The Great American Coin Shortage
What do coins and toilet paper have in common?
Thorsen speaks for many in the local coin-operated economy, a diverse, somewhat old-school community of businesses and consumers that has been in a state of agitation since COVID-19 interrupted the normal cycle of coins.
“It’s something I have to think about all the time,” says Queen Anne resident Dan White, whose apartment has a coin-operated laundry. Early in the pandemic, White had to frantically group-text friends to secure enough quarters for a weekend’s wash.
Another time, after White scored a precious stack from the change machine in a downtown bar, he was followed out and lectured by an employee. “He was like, ‘I’ll let it slide this time but you can never do that again,'” White says.
White is more systematic these days, periodically hitting a handful of businesses to get his rolls for the month. But the routine is time-consuming, with a furtiveness that often feels “weirdly like I was doing like a drug deal or something,” he says. “People that aren’t using quarters for a laundry machine have no idea that this is even happening.”
Since the COVID-19 outbreak, we have experienced periodic global shortages in products as varied as toilet paper, semiconductors, and cream cheese. These can usually be traced back to abrupt shifts in demand combined with disruptions in the supply chain, both of which are linked to the global pandemic. The coin shortage, however, seems to be especially unusual.
First, the coin shortage seems to be an exclusively American phenomenon; although the alleged cause is disruption to coin circulation due to COVID-19, every other country apparently experienced the same COVID-19 disruption without a coin shortage.
Furthermore, most other developed countries are far more coin-dependent than the United States; Europe has 1-euro and 2-euro coins in lieu of paper bills of those denominations, and Canada has their loonies and their toonies. If a coin shortage was going to strike any country, why would it be the US?
Second, one would think that a disruption to normal business activity should create a coin surplus, not a coin shortage. Consumers and businesses hold currency and coin so they have enough on hand for day-to-day transactions; if they are engaging in fewer transactions, they should prefer to hold less currency and coin, all things being equal. Also, if Covid is causing people to shift from cash to electronic payments, they should be getting rid of their coins, not accumulating more.
Data from Europe backs up this intuition; the Eurosystem reports an annual net issuance of nearly 1 billion euro worth of coins in 2018 and 2019, which collapsed to a net issuance of only 400 million euro in 2020 in the wake of Covid. Net coin issuance in Europe actually turned negative in the first quarter of 2021, meaning Europeans were, in aggregate, disposing of coins they no longer needed.
Finally, why has the shortage gone on so long? Many of the Covid-related shortages have been quite short in duration, disappearing as producers and consumers adjusted to new conditions. The coin shortage started at the beginning of the pandemic, and two years would seem to be more than enough time for everyone to adjust. At the outset, government officials expressed confidence that the shortage would be resolved quite soon, but there still seems to be no end in sight.
Our investigation will start with the Global Toilet Paper Shortage of 2020, to put together a potential model.
No one can ever be absolutely certain what caused the toilet paper shortage, but as the Washington Post reported at the time, there were two leading demand-side theories:
1. We’re buying too much toilet paper because we’re panicked there won’t be any when we need it.
2. We’re actually using way more than usual at home because most people are sheltering in place rather than using the facilities while at work, school, restaurants or other public places.
In this simple model, there are only two things that influence consumer demand for toilet paper: how much toilet paper we consume every day, and how much toilet paper we wish to hold as inventory, in our closets at home.
At the time, 4-week toilet paper sales were up 102% year-over-year, implying that either we suddenly doubled our rate of toilet paper consumption, or we built up an extra four weeks of at-home toilet paper inventory, or some combination of both.
Almost exactly a year later, the Wall Street Journal reported that toilet paper sales rose from $9 billion in a normal year to $11 billion in 2020, but 2021 sales were on pace to fall below $9 billion, as consumers worked through the excess inventory they built up in 2020. Clearly, most of that 102% spike was from our sudden desire to hold larger inventories, as a buffer against disaster, although undoubtedly the extra time at home has also elevated toilet paper consumption a bit.
It should be noted that most Covid-related shortages seem to be linked to higher actual end consumption, rather than stockpiling. For example, the most recent product to be in short supply, cream cheese, is actually perishable, and indeed the evidence seems to be that we are simply eating more of it than we used to.
