Hello. I’d like you to imagine for a moment that you a sheep farmer. Not some fancy, 21st-century sheep farmer with heritage breeds and heated barns – just your standard, hard-working sheep farmer living hundreds of years ago, in medieval Europe. OK, are you there, in your mind? Now, imagine that just down the road from you is a rival sheep farmer. The two of you have never gotten along. And now he’s accusing you of stealing some of his sheep. You are arrested and sent to court.
Peter LEESON: … and the court or the officials don’t have any real reason to think that the neighbor would make this up.
But nor is there enough evidence to convict. The judge doesn’t want to send an innocent man to prison, but he also doesn’t want to let a criminal go free. So he presents two options: You can either plead guilty or your case be turned over to a church court for a trial by ordeal.
LEESON: There were two basic types of ordeals in the period in question: ‘hot ordeals’ and ‘cold ordeals.’
That’s Peter Leeson.
LEESON: Hot ordeals consisted of trials by water. In which case, what they would do is —and by ‘they’ I mean clerics, it’s clerics, priests who were administering all of these ordeals—boil a pot of water, throw a stone or a ring into it, ask the defendant to plunge his arm into the water and pluck out the stone or the ring. Then they would wrap up the defendant’s arm and revisit it three days later. And if it was determined by the priest to be what they would called “foul” within the wrapper — which is to say, showing serious signs of having been burned — the idea was that the defendant was guilty of the crime. And if there was no signs of injury, then he was considered innocent of the crime.
Stephen J. DUBNER: OK. So, when we look back from our modern perspective, or modern as of today, at least, and kind of mentally ridicule these things as, you know, barbaric, or at the very least counterproductive, were they? Counterproductive?
LEESON: My argument is that they were not. My argument is that the ostensible purpose of ordeals, which was to find fact in criminal cases, is in fact what ordeals did, and they did so quite successfully.
* * *
ANNOUNCER: From WNYC, this is Freakonomics Radio, the podcast that explores the hidden side of everything. Here’s your host, Stephen Dubner.
Today’s program is about — well, it’s about “teaching your garden to weed itself.” Which, I realize, is a phrase that doesn’t make any sense to you — yet. But it will. We begin with Peter Leeson, who’s a professor of economics and law at George Mason University. Leeson has long been interested in the kind of topics that many economists aren’t so interested in. Anarchy, for instance.
LEESON: That work led me to some work on pirates, who are, of course, criminals and in consequence can’t rely on government to promote social cooperation within their organization. But I’ve also looked at the early modern institution in England of wife-selling, of human sacrifice, of the legal prosecution of insects and rodents. And a whole host of, I guess, things that most people would describe as peculiar, anyway.
Among these peculiar topics: trial by church ordeal in medieval Europe. Now from a modern perspective, a trial by ordeal sounds anything but rational. It couldn’t actually succeed at separating the guilty from the innocent, could it? Or could it? To find out, Leeson went looking for some data. One set of church records from 13th-century Hungary included 208 cases in which a defendant – like you, the accused sheep farmer — was summoned by a priest, led into the church, and was instructed to grab hold of a smoking-hot iron bar.
DUBNER: So if you were to just stop someone on the street, Peter, and say, “Hey, here’s a historical quiz: Two hundred eight people were sentenced to a trial by ordeal some many many centuries ago, and they had to grab a piece of hot iron that a priest was overseeing. How many of those 208 would you suspect were burned?” Most people would say, “I assume 208.” Yes?
LEESON: That’s what I would think, yes.
DUBNER: So tell us in reality the number.
LEESON: Two-thirds of these cases, of the 208, involve exoneration, which means that the defendant is found innocent because the supposedly red-hot burning iron didn’t burn him.
DUBNER: OK, so two-thirds of the more than 200 people who are commanded—in a church, by a priest—to grab onto a piece of red-hot iron are not burned.
LEESON: That’s right.
