet in practice, most investors probably aren’t planning on consuming everything in one fell swoop, and so Henderson has led readers to understate the economic impact of a wealth tax.

Recall the example: An investor who switches his $1,000 in capital from an asset yielding 3% to one that yields 4% will see his gross income jump from $30 to $40. Henderson reasoned that under a wealth tax of 3%, the investor got to retain $9.70 of the extra $10 in gross income, and concluded that the marginal income tax rate was therefore only 3%. (A reminder to avoid confusion: In order to keep the analysis comparable to the quotation from Kopczuk, I amended Henderson’s numbers to deal with a 3% wealth tax rather than a 2% version.)

Yet to repeat, this is only true if the investor consumes that $9.70. If instead the investor holds it another period, then it will trigger a second tax liability under the wealth tax, this time of ($9.70 x 3%) = 29 cents. And then if the investor carries the balance forward yet again, at the end of the third year he must pay another ($9.41 x 3%) = 28 cents in wealth tax. In contrast, under an income tax regime, if the investor just sits on his after-tax wealth after he earns it the first year, rather than deploying it to earn new income, then he owes no more additional tax on it.

In short, if our hypothetical investor had long-term plans for his wealth, then Henderson underestimated the burden of the wealth tax. In the limit, if the investor earned a one-shot return of $10 and then put it somewhere earning no return, it would asymptotically approach $0 over the years, as the government kept nibbling 3% annually at it. (For example, after 20 years of getting hit with the wealth tax, the original $10 in extra interest income earned that first year would have been whittled down to about $5.44.)

Let’s do one last example to illustrate the subtleties involved. Suppose our investor earned that extra $10 during this year (by moving his $1,000 into an asset that yielded 4% rather than 3%), and then wants to put the $10 under his mattress, where he intends to keep it for 50 years. How then would this extra $10 he earned this year, affect his long-term tax liability? Well, at the end of the first year he owes 30 cents. At the end of the second year he owes 29 cents on the remainder, and at the end of (say) the 25th year he owes 14 cents. However, when computing the burden from today’s perspective, those future tax payments need to be discounted. Since Kopczuk and Henderson both assumed a “safe” return of 3%, we can use that for a discount rate. (For example, the 14-cent wealth tax liability due in 25 years only has a present discounted value to our individual of 7 cents.)

Using this approach, the total wealth tax (in present-dollar terms) that the incremental $10 in wealth will cause our investor, over a 50-year time horizon, is some $4.89. In that sense, then, when our investor is considering whether to rearrange his portfolio in order to earn an extra $10, he faces a “marginal income tax rate” of about 49%.

Another way of showing the issue is to assume our investor wants to set aside a portion of his initial $10 in extra wealth, in order to cover all of the future wealth tax payments over the 50-year horizon. If he puts his earmarked “sinking tax fund” wealth into the relatively safe asset yielding 3%, then the investor must allocate $6.01 of his initial $10, just to cover the future wealth tax payments. Using this approach, the investor could understandably conclude that of his $10 in gross earnings—since he could only put $3.99 under the mattress “free and clear” for use in 50 years—he effectively paid the equivalent of a 60.1% marginal income tax rate.


Putting aside the moral problems with taxation—it’s theft, as a popular libertarian slogan reminds us—Elizabeth Warren’s proposed wealth taxes will have devastating consequences on capital formation, and will encourage investors to hold riskier assets than they otherwise would have. In order to illustrate the magnitudes involved, some analysts translated Warren’s proposals into “equivalent” income tax rates.

However, wealth and income are different concepts, and in general taxes on wealth and income will have different effects. For those investors with a short planning horizon, a modest wealth tax has a relatively modest impact on the decision to save for the future. However, for those with longer time horizons, even a seemingly modest wealth tax has an economic impact akin to a large income tax.

Reprinted from Catalyst.