Demand spikes do not typically result in shortages, as producers are usually eager to meet demand by producing more. However, it turns out that toilet paper production capacity is relatively fixed, and cannot easily respond to demand. As the WSJ recently reported:
[W]hereas companies were able to more quickly increase capacity for cleaning products, hand sanitizer and other in-demand items, doing so for toilet paper was less feasible given that making toilet paper in bulk requires four-story-tall machinery that costs billions of dollars and takes months to build.
Toilet paper companies were able to produce a bit more by running extra shifts on existing machinery, but they could not hope to meet a sudden 102% increase in demand.
However, demand spikes and a relatively fixed supply do not usually result in a true shortage complete with empty store shelves, as we saw with toilet paper. Usually prices adjust enough to encourage consumers to reduce or defer consumption and to encourage suppliers to increase production, until supply and demand are balanced.
To take just one example, housing prices since the pandemic adjusted sharply upwards (up 25% nationally), so while the inventory of homes for sale has been low, the market still clears, and homes are available for those able to pay.
Retailers like Costco could have raised prices on toilet paper, but they had little incentive to do so. Retailers rely on repeat business and reputation, which they do not want to jeopardize by appearing to profit from a crisis (and toilet paper is such a miniscule business for them, the extra profit from higher prices would have been very low). Most retailers were content to set purchase limits per customer and allow their shelves to empty temporarily.1
In summary, we had a spike in toilet paper demand from the sudden desire of consumers to build inventory, coupled with inability of toilet paper manufacturers to meaningfully increase production (at least in the short-term). That might have simply resulted in high prices, but retailers prefer to keep prices steady, so instead we ended up with empty shelves.
Before we move on to coins, let us first analyze money more broadly defined, which includes coins, paper money, and bank accounts.
In a way, money is like toilet paper. Demand for money fluctuates depending on how much inventory people want to hold. In most models, consumers try to hold enough money so that they can conveniently buy what they need to, but they also may choose to hold money as a store of wealth, instead of stocks or bonds.
Usually you will not want to hold too much money, since you are forgoing interest you could be earning if you put the money into bonds (for example). Lower interest rates lower that opportunity cost, so people are more willing to hold money today, with interest rates near zero.
Here is a graph of the historical “velocity” of money, which is the ratio of GDP to money supply. (It is called velocity because it measures how many times each dollar is spent per year on a final good or service; hence, it measures how fast money passes through the economy.) If people are holding more money, it will be reflected in lower velocity.
This graph is upside down; a lower ratio means that people are holding more money, and thus money is circulating more slowly.2 We can see that during the financial crisis of 2008 and the Covid outbreak of 2020, there were sudden spikes in demand for money, as people simultaneously decided they would feel more comfortable with more money in the bank, just as they also suddenly felt that they would be more comfortable with more toilet paper at home in early 2020.
In 2003, the late economist Milton Friedman observed that inflation had been quite volatile through the mid 1980s, even though velocity had been very steady. He notes that this is quite odd when you consider the quantity equation of money: PY = MV.
In this equation, P is the price level (which, when it goes up, is called inflation) and Y is the quantity of goods and services produced in the economy, and on the other side, M is the money supply and V is money velocity.
Simple algebra suggests that a stable velocity (V) should make it easy to produce a steady price level (P). This is the exact opposite of what actually occurred, which was a volatile inflation rate that reached double digits at times.
Perhaps more perplexing is that velocity has apparently become much more volatile since then, but inflation suddenly became very steady by historical standards, as we see below.
Friedman’s explanation is that central banking policy, properly executed, is like a thermostat. A thermostat does not need to know how cold it is outside, it merely targets an indoor temperature and turns the heat up or down if it is missing its target. A thermostat doesn’t care if it is 72 degrees in January or if there is a May snowstorm, it just needs to know the current indoor temperature and turn the heat up or down as necessary.
Likewise, a central bank does not need to know anything directly about the demand for money or the money supply. It recieves signals about inflation and economic conditions and tighten or loosen monetary policy accordingly. For example, the 2008 financial crisis might be analogous to a May snowstorm, and unconventional monetary policy like QE would be analogous to running the heater at record levels in response.
His theory is that central banks learned to target inflation like a thermostat, and ended up with much better results than when they had other theories about how to control inflation.