OK, how can this possibly be? How were two-thirds of the defendants not burned by a hot iron bar? Did God exonerate the innocent and punish the guilty? That’s not how Peter Leeson sees it. He explains this in two words: “Priestly rigging.” That’s right, the priests were arranging things so that most defendants who accepted the ordeal wouldn’t get burned by the hot iron bar or a cauldron of boiling water. Now, why would that be? Were the priests simply exercising a bit of human mercy? Did they maybe take bribes from some defendants? Not according to Peter Leeson. Here’s what he thinks was happening. Most people at the time likely believed in an almighty God who knew whether a defendant was guilty or innocent – and, accordingly, would burn the guilty man but protect the innocent man. For a defendant, knowing that God knows what you did or didn’t do, would affect your behavior. It changes the incentives.
LEESON: The key here is that because the priests know that the innocent person’s incentive is to undergo the ordeal, they also know that on the other side of it, the guilty person’s incentive is to decline the ordeal. The reason for that is exactly the flip-form of thinking. So now imagine that you did steal the sheep. Now you’re thinking, “Well, I know that if I undergo the ordeal, if I put my arm in the boiling water, I’m going to have my arm boiled to rags, because I am in fact guilty. God’s not going to perform the miracle. And in the process, on top of that, I’m going to be convicted of the crime.” It’s better for me to simply either settle with the accuser or to confess to the crime and enjoy a somewhat less harsh punishment.
DUBNER: OK, so that explains why, if I were guilty, I would decline the ordeal and accept my penalty. But if I’m innocent, I would undergo the ordeal. And then what?
LEESON: Well, that’s the key thing. So the priest now knows that the incentive of the innocent person only is to undergo the ordeal. The guilty person is going to decline. Because the priest knows that, conditional on you being willing to undergo the ordeal, you reveal, if you will, this private information that you have about your guilt or innocence. You reveal the fact that you’re innocent to the priest. Now, in order, of course, to be exonerated, the water needs to not boil you. And so the priest’s job is, conditional to knowing that you’re innocent, is to turn down the dial on the stove, so to speak, to ensure that the water doesn’t boil you and exonerates you, as you expect.
If Peter Leeson is right – and there’s no guarantee of that – then the medieval trial-by-ordeal was, rather than a barbaric expression of divine justice, a rather brilliant means of sorting the innocent from the guilty. In economist-speak, this is known as breaking down a pooling equilibrium into a separating equilibrium by using game theory. In our book Think Like a Freak, we give this practice a different name. We call it “teaching your garden to weed itself.” And that’s what today’s show is about. Let’s begin with two men, separated by many centuries. King Solomon built the First Temple in Jerusalem and was known throughout the land for his wisdom. David Lee Roth fronted the rock band Van Halen and was known throughout the land for his prima-donna excess.
[Live concert audio with applause]
David Lee ROTH: I’m gonna tell you baby. Rock ‘n’ Roll is my second favorite thing in the whole world.
[“Runnin’ with the Devil” by Van Halen]
Now let me ask you this: what could David Lee Roth and King Solomon possibly have in common? Here are a few possibilities: Number one: They were both Jewish. Number two: They both got a lot of girls. Number three: They both wrote the lyrics to a number-one pop song. And, number four: they both dabbled in game theory. OL, what’s your answer? Trick question: All four are true. Number one: King Solomon of course was Jewish; so was David Lee Roth. In fact, he says he learned to sing while preparing for his bar mitzvah.
[“Just a Gigolo” by David Lee Roth]
Number two: Girls, girls, girls. David Lee Roth says he “slept with every pretty girl with two legs in her pants” and “I even slept with an amputee.” King Solomon, according to the Bible, had “seven hundred wives, princesses, and three hundred concubines.”
[“Jump” by Van Halen]
Number three: the lyrics to a number-one pop song? Yes. David Lee Roth wrote the lyrics for most Van Halen songs, including their only number-one hit, “Jump!” King Solomon, meanwhile, is thought to have written several biblical books, including Ecclesiastes — which the folk singer Pete Seeger used as lyrics to his song “Turn! Turn! Turn!”—which, when recorded by the Byrds in 1965, was a number-one hit.
[BYRDS, Turn! Turn! Turn!]
“To everything, turn, turn, turn/there is a season, turn, turn, turn/and a time to every purpose under the heaven.”