The Bogus “Consensus” Argument on Climate Change

The Bogus “Consensus” Argument on Climate Change

One of the popular rhetorical moves in the climate change debate is for advocates of aggressive government intervention to claim that “97% of scientists” agree with their position, and so therefore any critics must be unscientific “deniers.”

Now these claims have been dubious from the start; people like David Friedman have demonstrated that the “97% consensus” assertion became a talking point only through a biased procedure that mischaracterized how journal articles were rated, and thereby inflating the estimate.

But beyond that, a review in The New Republic of a book critical of mainstream economics uses the exact same degree of consensus in order to cast aspersions on the science of economics. In other words, when it comes to the nearly unanimous rejection of rent control or tariffs among professional economists, at least some progressive leftists conclude that there must be group-think involved. The one consistent thread in both cases—that of the climate scientists and that of the economists—is that The New Republic takes the side that will expand the scope of government power, a central tenet since its birth by Herbert Croly a century ago.

The Dubious “97% Consensus” Claim Regarding Climate Science

Back in 2014, David Friedman worked through the original paper that kicked off the “97% consensus” talking point. What the original authors, Cook et al., actually found in their 2013 paper was that 97.1% of the relevant articles agreed that humans contribute to global warming. But notice that that is not at all the same thing as saying that humans are the main contributors to observed global warming (since the Industrial Revolution).

This is a huge distinction. For example, I co-authored a Cato study with climate scientists Pat Michaels and Chip Knappenberger, in which we strongly opposed a U.S. carbon tax. Yet both Michaels and Knappenberger would be climate scientists who were part of the “97% consensus” according to Cook et al. That is, Michaels and Knappenberger both agree that, other things equal, human activity that emits carbon dioxide will make the world warmer than it otherwise would be. That observation by itself does not mean there is a crisis nor does it justify a large carbon tax.

Incidentally, when it comes down to what Cook et al. actually found, economist David R. Henderson noticed that it was even less impressive than what Friedman had reported. Here’s Henderson:

[Cook et al.] got their 97 percent by considering only those abstracts that expressed a position on anthropogenic global warming (AGW). I find it interesting that 2/3 of the abstracts did not take a position. So, taking into account David Friedman’s criticism above, and mine, Cook and Bedford, in summarizing their findings, should have said, “Of the approximately one-third of climate scientists writing on global warming who stated a position on the role of humans, 97% thought humans contribute somewhat to global warming.” That doesn’t quite have the same ring, does it? [David R. Henderson, bold added.]

So to sum up: The casual statements in the corporate media and in online arguments would lead the average person to believe that 97% of scientists who have published on climate change think that humans are the main drivers of global warming. And yet, at least if we review the original Cook et al. (2013) paper that kicked off the talking point, what they actually found was that of the sampled papers on climate change, only one-third of them expressed a view about its causes, and then of that subset, 97% agreed that humans were at least one cause of climate change. This would be truth-in-advertising, something foreign in the political discussion to which all AGW issues now seem to descend.

The New Republic’s Differing Attitudes Towards Consensus

The journal The New Republic was founded in 1914. Its website states: “For over 100 years, we have championed progressive ideas and challenged popular opinion….The New Republic promotes novel solutions for today’s most critical issues.”

With that context, it’s not surprising that The New Republic uses the alleged 97% consensus in climate science the way other progressive outlets typically do. Here’s an excerpt from a 2015 article (by Rebecca Leber) in which Republicans were excoriated for their anti-science stance on climate change:

Two years ago, a group of international researchers led by University of Queensland’s John Cook surveyed 12,000 abstracts of peer-reviewed papers on climate change since the 1990s. Out of the 4,000 papers that took a position one way or another on the causes of global warming, 97 percent of them were in agreement: Humans are the primary cause. By putting a number on the scientific consensus, the study provided everyone from President Barack Obama to comedian John Oliver with a tidy talking point. [Leber, bold added.]