Central banking is a hot button topic, so we will consider it only to observe that the Federal Reserve tries to print just enough money to keep inflation constant and consumers adequately supplied, just as Kimberly-Clark and Procter and Gamble try to produce enough toilet paper to keep the price of toilet paper constant and consumers adequately supplied.3
The difference between the two is that the Fed can “print” money out of thin air with the push of a button, while Kimberly-Clark is constrained by the scarcity of heavy machinery required to print toilet paper. It should be no surprise that we have generally been able to get our hands on money when we need it, but the same cannot be said for toilet paper.4
In July 2020, in response to the outcry over the coin shortage, the government convened the U.S. Coin Task Force. The U.S. Coin Task Force studied the matter and pronounced that there are actually plenty of coins available, but they are just stuck, somehow. The official government statement on the coin shortage begins:
There is currently an adequate overall amount of coins in the economy. But business and bank closures associated with the COVID-19 pandemic significantly disrupted normal circulation patterns for U.S. coins.
The U.S. Coin Task Force created a website complete with video PSAs and hashtags (#getcoinmoving), urging consumers to get out and return their coins. Their statement goes into a bit more detail:
Coin circulation has emerged as a new disruption caused by the COVID-19 pandemic. Many have referred to this as a shortage; however it is not. There is more than $40 billion in coin already in circulation, most of which is sitting dormant inside America’s 128 million households. As people have changed their spending habits, and coin-intensive businesses and financial institution lobbies have been less accessible, the nation’s coin is pooling in change jars, in car cup holders and in shuttered businesses, making it difficult for the businesses of this country to get the coin that they need to support cash transactions.
This statement should leave one a bit skeptical. $40 billion of coin would be over $300 of coins per household; there is no way the average household has $300 in change lying around, “dormant”. Also, business has mostly returned to normal, but news outlets are clearly still reporting a general shortage, even while the U.S. Coin Task Force insists that no shortage has ever existed.
One clue as to the true nature of the coin shortage lies in the reporting on the crisis, which like our original Seattle Times story, consistently and inevitably comes back to coin-operated laundries. The Wall Street Journal wrote an article on the shortage back in August:
The night before she was scheduled to serve jury duty, Kuromi Hendricks realized the clothes she planned to wear were dirty. She couldn’t find the 12 quarters required for a wash and dry at her Boston apartment building.
By 11 p.m., desperation took hold. She hopped in a $10 Uber to the only place she knew was open and might have spare change—a 24-hour cafe with a couple of arcade games. Not wanting to use the business just for their coins, she ordered a lemonade, asked for about $5 in quarters, played a few rounds of pinball, and headed home to wash.
Aeryn Emmerich-Wise and her husband piled four loads of dirty laundry into their car for the nine-hour drive from the borough of Queens in New York City to visit relatives in North Carolina, after both sets of in-laws encouraged them to do so. Before that, Mrs. Emmerich-Wise didn’t wash her favorite pair of jean shorts for about three months. She finally gave up and ordered a new pair because she was having such a hard time finding quarters.
And here is a Philadelphia Inquirer story from last year:
Roughly once a week, Marta Rusek’s studio apartment is adorned with damp clothes laid out to dry.
With coins scarce all over the country, the Center City resident has favored air drying rather than the coin-operated laundry machines in her Spruce Street building.
“Basically,” she said, “my quarters are very near and dear to me.”
Rusek, 35, washes much of her wardrobe in her bathtub, much as she did in 2011 while serving with the Peace Corps in Gambia.
Most Americans rarely think about coins, but a subset of Americans are wholly reliant on quarters to be able to have clean clothes. The demand for coins seems to be less about making change at the grocery store and more about demand from coin-operated laundromats, the majority of which currently have no way of taking cash or any kind of electronic payment.
At the beginning of Covid, the U.S. Mint temporarily slowed down coin production, as they implemented social distancing. This would likely have caused some people to have trouble getting their hands on coins. If you really need coins, you might respond to a temporary disruption by hoarding coins, as a buffer against future disruption. Reported Bankrate:
When SAFE Credit Union heard the Federal Reserve would reduce the number of coins [a] financial institution could order, Baldi remembers thinking “whaaat?” While she was surprised, she knew the credit union didn’t want to experience the financial equivalent of a toilet paper shortage. Then, she had a quick aha moment. The Fed restrictions were per location and the credit union had only been ordering for one. “Suddenly, we discovered we had 20 other locations that could be used,” Baldi says.