Which brings us to number four, the game theory. The most famous story about King Solomon in the Bible involves two women who come to him with a baby – and a dilemma.
David SPERLING: These two women come to the king and the first one speaks up and said, “We women we live together, just the two of us, nobody else in the house, no men, just the two of us women.”
That’s David Sperling, a professor of Bible at Hebrew Union College.
SPERLING: “And we both gave birth more or less around the same time.”
OK, so two women, two newborn babies, in one house. One woman was sleeping next to her baby…
SPERLING: … and she crushed him to death, she rolled over on him and suffocated the kid.
Joseph TELUSHKIN: And in the morning the woman wakes up with a dead baby on her chest.
That’s Joseph Telushkin, author of many books including Jewish Literacy.
TELUSHKIN: But she claims that that isn’t her baby. She claims that the baby that the other woman is holding is the baby. So one woman is saying that it’s her baby and the other woman is saying not he live baby is mine, this woman accidently killed her own baby and then took mine.
It’s a case of she said, she said.
TELUSHKIN: How was Solomon to decide the case?
How was Solomon supposed to decide this case? If only he could create a separating equilibrium.
TELUSHKIN: So finally in a moment of desperation and to the shock of everyone in the court…
SPERLING: He calls for… he says to his servant, “Bring me a sword,” And he says, “Now…”
TELUSHKIN: … “we’ll cut the baby in half, each mother will get half.”
You know what happens next, don’t you?
TELUSHKIN: One mother cries out, “Don’t do that! Don’t do that, you’ll kill the baby!”
SPERLING: The second woman says, “No, go cut the baby up.”
King Solomon doesn’t actually slice the baby in half. He doesn’t have to: because he now knows who the real mother is. How? He figured that the second woman, the one who was cruel enough to go along with his baby-carving plan, was also cruel enough to steal another woman’s child. And, further, he knew that the child’s real mother would rather give up the baby than see it die. King Solomon had set a trap that encouraged the guilty and the innocent to sort themselves out.
TELUSHKIN: I grew up in an era of like Perry Mason shows where often if you’re a smart prosecutor or a smart defense lawyer you’re going to ask your questions in a very provocative way to get the person to finally say something that they, in the normal order of events wouldn’t have been willing to say. And Solomon, by coming up with such a surprising question, was able to do it. It doesn’t make sense that a king would say, “Bring a sword and cut a baby in half.” So it introduces such a surprising and peculiar element that people get shocked. And in their shock they reveal, you know, the expression their true colors. They reveal who they really are.
So King Solomon was pretty clever. Is it possible that David Lee Roth was even cleverer? By the early 1980s, Van Halen had become one of the biggest rock bands in history, they had these extravagant live shows: a huge stage set, booming audio, and spectacular lighting effects. Their contract carried a 53-page rider that laid out the technical requirements for all this, as well as other demands:
Mike PEDEN: You know, it lists all kinds of food that they want…
Mike Peden worked for a concert promoter in Syracuse, New York.
PEDEN: Let’s see, some of the strange ones… you know, various doughnuts, a dozen hard-boiled eggs. They want Froot Loops and Raisin Bran. They want real knives and forks, they don’t want plastic ones.
Patrick Whitley was Van Halen’s production manager at the time. Which means he was responsible for that contract rider.
Patrick WHITLEY: And I think I actually typed that sheet…
DUBNER: It looks like an IBM Selectric, maybe? Is that what you owned?
WHITLEY: Yeah. It probably was. Yeah.
DUBNER: OK, and then I really love that you made sure you got your vegetables, or you made sure that somebody got their vegetables. Even days there were brussel sprouts, broccoli, cauliflower, mushrooms, spinach. And then odd days, peas, green beans, corn, carrots and tomatoes. And on it goes. You needed some whiskey, beer and wine, I gather, yes?
WHITLEY: Of course.
DUBNER: Some KY Jelly. What’s that for?
WHITLEY: For fun.
There was also a section for “munchies.” Here’s Mike Peden again:
PEDEN: …potato chips with assorted dips, nuts, pretzels….