Notice already that Leber is helping to perpetuate a falsehood, though she can be forgiven—part of David Friedman’s blog post was to show that Cook himself was responsible (Friedman calls it an outright lie) for the confusion regarding what he and his co-authors actually found. And notice that Leber confirms what I have claimed in this post, namely that it was the Cook et al. (2013) paper that originally provided the “talking point” (her term) about so-called consensus.

The point of Leber’s essay is to then denounce Ted Cruz and certain other Republicans for ignoring this consensus among climate scientists:

All this debate over one statistic might seem silly, but it’s important that Americans understand there is overwhelming agreement about human-caused global warming. Deniers have managed to undermine how the public views climate science, which in turn makes voters less likely to support climate action.

Now here’s what’s really interesting. A colleague sent me a recent review in The New Republic of a new book by Binyan Appelbaum that is critical of the economics profession. The reviewer, Robin Kaiser-Schatzlein, quoted with approval Appelbaum’s low view of consensus in economics:

Appelbaum shows the strangely high degree of consensus in the field of economics, including a 1979 survey of economists that “found 98 percent opposed rent controls, 97 percent opposed tariffs, 95 percent favored floating exchange rates, and 90 percent opposed minimum wage laws.” And in a moment of impish humor he notes that “Although nature tends toward entropy, they shared a confidence that economies tend toward equilibrium.” Economists shared a creepy lack of doubt about how the world worked. [Kaiser-Schatzlein, bold added.]

Isn’t that amazing? Rather than hunting down and demonizing Democratic politicians who dare to oppose the expert consensus on items like rent control—which Bernie Sanders has recently promoted—the reaction here is to guffaw at the hubris and “creepy lack of doubt about how the world [works].”


From the beginning, the “97% consensus” claim about climate change has been dubious, with supporters claiming that it represented much more than it really did. Furthermore, a recent book review in The New Republic shows that when it comes to economic science, 97% consensus means nothing, if it doesn’t support progressive politics.

Republished from the Mises Institute

Do We Really Have a Decade Left to Solve Climate Change?

Do We Really Have a Decade Left to Solve Climate Change?

Wise alecks on social media noted with amusement how Beto O’Rourke recently claimed humans had only ten years to act on climate change, thus one-upping Alexandria Ocasio-Cortez who had previously gone out on a limb by putting the deadline at twelve years. Snark aside, it’s important to point out that the “consensus science” as codified, for example, in the periodic reports from the United Nations do not support such a cliff-hanger mentality at all.

Our Political Figures Ignore the IPCC

The quickest way to make this point is to reproduce something I highlighted several years ago in an IER post where I caught Paul Krugman just making up stuff about climate change. Specifically, the following table comes from the latest Intergovernmental Panel on Climate Change (IPCC) report, the AR5 (Table SPM.2):

Source: IPCC AR5, Working Group III, Summary for Policymakers

To make it easier to read, I’ll excerpt the relevant left and right portions of the table below:

SOURCE: Adapted from IPCC AR5, Working Group III, Summary for Policymakers, Table SPM.2

There’s a lot of information in the table, but let me summarize the important elements vis-à-vis the recent claims from O’Rourke and Ocasio-Cortez. The beige cells in the adapted table above show the percentage increases in the total (undiscounted) mitigation costs necessary to achieve the far-left (white cells) atmospheric concentrations of greenhouse gases in the year 2100, for the years 2030-2050 and also for 2050-2100, for two different scenarios of total emissions (either below 55 gigatons of CO2-equivalent, or above).

In other words, the beige cells show us how much a delay of government action through the year 2030 will increase the cost necessary to achieve the specified atmospheric concentrations for the year 2100 (white cells). Specifically, the beige cells show that by “doing nothing” about climate change until the year 2030, even in a high-emission baseline scenario, the IPCC’s best guess of the cost of achieving the aggressive outcome rises by 44 percent in the years 2030-2050 and 37 percent in the years 2050-2100.