So SAFE ordered coins for all their locations and stored them in a third-party coin vault to avoid overwhelming the individual branches. “We had almost no impact from it from a member standpoint,” Baldi says.
The only way for the SAFE Credit Union to fill their vaults with coins is by removing coins from circulation; in this case, they are taking new coins that would have ordinarily entered circulation and instead placing them in their vaults. It seems perfectly reasonable to assume that every other coin-dependent business and consumer had the exact same idea at the exact same time. If everyone tries to hoard coins simultaneously, then the result will be a shortage of coins in circulation, just as toilet paper hoarding resulted in a shortage of toilet paper on store shelves.
This is also a good illustration of the bullwhip effect, whereby small fluctuations in demand get amplified through the supply chain. It might not be so bad if it was just consumers and laundromats ordering directly from the U.S. Mint and building inventories for themselves, but distribution is intermediated by banks, who take it upon themselves to distort their own orders and build their own additional coin stockpiles, worsening the problem. This is no different from what retailers and distributors of ordinary goods like shoes are reportedly doing today.
To the credit of the U.S. Mint, they have been hard at work trying to produce enough quarters to satisfy the spike in demand. According to the Philadelphia Inquirer:
At the Philadelphia Mint — the nation’s largest producer of coin currency — 14 presses, each producing 750 coins a minute, are running seven days a week to compensate for the pandemic-caused coin supply problems that turned quarters, nickels, and dimes into rare commodities.
“I’ve never seen anything like this in the past,” said a veteran Mint worker who maintains the coin presses and asked not to be named because he was not authorized to speak publicly.
During each two-week pay period, he said, “I’m putting in about 30 hours' overtime.”
We can presume from this article that coins are as difficult to produce as toilet paper, with the only flexibility available being the ability to run extra shifts. The U.S. Mint reported shipments of 2.9 billion quarters in 2020, up from 1.8 billion the year before. They managed to increase annual production of quarters by 60%, despite falling behind in the first half of the year.5
And yet, 1.1 billion extra quarters is a mere $275 million worth of coins. The coin-operated laundromat industry does nearly $5 billion of revenue per year, so $275 million is only 3 weeks worth of quarters, split between everyone in the supply chain, including consumers, laundromat owners, and banks.6 It is easy to understand why the coin shortage lingers on, nearly two years later.
At this point we must consider potential solutions, and we will start with Patrick McKenzie’s original tweet; he believes that we should let the quarter trade above face value, to balance supply and demand. He is of course correct that this would solve the issue, as surely the SAFE Credit Union and other hoarders would reconsider if they had to pay a large premium to get their hands on extra coins.
Also, if everyone felt that they would always be able to get their hands on coins when they needed them, just by paying a premium over face value, they would feel less need to stockpile quarters.7
This solution has the same issue we identified with retailers of toilet paper; as with retailers, the U.S. Mint is incentivized to avoid bad publicity, hence the press releases telling us that the coin shortage is not really a coin shortage. Charging banks fifty cents for a quarter likely will not give them the publicity they are looking for.8
Furthermore, moving down the supply chain, it is not clear that most businesses even have the ability to charge consumers a premium for a quarter. They rely on change machines that are probably only capable of giving four quarters for a dollar.
Finally, there is an online market for quarters ($79 + shipping for $50 of quarters on eBay as of this writing), and banks and businesses have offered bonuses to get consumers to turn in coins. The existence of an active black market for quarters does not seem to have solved the issue.
The libertarian Cato Institute, in typical fashion, suggests that the true problem is that the government is preventing the private sector from executing a supply-side solution:
Although history is rife with examples of the private sector stepping in to provide alternative money when official currencies have been in short supply (e.g., Selgin 2008; Champ 2007), that didn’t happen here. Why not? The most likely answer is that the U.S. government is not one to tolerate competition.
Perhaps the government usually does not tolerate competition, but anyone who thinks that private businesses cannot create their own coins and tokens has never been inside a casino, or even a Chuck E. Cheese.
The problem with a private sector supply side solution is that laundromats use quarters precisely to avoid having to manufacture their own tokens at great expense. Laundromats already have the ability to solve their quarter dependency by installing machines that take cash and credit card, no coin manufacturing needed. The Seattle Times article suggests that in many cases they already are, but it is a process that takes time and costs money.
