Data Radicals logo

Premature Enumeration

Tim Harford, Host of Cautionary Tales and Author of The Undercover Economist

Tim Harford

Tim Harford is a senior columnist at the Financial Times, the presenter of BBC Radio 4’s More or Less, the series 50 Things That Made the Modern Economy, and the podcast Cautionary Tales. His books include the bestselling The Undercover Economist. At the 2019 New Year Honours, Tim was made an OBE for services to improving economic understanding.

Tim Harford

Tim Harford

Host of Cautionary Tales and Author of The Undercover Economist

Satyen Sangani

As the Co-founder and CEO of Alation, Satyen lives his passion of empowering a curious and rational world by fundamentally improving the way data consumers, creators, and stewards find, understand, and trust data. Industry insiders call him a visionary entrepreneur. Those who meet him call him warm and down-to-earth. His kids call him “Dad.”

Satyen Sangani

Satyen Sangani

CEO & Co-Founder

Alation

Producer: (00:02) Hello and welcome to Data Radicals. In today's episode, Satyen sits down with fellow data radical Tim Harford. Tim is a self-proclaimed nerd storyteller and brings data-driven stories to the masses. In this episode, he shares how using data will put you ahead of the curve, how to foster a spirit of curiosity around data, and shares real-world examples of change in action.

Producer: (00:27) This podcast is brought to you by Experian and our data quality solutions. At Experian, our top priority is helping thousands of business users and their data stewards unlock the power of their data. Our solutions allow you to analyze, enrich, profile, validate, and standardize your organization's data to transform it into a reliable asset. Our clients tell us that trustworthy data has supercharged their data governance programs, improving ROI, collaboration, performance, and security. Get started at edq.com/free-trial.

Satyen Sangani: (01:08) Today on Data Radicals, we're joined by Tim Harford. Tim is an economist, journalist, and broadcaster. He's the author of How to Make the World Add Up, or The Data Detective, Messy, and the million-book-selling The Undercover Economist. Tim is a senior columnist at Financial Times and the presenter of the BBC Radio's More or Less, How to Vaccinate the World, and 50 Things that Made the Modern Economy, as well as the podcast — and my personal favorite — Cautionary Tales. Tim has spoken at TED, PopTech and the Sydney Opera House. He's a member of the Nuffield College at Oxford and an honorary fellow of the Royal Statistical Society, and he was awarded an Order of the British Empire for services to improve economic understanding in the New Year's Honours of 2019. Tim, welcome to Data Radicals.

Tim Harford: (01:56) It's my pleasure to be on the show.

Satyen Sangani: (01:58) It's great to see you again and I'd love to spend as much about this podcast talking about you and how you do your work as we do the work itself, because I think you do the style of work that many of our audience members aspire to do, which is to unearth really interesting insights and then to present them out in a way that captivates people's imaginations.

So let's start with you. You're a writer, a journalist, a broadcaster, an economist, a podcaster. How do you describe yourself?


Tim Harford's Professional Background

Tim Harford: (02:26) I generally would go for something like “nerd storyteller.” I don't know if that's the best description. I trained as an economist rather than as a data scientist or a statistician or, indeed, as a journalist but I'm a communicator. I write but I write in different media. So I write books — I've just finished my tenth book — I write newspaper columns, and I write radio scripts and then, having written the radio scripts, I guess I also have to read the radio scripts into the microphone. So, I'm a radio guy and I'm a print guy but I suppose “nerd storyteller” covers most of it because I am pretty nerdy and I am fascinated by stories.

Satyen Sangani: (03:02) And do you write all of your own scripts for everything that you produce and are a part of? Are you able to generate all of that content yourself?

Tim Harford: (03:11) No, not everything. So I write all my columns, I write my books. Cautionary Tales is written with a colleague, a very old friend of mine called Andrew Wright, who people will hear him credited at the end of each Cautionary Tale. We pretty much split it up so he will just write some of the scripts and I will write other scripts and then we'll edit each other’s scripts. So that's the division of labor there, which works pretty well. And then with More or Less, the BBC Radio program, that's more complex because I will get chunks of script for items on the show from reporters, from producers, and they will be in very different states of repair. So some of them are very rough, depending on the experience of the reporter, and some of them are not very good, some of them are brilliant, some of them are absolutely perfect, and you just work on them and try to get them into a consistent form with a More or Less voice.