Now to be sure, the rhetorical point of the above table in the AR5 was to encourage support for climate mitigation policies. The people who put together this table for policymakers wanted to show, “Hey, since we’re obviously going to have to deal with climate change eventually, we might as well get going, because the longer we delay, the more expensive it will be.”

IPCC: An Inconvenient Truth

My modest point here, however, is to show that this table now poses an awkward stumbling block for those—like O’Rourke and Ocasio-Cortez—trying to scare people into supporting ludicrously expensive and aggressive proposals to “fight climate change.” If O’Rourke and Ocasio-Cortez were anywhere close to being correct when issuing their ever-shrinking windows for action, then in the IPCC table above, the beige cells should have all had infinity signs, and in a footnote it would explain: “If we wait until 2030 to begin mitigation efforts, we will all die.”

But that’s not what the UN report told us. Instead, it reported that yes, the costs of achieving various climate targets (as measured by atmospheric concentrations of CO2 in the year 2100) would be higher due to delay, but even in a pessimistic scenario, the best-guess of the cost increase was 44 percent.


In this post, I highlighted one particular table from the most recent UN report on the science of climate change in order to show just how baseless are the recent claims that humans have years to act on climate change. Over at Reason, Ronald Bailey marshals more evidence—again, from the very “consensus science” documents we are supposed to rely on—to show that these claims are nonsense.

This whole episode is yet another example demonstrating the farce of the climate change policy debate. Whenever a critic disagrees with the most radical proposals that would—according to their own proponents—transform Western society, the critic is berated as a science denier. And yet, even a cursory examination of the actual technical reports shows that the prophets of doom are the ones who are spouting forth unsupported claims.

Republished from

The Upside-Down World of MMT

The Upside-Down World of MMT

[Editor’s note: MMT is back in the news, championed by Congresswoman Alexandria Ocasio-Cortez and former Bernie Sanders advisor Stephanie Kelton. Economists like Brad DeLong and Paul Krugman are giving MMT at least faint praise, and even National Review has favorable things to say. Ironically, MMT is neither modern nor truly “monetary;” instead it is a combination of tired fiscal and monetary policies. Our Senior Fellow Robert Murphy first wrote this article debunking MMT in 2011, but every word applies today.]

Modern Monetary Theory (MMT) is a hip economic/financial paradigm apparently sweeping a world unsatisfied with mainstream economics. Over the past year, I have been hearing a growing number of people refer to MMT: either fans who think it blows up my Austrian views, or foes who think it deserves a full-scale critique.

MMT’s underground popularity derives from its seeming mathematical rigor, its disagreement with the obviously flawed doctrines of standard neo-Keynesian orthodoxy, and its underlying message of hope that the perceived constraints on government deficit spending are an illusion. The MMT proponents tell us that fiat monetary systems have removed the shackles associated with the gold standard, and that our economic recovery is limited only by our failure to understand how modern money and banking work.

After my admittedly brief exploration, I have concluded that the MMT worldview doesn’t live up to its promises. However, as an Austrian economist I know how annoying it is when “big guns” in the economics profession reject my own position as nonsense without even taking the time to spell out what is supposedly wrong with the Misesian approach. Therefore, in the present post I’ll try to fairly summarize a major plank in MMT thought and show why it is misleading at best, and downright false at worst.

Background on MMT

One thing I should make clear upfront is that MMT is not the same thing as neo-Keynesian economics, as expounded by the likes of Paul Krugman. In fact, Krugman has actively criticized the MMTers himself (to which they responded here and here, to list just two instances).

MMT is linked to the older doctrine of “chartalism,” for readers who are more familiar with the latter term. The fascinating aspect of MMT is that it turns standard views on their head. For example, MMTers hold that the sovereign issuer of fiat currency can never become insolvent. For the MMTers, the point of taxation isn’t to raise revenue for the government, but rather to regulate aggregate demand.