There are other potential supply side solutions. We could perhaps build extra manufacturing capacity at the Mint, to handle future spikes in demand, or we could mint enough coins to establish a Strategic Quarter Reserve, to be tapped whenever businesses and consumers want to hoard quarters.
The problem with supply side solutions is that they are bound to be very expensive, relative to the coin-operated laundromat industry they would primarily be supporting. We might be fine with spending hundreds of millions of dollars a year to support a $5 billion a year industry, but we would like to find a cheaper or simpler solution if we can.
If the supply side is out, perhaps we can find a solution on the demand side.
The challenge is that the demand for coins for use in transactions has been declining, as we switch to electronic payment; as we see below, the increase in coin production in 2020 is barely visible against the longer secular decline in coins (at least when considered relative to population and economic growth). In the meantime, the coin-operated laundry industry has continued to grow at the same pace as the broader economy.
Laundromats and other coin-operated machines piggybacked off of the coins we were already using every day,9 but as electronic payment gains share and inflation renders coins obsolete, we make fewer coins for everyday use. Machines that are exclusively coin-operated are hostage to the general availability of coins.
One possible solution is to subsidize any laundromat that wishes to eliminate their dependency on coins and install an electronic payment system. It costs hundreds of millions of dollars a year to manufacture extra coins; presumably it would be cheaper to just pay once to install electronic payment systems at a few thousand laundromats. In this case, it is probably easier to reduce demand than to increase supply.
If we are feeling less generous, we could simply mandate that laundromats of a certain size accept cash or electronic payment by a certain date, and perhaps couple it with mandates regarding parking meters and other coin-operated machines. If most vending machines today accept cash, surely it is possible for most other machines to accept cash or card with a small investment as well.
There is one other solution we must consider. I must caveat this by saying that this is my least favorite idea, but we must examine it for the sake of completeness.
Most other developed countries have high value coins in common use. Japan has a 500 yen coin (about $4.40), and Europe has a 2 Euro coin ($2.25). Canada and the UK both have coins worth about $1.50. The US is something of an outlier, with a maximum coin value of $0.25.
It was not always this way. The last time the US had a coin shortage was the early 1960s, and that came about in the traditional way, because the value of the silver contained in a quarter approached face value, causing people to hoard the coins to be melted down later.10 Adjusted for inflation, 25 cents back then is equivalent to nearly $2.25 today, so the quarter was a high value coin by modern standards.
If you have a high value coin, then curing any shortage becomes much simpler, because you get more value per coin minted, by an order of magnitude. We mint billions of extra quarters and barely make a dent in our shortage, because we are all running around with rolls of coins, heaping them by the dozen into washing machines and dryers. If you go to a laundromat in Europe, you can do a load of laundry for about three coins.
This may also explain why the coin shortage is unique to the United States. High value coins mean that coins are more frequently used in everyday transactions, which makes it easier to absorb fluctuations in demand from the users of machines that require coins. There are many ways to substitute away from coins in regular transactions, from nudging customers toward electronic payment to adjusting prices, which frees up coins for everyone else. Also, the machines that use coins become much more coin-efficient.
The problem is that coins are not user friendly. Coins are inconvenient to store and to use. High value coins have been phased out over time in favor of paper money, mostly through inflation. Efforts to introduce a dollar coin alongside the paper dollar have consistently fallen flat. High value coins may save the government money by replacing short-lived paper money with long-lived coins, but at the cost of inconveniencing the user.
From the point of view of coin-operated laundromats and their customers, however, phasing out dollar bills in favor of dollar coins should solve the problem of future coin shortages. Laundromats would have to install new coin machines that accept dollar coins, but once that was complete, they would be more resistant to future coin shortages.
The question of the dollar coin comes up from time to time, and the last effort in 2012 was led by the Dollar Coin Alliance, an organization made up of:
an eclectic coalition of metal companies, mines, merchandisers, vending interests, the United Steelworkers union, a government watchdog group and even an association for car washes.
This is exactly what we would expect; the organizations that favor high value coins are those that supply the materials for those coins and those that run coin-operated machines.
And who is to blame for defeating the dollar coin? As Bloomberg reports:
This led some laundromats to join the armored-car industry and paper money–printing industry in Americans for George, which lobbied (successfully) to preserve the print dollar in 2012.
