In 1977, Joan and Richard Sweeney wrote an article called “Monetary Theory and the Great Capitol Hill Baby Sitting Co-op Crisis” which describes the economic woes of a certain babysitting co-op whose members traded scrip between them in exchange for babysitting one anothers’ children. In 1998, Paul Krugman wrote a commentary on it called “Baby-Sitting the Economy.” He interpreted the events it describes as a microcosm for the economy as a whole, and he analyzed it according to his Keynesian framework. Finally, in 2014, Pascal-Emmanuel Gobry confidently predicted Bitcoin’s failure based on his application of Krugman’s analysis to Bitcoin.

This is a confused and convoluted topic, and there are a few separate issues to unravel in all this. There is the original article, which makes some vague comparisons to the U.S. economy, that is very modest about the broader implications of the events it describes. Then there is Paul Krugman’s Keynesian interpretation of it. Finally, how does Bitcoin fit into everything? I will show that both Krugman’s article and Gobry’s follow-up are confused, do not establish their case, and say nothing about Bitcoin.

First off, I enjoyed the original article. It is a well-written, fun piece, which is brief and incisive. To summarize, the members of the Capitol Hill Baby Sitting Co-op were all issued scrip, which they all pledged to accept from other members as payment for babysitting one another. An officer of the co-op would accept notifications from the members about their availability to babysit as well as their need for sitters and would match them up to one another. In other words, the scrip acted something like a currency that enabled all members to trade babysitting time.

Of course, the number of available babysitters did not always match up with the number of people who wanted to go out in a given evening. The Sweenys describe periods of both chronically too many babysitters and too few. These imbalances were caused by the way people valued the scrip. Within the co-op, the value of having a balance of scrip is flexibility. One can go out several times at one’s leisure without worrying about replenishing the supply immediately if one does not have time to provide care for other children. Everyone desires a certain flexibility, so there is a need for a certain amount of scrip per capita to be issued. Clearly it would be very inconvenient if one unit of scrip were available — no one could babysit or go out twice in a row and no two couples could go out at the same time. Everyone would want more scrip, and the one holding it would not be very willing to spend it—this is what is known as a liquidity trap. On the other hand, if there is too much scrip per capita, then everyone feels flexible enough and would prefer to go out rather than try to accumulate more.

When there was too little scrip,

holders were reluctant to squander it by going out. Those who wanted to go out but didn’t have scrip were desperate to get sitting jobs. The scrip-price of baby sitting couldn’t adjust, and the shortage worsened. The co-op even passed a rule that everyone must go out at least once every six months. The thinking was that some members were shirking, not going out enough, displaying the antisocial ways and bad morals that were destroying the co-op.

And when there was too much (which was the case at the time the article was written),

the price of baby sitting is constitutionally pegged at one unit of scrip for every one-half hour of baby sitting. Hence, this system of price controls means the inflationary pressure does not drive up the scrip-price of baby sitting, inflation is suppressed, and shortages are found.

The scrip was created and destroyed according to various provisions in the co-op’s rules. The co-op managers seem to have had little inkling that there would be an optimal supply of scrip in their organization and did nothing to ensure that the amount of scrip per capita which was created matched that which was destroyed each year. Early in its history, the supply of scrip was too small and decreasing, and later it was too large and increasing. Consequently, although they once briefly hit the “sweet spot” that balanced the supply and demand for babysitting within the co-op, they could not maintain it very long.

There at least two possible ways of dealing with a suboptimal supply of scrip: the supply of scrip per capita could be adjusted or the price of the scrip could be adjusted. If everyone has ten units of scrip worth an hour of babysitting, that is the same as if everyone had twenty units each worth half an hour. Thus, if people felt as if they had accumulated enough scrip and preferred to spend rather than save it, the co-op board could require everyone to turn in some scrip or declare that each unit is worth less babysitting time. There is not necessarily a single ideal supply of scrip that would work permanently because people will have different babysitting needs at different times, so there would have to be continual adjustments.