And, in the middle of the munchies section:
PEDEN: M&M’s, and then in capital letters and underlined and in parentheses it says: “WARNING: ABSOLUTELY NO BROWN ONES.” And then it just goes on to say 12 Reese’s Peanut Butter Cups, and 12 assorted Dannon yogurt on ice. So, if you’re reading through there you could easily, if you’re not paying attention, skip that.
Mike Peden’s sister, Donna, worked as a caterer on the 1982 Van Halen concert in Syracuse.
PEDEN: So my sister went out and bought, I think, three or four bags of M&M’s, and she sat there and with rubber gloves on removed all of the brown ones. And not happy about it either. She actually hates M&M’s and will not eat any to this day because of that.
So why did Mike Peden think that Van Halen demanded that all the brown M&M’s be removed?
PEDEN: Well, we thought that it was just extravagance, that it was just David Lee Roth and Van Halen being David Lee Roth and Van Halen.
Coming up on Freakonomics Radio: was it really just extravagance? David Lee Roth gives his version of the M&M story:
ROTH: This was touted wildly and widely as simple rock start misdemeanor excess, and being abusive of others simply because we could. And who am to get in the way of a good rumor?
And you’ll hear Steve Levitt describe our attempt to teach the garden to weed itself – of terrorists.
Steve LEVITT: Yeah, that was one of my favorite things of all time.
But we didn’t tell the whole story, did we?
LEVITT: No we didn’t, we lied, and that was what was so fun about it
One more thing: If you are not already a subscriber to Freakonomics Radio — you should be. Just sign up, for free, at iTunes, and you’ll get the next episode in your sleep.
* * *
So why did Van Halen require that the brown M&M’s – dark brown and light brown – be removed from the M&M bowl in their backstage munchies? Was it nothing more than rock-star excess – or, perhaps, a clever way of finding out some information that would otherwise be hard to find? Remember, at the time, Van Halen put on one of the most ambitious live rock shows of anyone.
Steve LEMON: There were two names that you as a house guy didn’t want to hear were really coming.
Steve Lemon was a rigger on the Van Halen production crew, suspending the lights and sound equipment above the stage.
LEMON: One was of course KISS and the other one was Van Halen. Because these guys always brought the biggest shows. They were going to challenge you and your local team.
WHITLEY: We would have a thousand lights, which was this magic, sort of amazing number that nobody had ever had. We had more trucks.
Patrick Whitley again, the production manager:
WHITLEY: Towards the ’84 period, we’d be touring with a rolling stage and a grid to suspend the lights from because the lighting system would move. So we sort of considered ourselves the innovators and the inventors of a lot of the standard practices of how people tour nowadays.
The upside of this innovation was obvious. But there was a downside too: It could be dangerous. You didn’t want a light tower falling on you, or the stage collapsing under the weight of all that gear. To that end, the band had to trust that the local promoter in each city took seriously the contract rider that listed all the technical requirements for this massive stage show.
ROTH: The promoters frequently didn’t read the contract rider and we would have structural, physical issues because, hey, there wasn’t the proper electricity, load bearing, stress, etc.
So David Lee Roth, as he explains in a 2012 video, claims that he came up with a trick – kind of like King Solomon’s trick – to figure out if a promoter had read the rider carefully.
ROTH: If I came backstage, having been one of the architects of this lighting and staging design, and I saw brown M&M’s on the catering table, then guaranteed the promoter had not read the contract rider and we had to do a serious line check because frequently we had danger issues.
So the brown M&M clause, according to David Lee Roth, at least, wasn’t just a prima donna move. It was a clever way to teach the garden to weed itself – to let a bad concert promoter reveal himself as bad when, of course, he’d never come forward and admit to being bad. Inspired by stories like these – and the bright minds of David Lee Roth and King Solomon and even the medieval priests who may have rigged the ordeals – Steve Levitt and I thought maybe we could entice some bad guys to reveal themselves.
DUBNER: Hey Levitt?
DUBNER: You remember that story we wrote in SuperFreakonomics about why terrorists should buy life insurance?
LEVITT: Yeah, that was one of my favorite things of all time.