It's a divisive issue on Planet Laundry, the world's leading forum for laundromat owners. On the one hand, some laundromat owners fear that the quarter is going to be destroyed by inflation: Soon enough, people will treat it like the penny and simply stop carrying it, or laundry will require so many quarters that the costs associated with capturing them will rise. On the other hand, it could be expensive for owners of older laundromats or laundry machines to convert to machines that accept dollar coins.
In all likelihood, consumers mostly ended up transferring inventory from store shelves to their own homes, with very little impact from greater production or consumption.
M2 is just one specific measure of money supply; other measures only consider currency, or include different kinds of transaction accounts. Velocity is sensitive to how broad your measure of money supply is, so do not take this graph to be that important. As technology advances, the cost of “going to the bank” changes (it becomes virtual) and different financial products are invented, both of which can be expected to change velocity quite significantly.
Extending the analogy, one might think that the Fed should be no more concerned that printing money during a crisis will lead to a devaluation of the dollar any more than the executives at Kimberly-Clark were concerned that running the production lines 24-7 during Covid would lead to a devaluation of toilet paper. Sure, demand will probably fall in the future, but there are obviously ways to handle that.
Jay Powell has a tough job, but let us take a minute to consider the unsung heroes throughout our supply chains, who have no way to conjure solid goods out of thin air in the same way that the Fed can create money.
The production of other coins grew by much lesser amounts, suggesting that this is mostly a story about quarters and laundromats.
This is a very rough estimate, assuming that laundromat users are the only ones interested in hoarding quarters; in reality, there are other businesses that rely on coins, from arcades to convenience stores, although we must also consider that some laundromats do not require quarters.
Noah Smith links to a 1989 paper by Harvard economist Martin Weitzman, which uses a then-current example Soviet soap shortages. At the time, soap was hard to find on store shelves in the Soviet Union despite normal consumption patterns and high production. The culprit was the extra demand from hoarding encouraged by price controls, exactly as we see today in quarters:
"Hoarding psychology," or what might better be called "defensive hoarding," can be thoroughly analyzed as an economic phenomenon by extensions of standard economic theory. In shortage equilibrium, everyone must expend effort to locate and hoard goods because everyone else is expending effort to locate and hoard goods.
Some cite breakdowns in the distribution system. (Railroads seem particularly to be accused.) Others blame a "hoarding psychology" that causes panic buying and is somehow related to the deficit and monetary overhang. Theft by workers, sabotage, and speculation by cooperatives are also candidates. The officials seem unified only on promising increased production to meet the shortage and on calling for formation of committees to investigate formally the problem.
Some informal investigation reveals an interesting, if perhaps not unexpected, fact. Although few official figures are available, observations, conversations, and anecdotes suggest strongly that Soviet people are hoarding soap and other commodities in massive amounts. Significant parts of bathrooms, closets, hallways, and other areas have been given over to storage.”
The Federal Reserve does place limits on the amount of coins that bank branches can order, roughly based on their historical order patterns. As we saw with the example of the SAFE Credit Union, which ordered for twenty branches when historically they had only ordered for one, limits are a very imperfect way to ration scarce goods, and coins inevitably get hoarded by those who do not really need them.
There is an additional mystery here as to why we use small denomination coins at all. When the half penny was discontinued in 1857, the penny became the smallest denomination; $0.01 then would be $0.32 today, adjusted for inflation. Even by WWI, a penny then would be worth the same as a quarter now, adjusted for inflation; a penny in 1960 would be worth a dime today. Why do we need any coins worth less than a quarter or a dime now, when we didn’t need them back then?
Many countries have recently gotten away from the penny (or its local equivalent), but that seems to stop well short of what could be done to get rid of small change. Most coverage seems to focus on the role of special interests, e.g. the role of the zinc lobby in preserving the penny, but there is likely some form of psychological anchoring involved as well. Tyler Cowen once observed that the average nominal price of a share of stock (which is generally chosen by the issuer through stock splits) is roughly the same today as it was a century ago (around $50), but $1 then might be equivalent to $20 today, so the average stock price back then would be over $1,000 in today’s money. The same principle likely applies in both cases, which is that we usually anchor to nominal rather real prices.