The co-op, of course, operates as a centrally planned arrangement characterized by strict price controls. As long as the price is fixed, then the central board must meet to adjust the price as needed. Yet although the Sweenys clearly and repeatedly identify price controls as being fundamental to the co-op’s problems, Krugman does not mention that his interpretation of the events is very different from the Sweenys and does not present price changes as a possible cause of the co-op’s problems.1 He assumes that a change to the supply of scrip is the only possible solution, and in fact, the word price does not appear in his article even once! This is rather odd because the Sweenys explain all of the co-op’s economic woes in terms of price controls, so Krugman clearly should have responded to this in order to establish his own interpretation of the events.

Of the babysitting co-op’s story, Krugman says:

Its story tells you more about what economic slumps are and why they happen than you will get from reading 500 pages of William Greider and a year’s worth of Wall Street Journal editorials. And if you are willing to really wrap your mind around the co-op’s story, to play with it and draw out its implications, it will change the way you think about the world.

However, he has utterly failed to show that the co-op is analogous to the American economy and, granting that it is a valid analogy, has failed to draw out its implications as a consequence of price controls. For if indeed the co-op’s problems can be explained by price controls, then what is the relevance to the American economy? The American economy is at least relatively free of regulation and could be made even more free. So in order to make the co-op into an analogy, Krugman would either have to argue that American recessions are caused by central planning, or that the Sweenys are incorrect in their economic analysis of the co-op, or that despite having different causes, American recessions have the same effects and cures as that of the co-op. He does not bother with any of that.

Even if he did manage to establish an analogy, we may still ask, “Why not simply eliminate the price controls?” Ceteris paribus, if the co-op board eliminated the price controls entirely and allowed people to set their own prices, then people could respond to imbalances in the demand and supply of babysitters by offering different scrip prices. The final result would be similar to one in which the central planning board had declared a price change, without any centralized decision being required. Krugman has done nothing to show in his article that deregulation, rather than monetary policy, is a valid response to a recession. One might argue that he does not need to show that because ideas such as the liquidity trap are well established in economics. But I still claim that his use of the babysitting co-op is a red herring which teaches nothing about Keynesian economics.

Gobry’s piece does nothing to resolve the problems with Krugman’s article. When he says, “The more people tried to hoard coupons, the less people were willing to go out and get their babies sat,” he is still talking about a problem that can be entirely explained in terms of price controls, and of which no relevance has been shown to Bitcoin. Just imagine if, for some reason, we still had to pay 5,000 bitcoins for a pizza. Clearly this would cause a severe disruption to the Bitcoin economy. There would not be enough bitcoins to go around and bitcoins would be so inconvenient that no one would want to improve the bitcoin infrastructure. That would be a very severe liquidity trap! But because the exchange rate of bitcoins is set by the market rather than fixed, nothing like this has ever been a problem.

Now I let Krugman off the hook for disregarding deregulation as a solution, because that may not have been important to his audience or intentions with the article. But Gobry is most certainly not off the hook, first of all, because there aren’t a lot of Keynesians into Bitcoin in the first place, and second, because the Bitcoin economy is not centrally planned and has never experienced a liquidity trap — exactly what an anti-Keynesian would expect! Gobry ought to provide evidence that his theory will apply in the future even though it has not in the past, but the way he deals with this problem is totally inadequate.

Gobry lamely cites Bitcoin’s future release schedule:

“Once the supply of Bitcoin is fixed, at some point the Bitcoinomy will run into the same problem as the baby-sitting co-op: there won’t be enough currency, and there will be a recession, and there will be a liquidity trap,” he says.

At the time of this writing, demand for Bitcoin has gone up roughly 160,000 times, as measured by its price, since the day that two pizzas were bought for 10,000 BTC. Moreover this price increase corresponds to the growth of the Bitcoin economy. Yet there are now only about three times as many bitcoins that have been claimed as in those days.