DUBNER: But we didn’t tell the whole story, did we?
LEVITT: No we didn’t, we lied, and that was what was so fun about it
Lying, we should say, in the service of a greater good. Steve Levitt, my Freakonomics friend and co-author, had identified many bad guys in the past – cheating schoolteachers, collusive sumo wrestlers – by finding patterns in the data.
LEVITT: The bad guys I’ve caught in the past, it was so easy. It was like shooting ducks in a barrel to catch sumo wrestlers. I mean, the data are right there, you can understand the incentives. It’s a really, really simple problem. And although you know that I don’t actually think that terrorism is a very big problem, and I feel that way too much time, and effort and manpower, and economic distortions happen because of terrorism, just from a purely intellectual perspective. For me, the idea of catching terrorists was really fun, because it was a such an incredibly hard problem.
DUBNER: Now, you had tried, you had met with people like the CIA, for instance in this country before, and that didn’t work out so well did it?
LEVITT: Yeah, no I didn’t really make much headway. I was invited and it was nice, the CIA had me out for the day, and we had a fun time. But I couldn’t convince folks there or at any of the American banks to work with me on my pet idea about catching terrorists using retail banking data until we stumbled on to a British bank that amazingly was willing to give it a run.
DUBNER: So in SuperFreak we describe this algorithm that was loaded into the computers of a big bank and which was able to sift through billions of data points and identify a relatively teeny handful of potentially deadly terrorists.
LEVITT: That’s true.
DUBNER: And there was one variable that we wrote about in SuperFreakonomics that we, I would say, highlighted, above all other variables. And that was whether or not a given bank customer had bought life insurance from that bank. Can you explain that variable and why we presented it in the book as we did?
LEVITT: Yeah, so we even, as you say, not just highlighted, we put it in the subtitle to the book was “Why Suicide Bombers Should Buy Life Insurance.” And here was the logic. In general, you wouldn’t think buying life insurance would be a great idea for a terrorist because, number one, they tend not to have strong attachments to other people, and number two even if you did kill yourself, you know, detonating a bomb in the Tube in London, the chances that the life insurance company would actually pay you off in that setting is probably low. So it seems silly to have life insurance if you’re a terrorist. Which is exactly the reason we argue that you should have life insurance, because you have data snoops like us going through the data, and if we see a young man who might potentially be a terrorist but he has life insurance, well it’s not likely that he’s a terrorist because it doesn’t make sense for a terrorist to have life insurance. So it’s a reverse logic of throwing us off the trail. And that is the sort of logic we threw out in the book. Now I have to say it was a little bit uncomfortable at times, because as many astute readers said to us, they said, “Well wait,” that doesn’t make all that much sense to me because most insurance, life insurance policies, if you do commit suicide, as long as it’s been year or two since you bought your policy it’s still covered by the policy and you will get paid. So I don’t really understand what you guys are talking about. So it was a little bit embarrassing that we had to write back and say well you’re right, that is the way these policies work, but maybe it will make sense to you down the road.
DUBNER: And that wasn’t the only kind of challenging part of that because when we went on book tour in the U.K. I guess in 2009 or 2010, you had a lot of people saying to us in interviews on TV and in newspapers and so on, what fresh kind of idiots are you that you would go to the trouble to work for years on an algorithm that would find terrorists and then in the book tell these same people exactly how to evade it, which is to say they should go down to the bank and buy some life insurance?
LEVITT: I know, what kind of horrible, base traitors were we, Dubner, that we worked so hard. And we put together this algorithm and then to just give it away because we wanted to make money selling a book? It was really incredibly damning criticism. I mean how could a person possibly respond to that? We would just kind of hang our head and say, you know, well, you know, and hem and haw. And what could you say? It was true. We were horrible, horrible people.
DUBNER: But there was a bigger truth.