It would be foolish for me to try to summarize the MMT position, as I am sure I would offend its proponents by my imprecision. As Morpheus said of the Matrix, I cannot tell you of the worldview of the MMTers; you must see it for yourself. Warren Mosler’s website is reputed to be the best one-stop shop, and the comments at my open-ended blog post are filled with suggested readings from actual MMTers.

The Counterintuitive MMT Position on Government Deficits

To illustrate my problems with MMT, let’s focus on a specific issue: the debate over the government budget deficit. With Austrians and other libertarian types calling for immediate cuts in spending, while Keynesians call for future spending restraint and tax hikes to slow the increase in debt down the road, the MMTers come along and say both sides are ignorant.

According to many proponents of MMT, “deficits don’t matter” when a sovereign government can issue its own fiat currency, and all the hand wringing over the government’s solvency is absurd. In fact, the MMTers claim that given the reality of a US trade deficit, a sharp drop in the government’s budget deficit would hamper the private sector’s ability to save. Thus, the Austrians are unwittingly calling for a collapse in private saving when they foolishly demand government austerity.

I have scoured the websites of a few prominent MMTers and here is the best explanation of this reasoning that I could find. The quotation below is somewhat lengthy and contains equations, but reproducing it is the only way to be sure I am not misrepresenting the MMT position:

The national accounts concept underpins the basic income-expenditure model that is at the heart of introductory macroeconomics. We can view this model in two ways: (a) from the perspective of the sources of spending; and (b) from the perspective of the uses of the income produced. Bringing these two perspectives (of the same thing) together generates the sectoral balances.

So from the sources perspective we write:

GDP = C + I + G + (X — M)

which says that total national income (GDP) is the sum of total final consumption spending (C), total private investment (I), total government spending (G) and net exports (X — M) [i.e., exports minus imports].

From the uses perspective, national income (GDP) can be used for:

GDP = C + S + T

which says that GDP (income) ultimately comes back to households who consume (C), save (S) or pay taxes (T) with it once all the distributions are made.

So if we equate these two perspectives of GDP, we get:

C + S + T = C + I + G + (X — M)

This can be simplified by cancelling out the C from both sides and re-arranging (shifting things around but still satisfying the rules of algebra) into what we call the sectoral balances view of the national accounts.

(I — S) + (G — T) + (X — M) = 0

That is the three balances have to sum to zero. The sectoral balances derived are:

  • The private domestic balance (I — S) …

  • The Budget Deficit (G — T) …

  • The Current Account balance (X — M) …

A simplification is to add (I — S) + (X — M) and call it the non-government sector. Then you get the basic result that the government balance equals exactly $-for-$ … the non-government balance (the sum of the private domestic and external balances). This is also a basic rule derived from the national accounts and has to apply at all times.

For the purposes of our discussion, let’s simplify things by taking out the international-trade aspect. (We can justify this by looking at the world as a whole, which obviously can’t run a trade deficit or trade surplus,1 and then analyzing the effects of changes in the total budget deficits of all the various governments.)

So if we take out exports and imports, and rearrange the remaining terms, we derive this equation:

G − T = S − I

That is, the amount of government spending minus total tax revenue, is necessarily equal to private saving minus private investment. The MMTers might succinctly express this relationship in words:

Government Budget Deficit = Net Private Saving.

This equation underpins the MMTers’ disdain for the tea party’s call for fiscal austerity. We derived the above equation through accounting tautologies, not by relying on any particular economic theory, so it should be impregnable. And gosh it sure looks like if the government were to reduce its budget deficit, then the private sector’s saving would necessarily go down. Yikes! Have the Austrians been unwittingly advocating massive capital destruction without realizing it?

Of Course You Don’t Need the Government in Order to Save

When I first encountered such a claim — that the government budget deficit was necessary to allow for even the mathematical possibility of net private-sector saving — I knew something was fishy. For example, in my introductory textbook I devote Chapter 4 to “Robinson Crusoe” economics.