Tim Harford: (04:06) But that sometimes requires absolutely no effort from me at all and sometimes I occasionally find myself thinking, “I wish I'd written this right from the start because it would have been easier,” but that's what working in a team is like and part of the job there is (and part of the joy of it is) to be able to work with people who produce work that you just think, "I could never have done this myself. This is brilliant." And also to be mentoring more junior members of staff and advising them and showing them the edits that you're making and talking to them about why you're making it, which is not easy. I always used to originally just work by myself on my own scripts but now I try to get better at that.

Satyen Sangani: (04:45) Yeah. Writing is obviously a craft and you certainly have the ability to practice that craft. As I looked through your work, one of the things that I really appreciated was how prolific you are and how many different stories you get. Can you tell us a little bit about the story acquisition process? Because I would imagine some of it you get directly, some of it's brought to you. How do you know when something's a good story and how do you know when you've got something great?


How do you pick a topic as a storyteller?

Tim Harford: (05:12) Yeah, it does depend a lot on the medium. So for a newspaper column, it's often a question I have in my mind or I'm having a conversation with somebody. We'll be talking about something and I'll think that's an interesting distinction or that's the question that's on people's minds, and so it will come like that. For More or Less on BBC Radio 4, almost all of it is generated by our listeners. They will ask us questions and we'll try and answer them, and that's very powerful because it starts with the question. Very often I think that data journalism starts either with "How can we plot a cute graph?" Or there's a fact-checking element: "How can we debunk this stupid thing that somebody said?" And both of those are important.

Tim Harford: (06:04) But I think to start with the question is the most powerful thing of all, and then for Cautionary TalesCautionary Tales for those who don't know, is a podcast about things going wrong, it is stories of things going wrong and the lessons that we can learn from those disasters. So that's very simple. I just have my eye out for disasters, these are historical incidents, so they are often very well documented, they might be 10 years old or 100 years old, some of them even older than that, and what I'm looking for there is not so much the story, because the story itself is often fairly obvious. It's the twist or the surprise, or maybe the connection between one story and another that gives it some extra richness and depth.

Satyen Sangani: (06:44) Do you have any particular work that you are particularly proud of or are really proud that you've done over your career, is there one work that if there was nothing else somebody could take from the Tim Harford compendium they should look up and read?

Tim Harford: (07:00) Well, actually you asked two different questions now, one of which is what should people pick out, and the other is what am I most proud of...

Satyen Sangani: (07:06) I did, I did.

Tim Harford: (07:07) They are not necessarily the same question! I would suggest maybe people, if they wanted to pick something up, start with Cautionary Tales, grab an episode of Cautionary Tales and have a listen; the last half an hour. I think there are good stories, there's a lot to offer, and I'm really enjoying them.


How the pandemic changed public perception of data

But that's not the work I'm proudest of, the work I think I'm proudest of is the work I did with the team from More or Less, which was very much a team effort during the pandemic, and in particular during the first lockdowns in the UK, which were basically April/May 2020. And we all have our own memories of what was happening there in different places, but all across the world, everybody was affected. And we had just been moved to a more prominent slot in the radio schedule on Radio 4. And radio listening is very — your audience is very influenced by what time you broadcast, it makes all the difference. And so it had some meetings and we were talking about what we were gonna do with this new slot, it was all very exciting. And it was gonna start, I think, in late April.

Tim Harford: (08:13) And then we just got a call, "Can you get on air next week? Can you go immediately? This is a number story, everybody wants to know what's going on, we are all confused by the data." And so there was this scramble in a much more prominent spot than we'd ever been before by a team who were trying to work mostly under duvets from home. I was working under a duvet from home, although my producer and the studio engineer were going into the office. They were regarded as essential workers, we were like physicians and nurses and firefighters and people working in food. We were regarded as doing essential journalism, which sort of simultaneously felt a bit absurd, and then you thought, "Well, no, maybe this is important."