Can that relatively tiny increase really explain economic growth by a factor of 160,000? These numbers are obviously way out of proportion, and I doubt that any similar multiplier effect has been observed anywhere else. The observed history of Bitcoin is very bizarre from a Keynesian perspective but far less so from that of the Austrian. Why has no liquidity trap occurred? Bitcoiners have already had to adjust prices by enormous factors that are virtually unaffected by the increase in Bitcoin’s supply, so why shouldn’t they be able to do this again as necessary once Bitcoin’s supply is fixed? If Gobry wishes to show they cannot do this, he really ought to provide some evidence.

Furthermore, because everyone knows Bitcoin’s future release schedule, everyone behaves now in expectation that more bitcoins will be in circulation in the future. Therefore the newly mined bitcoins should not be expected to have any effect on the Bitcoin economy to prevent liquidity traps. In an economy in which everyone knows that new coins will be released soon, everyone will just try to save more than they would otherwise because they would know that their savings will have less of an effect than they might otherwise expect. Bitcoin ought to be experiencing a liquidity trap already, yet Bitcoin prices have instead consistently changed to reflect its growth.

I close with a side-issue. Would the original babysitting co-op have solved its problems by allowing for market prices? I’m not so sure. There is an important reason that both the American economy and the Bitcoin economy both differ from the babysitting co-op. The babysitting co-op is a contractual arrangement and the babysitting scrip is not a currency and it does not function as money. The value of the co-op is enabling people to receive babysitting from within a group of people who already have some trust for one another. However, this is not their only option: they can also leave the co-op and pay teenagers with dollars to babysit. Above, when I described such a possibility, I said “ceteris paribus” but, in fact, all things are not equal, because changes to the rules of the co-op can change its competitive advantage over its alternative. In order to stay competitive, the members of the co-op must agree to some obligations to one another, and I would argue that agreeing to honor the scrip upon certain terms is a necessary part of that obligation.

As long as they have all pledged to accept scrip at a certain rate for babysitting, then they have given some real evidence that the co-op is worth joining and holding its scrip is worthwhile, but without a fixed price, there is effectively no such pledge. It is possible that some system of rules would enable the system to function still, but merely eliminating the fixed price for scrip would transform it into nothing but an appcoin. Consequently, there would be a self-reinforcing trend of members reducing their holdings of scrip in expectation that others will do the same, and demanding higher prices for the same reason. This in turn would reduce the expected future value of the co-op, thus induing people to leave it, thus, further reducing its value. The co-op would not be able to maintain itself on those terms.

But once again, there is no corresponding problem with Bitcoin. Bitcoin users have no need to make pledges or guarantees to one another because Bitcoin works without requiring a contractual relationship between its users. People can be expected to follow Bitcoin’s conventions (like accepting the greatest-difficulty chain) without making promises to one another, because Bitcoin has its incentives in the right place. Because there are real reasons to expect Bitcoin to be useful as money and for the Bitcoin network to grow, people should be expected to demand continually lower prices in Bitcoin, not higher ones.

Thus, a babysitting co-op is very different from a money economy (Bitcoin or otherwise). It is different for reasons that have been noted neither by the Sweenys, nor Krugman, nor Gobry. However, both Krugman’s and Gobry’s arguments depend on an invalid analogy between the two, whereas the Sweenys’ do not. Krugman misrepresented or misunderstood the original article and failed to discuss the price control imposed by the co-op, which means that he fails to show that the co-op’s situation has anything to do with the American economy. Finally, Gobry has merely repeated Krugman’s claims without showing any relevance to Bitcoin, and failed to deal cogently with the evidence that Bitcoin is not at risk of a liquidity trap.


  1. I suspect that Krugman did not go back and review the original article when he wrote his 1998 column because he makes a few errors that he probably would otherwise have caught. For example, he says that one unit of scrip was worth an hour of babysitting time, when it was really worth one half-hour. He also cites the article incorrectly and gives its year of publication as 1978 instead of 1977. He does not quote from the original article to support his case or note that his interpretation of the events differs somewhat from that of the authors. None of that has any necessary effect on the logic of his position, but I think these observations suggest that his column may be strongly colored by his own preconceptions without a fresh look at the original piece. Gobry also shows no evidence of having read the original article. 





comments powered by Disqus