To explain the importance of saving and investment in a barter economy, I walk through a simple numerical example where Crusoe can gather ten coconuts per day with his bare hands. This is his “real income.” But to get ahead in life, Crusoe needs to save — to live below his means. Thus, for 25 days in a row, Crusoe gathers his ten coconuts per day as usual, but only eats eight of them. This allows him to accumulate a stockpile of 50 coconuts, which can serve as a ten-day buffer (on half-rations) should Crusoe become sick or injured.

Crusoe can do even better. He takes two days off from climbing trees and gathering coconuts (with his bare hands), in order to collect sticks and vines. Then he uses these natural resources to create a long pole that will greatly augment his labor in the future in terms of coconuts gathered per hour. This investment in the capital good was only possible because of Crusoe’s prior saving; he wouldn’t have been able to last two days without eating had he not been able to draw down on his stockpile of 50 coconuts.

This is an admittedly simple story, but it gets across the basic concepts of income, consumption, saving, investment, and economic growth. Now in this tale, I never had to posit a government running a budget deficit to make the story “work.” Crusoe is able to truly live below his means — to consume less than his income — and thereby channel resources into the production of more capital goods. This augments his future productivity, leading to a higher income (and hence consumption) in the future. There is no trick here, and Crusoe’s saving is indeed “net” in the sense that it is not counterbalanced by a consumption loan taken out by his neighbor Friday.

So how in the world are we to interpret the MMTers’ proclamation that “net private saving” necessarily equals the government’s budget deficit (if we ignore international trade)?

When I raised this question on my blog, Nick Rowe — who is a very sharp economist — defended the MMT statement in this way:

Robert [Murphy]: “In particular, I think it is crazy when people say that if the federal government runs a budget surplus, then by simple accounting the private sector can’t save.”

[Nick Rowe:] That’s perfectly correct, and standard, once you do the translation. Assume [an economy closed to international trade]. Define “private saving” as “private saving minus Investment” … which is how MMTers normally use the word “saving”, or sometimes “net saving”. Then it’s just standard National Income Accounting. Y=C+I+G, and S=Y-T-C, therefore S-I=G-T.

And there you have it: When MMTers speak of “net saving,” they don’t mean that people collectively save more than people collectively borrow. No, they mean people collectively save more than people collectively invest.

I’m not trying to make fun of Nick Rowe; he is a professional economist who has written some very nuanced posts relating MMT to more orthodox mainstream economics. But look at what he was forced to type: “Define ‘private saving’ as ‘private saving minus investment.'” As I noted in my response to Rowe, if we define “private saving” as “private saving,” then my critique of MMT stands. (That’s supposed to be funny, by the way — at least insofar as economics can be funny.)

Now Nick Rowe and the MMTers are certainly correct when they observe that “private saving net of private investment” can’t grow without a government budget deficit (again if we disregard foreign trade). But so what? The whole benefit of private saving is that it allows for more private investment.

This is the fundamental problem with relying on macro-accounting tautologies; people often bring in causal arguments from economic theories without realizing they are doing so. Let’s look again at the equation causing so much confusion:

G T = S I

As a free-market economist, I don’t need to run from this tautology. I can use it to underscore the familiar “crowding out” critique of government deficit spending. Specifically, if government spending (G) goes up while tax revenue (T) remains the same, then the left-hand side of the equation gets bigger as the government budget deficit grows. So the accounting tells us that the right-hand side must get bigger too. It may happen partially because people cut down on consumption and save more (due to higher interest rates and their expectation of higher tax burdens in the future), but it may also happen because private-sector investment goes down. In other words, as the government borrows and spends more, the equation tells us we might see lower private consumption, rising interest rates, and real resources being siphoned out of private investment into pork-barrel spending projects. I can tell my “story” of the dangers of government deficit spending with that equation just fine.