LEVITT: There was. This was actually…I’ve hatched many plans in my life but very few as ambitious and exciting as this one. And it must have been, I don’t know now, seven or eight years ago that I first had the idea that if you wanted to catch terrorists it would be very difficult to do it with data, you really needed the terrorists to raise their hand in the air and announce that they themselves were the terrorists. Okay, but not easy to do that. Why would a terrorist come forward and say they were terrorists? So what it required was a trick. Okay? And this whole thing in the SuperFreakonomics book about life insurance was just a complete and total lie. It was made up from beginning to end. Nobody buys life insurance from their bank. I mean, there was some products that people could buy, but nobody purchased it. I mean I’m guessing that a handful of all of our listeners on this podcast have ever bought life insurance from the bank. So what in the world, why would we make this up, what were we talking about. Well here’s the idea: If nobody buys life insurance from the bank, but we managed to get the tabloids in the U.K. and the TV stations to say look these guys are looking for terrorists, and they say if you’re a terrorist and you buy life insurance then you’ll be off their radar screen. I mean, if I’m not a terrorist, I don’t pay any attention to it. If I’m a terrorist I think twice and I say, hmmm, maybe if I buy life insurance that will get me off their radar screen. And if you are a really, really dumb terrorist, hopefully what you do is you go to the bank and you purchase life insurance. Because of course it’s a trick. And we’re watching to see who, after SuperFreakonomics, comes out and the tabloids write about it shows up at the bank and buys life insurance. And our guess is that the kind of person that buys life insurance when they think it gets you off the hook as a terrorist is much more likely to be a terrorist than a regular person. In other words, by pulling off this scam we manage to get some set of terrorists, the really, really dumb ones, to go to the bank and essentially announce, ‘I am a terrorist.’
DUBNER: Levitt, let me ask you this, so this example of teaching your garden to weed itself with planting this action in the minds of guilty people that only guilty people would respond to, is easily the most outlandish and intricate of all the examples that we’ve given compared to David Lee Roth, and King Solomon, and even compared to the medieval ordeals. This is easily the most intricate. And it seems way beyond anything that the average person would ever need to really think about. But I’m curious, could you distill the lessons that you learned from this trick to give listeners a way to think about a way in which they might someday need, or be able to, teach some kind of garden to weed itself?
LEVITT: Yeah, the teaching the garden to weed itself is the ultimate expression of using incentives. And the basics it comes down to is thinking about a way in which you get people who don’t want to tell you something to tell you something by accident. And there are examples, certainly what we’re doing is to completely original. Police will often keep secret many details of the crime scene in the hope that they can get the potential suspects to start to mention details that they couldn’t possibly know if they hadn’t been there. That’s a great example again of the same idea of how you get people to reveal themselves. I wouldn’t say it’s easy. I wouldn’t say these situations come up often, but it’s a tool. It’s a tool in the toolkit, which says if it’s too hard to figure out who the bad guys are just by snooping around, I need the bad guys to tell me who they are. And again, in every setting it will be different how you do it. But it’s simply the knowledge that that’s your last gasp. When you can’t get people to do it otherwise, you actually have to get them to tell you. It at least is the first step on the path of figuring out a way to get the bad guys to come forward and tell you who they are.
So it’s a good way to “get the bad guys to come forward and tell you who they are.” Fair enough. But let’s not be naïve here. Bad guys can teach their garden to weed itself, too. You know the Nigerian email scam? Of course you do. It’s famous: you get an email from some deposed Nigerian government minister who has millions of dollars locked up in some bank and needs help getting it out. Help from you. For which you will, of course, be handsomely rewarded. So here’s a question: if the Nigerian email scam is so famous, why would a Nigerian scammer say he’s from Nigeria? That’s what a Microsoft researcher named Cormac Herley wanted to know. So he investigated. His conclusion? These scammers are actually quite clever. When they send out all those bait emails, what are they actually searching for? They are looking for someone so gullible that that person will end up sending thousands of dollars to a faraway stranger based on some kooky email about a fictional fortune. But how are the scammers supposed to sift the truly gullible from everyone else? By sending out such a kooky email that only a gullible person would take it seriously. An email that anyone with an ounce of sense or experience would immediately trash. So think about that the next time you get one of those Nigerian scam emails. Your first instinct may have been to think how stupid the scammer is. But now you’ll know better. Now you’ll know that this is exactly the kind of stupid we should all aspire to be.