Of course, the Keynesians and MMTers would have a different spin on the result of higher government spending in our current economic environment, but that’s not really the issue here. My point is that the national-income accounting tautologies aren’t a good critique of the tea party after all. Those equations are just as consistent with economic theories claiming that government spending cuts will lead to faster economic growth. The fans of MMT should therefore stop pointing to those identities as if they prove the futility of government austerity during an economic downturn. Those tautologies, and the cherished equations of the three sectors, are consistent with post-Keynesian and tea party economics.

As a final way to illustrate the non sequitur of the equations involving government budget deficits, note that we could do the same thing with, say, Google. Go back through all the equations above, and redefine G to mean “total spending by Google.” Then C would be “total consumption spending by the-world-except-Google,” and so on.

After doing this, we would be able to prove — with mathematical certainty — that unless Google were willing to go deeper into debt next year, the world-except-Google would be unable to accumulate net financial assets, in the way MMTers define that term. The proper response to this (perfectly valid) observation is, Who cares?2

Not All Spending and Income Are Created Equal

Thus far I have accepted the MMT premises on their own terms, and shown that MMT’s proponents often read more into their neutral accounting relationships than is justified by the relationships per se. However, in this final section I want to point out something even subtler.

One way to describe MMT is that is a “nominal” model of the economy, looking at flows of money without inquiring too deeply about the economic significance behind the flows. This article is already lengthy, so let me illustrate the problem with an analogy.

Suppose Tabitha has an income of $100,000, out of which she consumes $90,000. Tabitha takes her savings of $10,000 and lends it at 5 percent interest to Sam, who signs over an IOU promising to pay Tabitha $10,500 in 12 months.

Now let’s stop and ask, did Tabitha save money in this scenario? Yes, of course she did. Another question: did Tabitha accumulate net financial assets? Yes, of course she did: she is holding a legally binding IOU from Sam, which possesses a current market value of $10,000 and will grow in value over time as the payoff date approaches. (Changes in Sam’s solvency and interest rates of course might inflict capital gains or losses along the way.)

Now let’s tweak the scenario. Suppose I tell you that Sam plans to raise the money needed to repay his loan by selling services to Tabitha. For example, suppose Sam used the $10,000 loan to buy equipment that he will then use to perform landscaping work on Tabitha’s property over the course of a year. Every month Tabitha pays Sam a fee for his services, and after the 12th month Sam takes these fees, which are equal to $10,500, and hands them back to Tabitha.

In this revised scenario, is it still true that Tabitha acquired a net financial asset when she bought the $10,000 IOU from Sam in the beginning? Yes, of course it is. Tabitha voluntarily purchases the landscaping services from Sam; the flow of money back and forth is a bookkeeping convenience. Economically, what happened is that Tabitha exchanged a stock of present goods up front for a stream of services over the course of the year.

Now let’s tweak the scenario one last time: Suppose that Tabitha lends $10,000 to Sam, who gives her an IOU promising $10,500 in 12 months. After the year passes, Sam walks up to Tabitha and sticks a gun in her belly, demanding $10,500 in cash. She hands it over to him, and then he gives it right back and tears up his IOU.

In this scenario, did Tabitha acquire a net financial asset when she originally lent the money to Sam? No, not really — especially if she knew how he planned on “repaying” her. In this case, Tabitha’s savings of $10,000 would have simply been confiscated by Sam. He can go through the farce of giving her an IOU and then robbing her in the future to “redeem” it, but economically that is equivalent to him simply robbing her of the $10,000 upfront. From Tabitha’s viewpoint, her $10,000 in savings vanished, while Sam’s consumption can rise by $10,000 without increasing his own indebtedness.

Now let’s expand the groups. Instead of the individual Tabitha, consider the group of all Taxpayers. And instead of the individual thief Sam, consider the institution Uncle Sam. The MMTers correctly tell us that the Taxpayers can’t accumulate “net financial assets” — i.e., drawing on income streams that originate outside the group — unless Uncle Sam runs deficits and issues them bonds.

But what is the point of accumulating bonds that will only be redeemed when Uncle Sam coercively raises the necessary funds from the same group of Taxpayers in the future? Any individual taxpayer can justifiably look at a Treasury bond as a net asset, because his or her own tax contributions will not vary significantly based on his or her investment decisions regarding Treasuries. But the private sector as a whole surely shouldn’t naively assume that if the government runs a $1.6 trillion deficit this year, this foretells of a shower of new income flowing “into the private sector” down the road.

I hope I’ve convinced the reader that something is very fishy with the MMT conclusions regarding private saving and government budget deficits. The error crept in at step one, with the equation GDP = C + I + G + (X M). The only justification for measuring “output” (left-hand side) by the summation of spending (on the right-hand side) is that in a market exchange, the “value” of something is whatever the buyer spends on it.

However, if the government can raise revenues through present taxation or by borrowing now and paying back with future taxes, then this justification falls away. It’s simply not true that $1,000 in private consumption or investment spending is an equivalent amount of “real output” to $1,000 spent by bureaucrats who raised the money without the consent of their “customers” and who may very operate under a “use it or lose it” appropriations process.


The MMT worldview is intriguing, if only because it is so different from even the way conventional Keynesians think about fiscal and monetary policy. Unfortunately, it seems to me to be dead wrong. The MMTers concentrate on accounting tautologies that do not mean what they think.

  • 1. In his third bullet point, the MMT writer Bill Mitchell incorrectly referred to (X M) as the “current account balance,” when strictly speaking it is the trade balance since we are talking about GDP rather than GNP. This is a very subtle distinction that is unimportant for this article, but the interested reader can read Greg Mankiw’s explanation.
  • 2. Note too that Google’s lack of a printing press isn’t relevant for the establishment of the accounting tautologies. The MMTers’ sectoral equations are true whether the government has a fiat currency or gold commodity money

Republished from

The Idea That the Fed Is ‘Independent’ Is Absurd

The Idea That the Fed Is ‘Independent’ Is Absurd

President Donald Trump sparked controversy — as is his wont — when he recently told CNBC that he was “not thrilled” with the Federal Reserve’s announced hikes in short-term interest rates, which he claimed would hinder the economic expansion for which his administration had worked so hard. “I’m letting them [the Fed] do what they feel is best,” he added, but this assurance was not enough to prevent journalists and policy experts from pronouncing Trump’s remarks as unprecedented interference with the central bank’s independence.

It may be unusual for a president to openly voice such criticism, but it wouldn’t be the first time one has pressured the Federal Reserve for short-term political gain. In 1965, President Lyndon Johnson considered firing then-Fed Chairman William McChesney Martin, but upon learning this would probably be illegal, he opted instead to dress down the recalcitrant central bank chief at his Texas ranch. By Martin’s later account, a heated argument erupted that resulted in the president shoving him against a wall. According to financial journalist Sebastian Mallaby, as LBJ pushed Martin around the room, he yelled, “Boys are dying in Vietnam, and Bill Martin doesn’t care.”

Better known is President Richard Nixon’s tape-recorded collaboration with Fed Chairman Arthur Burns, Martin’s replacement, who maintained an easy-money policy to stimulate the economy before the 1972 election, which contributed to Tricky Dick’s landslide victory and fueled price inflation for the rest of the decade. In terms of the resulting capital destruction and economic dislocations, this episode is one of modern U.S. history’s greatest object lessons about the risks of executive power reaching beyond its constitutional authority.

For another example showing that Trump’s behavior is nothing new, consider that President George H. W. Bush had a running public dispute with then-Fed Chair Alan Greenspan over monetary accommodation. Bush would later blame “The Maestro” for his 1992 reelection loss.

Read the rest at the

Pin It on Pinterest