And we were just trying to make sense of what was happening to everybody, and give people an anchor, and maybe some kind of compass as to where this was all going. And it sounds strange because it was a very traumatic time for many people, a very frightening time, but there was something refreshing about the fact that people actually wanted to understand the world using numbers. Coming from some very vitriolic and divisive political campaigns where numbers only ever seem to be propaganda and no one would actually pay any attention to numbers, they just thought, “Well, it's just more people trying to sell me ideas that I don't believe in,” and that changed briefly.

Tim Harford: (09:45) People wanted to understand this new thing and they realized that data were the way that they were going to do that. And I still run into people who tell me that just listening to that series of radio programs — we broadcast every week for about three months, which was much longer than we would normally do, and just listening to that really helped to send to them and keep them going in the most extraordinarily confusing time.

Satyen Sangani: (10:09) Because in those confusing times people are looking for truth and some mooring. And this gave them facts over ideology or some mythical explanation as to what was happening or why.

Tim Harford: (10:30) Yeah, it's easy to forget now, but at the time, we didn't know the most basic things about this virus. In particular, we didn't know how dangerous it was, we didn't know how fast it spread, we didn't know who was most at risk. There was a theory out there that turned out to be wrong, but was not absurd, that maybe lots of people had already been exposed, but hadn't had symptoms, it was just like a cold and therefore maybe it would all soon be over, and I think a lot of people embraced that theory because it felt like very good news.

And just trying to answer these really basic questions. Like how many cases have there been, how many cases result in death? This long COVID thing that people are starting to talk about, how do you make sense of that? How do you make sense of that after three months? Because there isn't time to really get a feeling of what long COVID might even be. We didn't know the answers to any of those questions, we were just trying to figure that out in real time and doing our best. We have limited data, it felt like important work. And, of course, we were not ones doing the work, but we were the ones trying to help people access that work and understand what was going on.

Satyen Sangani: (11:38) Yeah, this is such a great story because it's... One of the things that makes me a believer, so we have this podcast, the Data Radicals, which is all about spreading data culture and how people are able to sort of proselytize, if you will, the religion of data inside of their companies. And if that's what they're trying to do, one of the things that struck me is that when people tell these stories about how they were able to drive change, often they're doing so in moments of crisis. It's not the day-to-day report that reports the news that gets people to change and pay attention to the data or change their habits. It is often these sort of pivotal moments where data is introduced at a time where people either didn't know that they wanted it or didn't know they could get it. Is that something that you've seen in your reporting as well?

Tim Harford: (12:24) No, I think that's right. But I think that's a real shame because, of course, if the moment of crisis is the moment where you suddenly say we need some data, then it's too late, really. In the case of the pandemic, most of the data that we would have wanted, there was no way we could have collected it anyway, but there were a lot of elements to that where we just had very, very poor numbers, very poor statistics. And that's something that became a serious weakness.

For example, Alexis Madrigal, who was spearing a volunteer data effort in the US, told me that we didn't actually know how many hospitals there were in the US. We didn't know how many hospitals there were in the US, so that's a really — that's a basic question. In the UK, we didn't know how many people were receiving nursing care and nursing homes were, of course, highly vulnerable to the virus, where a lot of the deaths occurred, and there was just no centralized database. If we don't know how many people are receiving the care, it's really hard to know how to reach them, how to protect them. So these are quite basic things. On the other hand, there are examples of really heroic efforts to gather the data in an awful hurry.

Tim Harford: (13:45) So one of the ones that I describe in The Data Detective — I think in the paperback edition or the softcover edition of The Data Detective because of the timing — was the recovery trial. So the recovery trial was set up in the UK and the idea came on a bus journey across London, between a conversation between a couple of scientists on a bus journey in February 2020, saying, look at what's happening in Italy, you can see the TV footage coming out of Italy, we've seen the footage coming out of Wuhan, something terrible is coming, and what will happen is intensive care all over the world will be swamped and physicians all over the world will do their best, they'll try stuff that they hope might work. And then some of the patients will die and some of the patients will pull through and nobody will really have gathered enough data to know whether any of this stuff is working.


Embracing pandemic data: the recovery trial

Tim Harford: (14:42) And so the idea of the recovery trial was to be very systematic about collecting the data as patients came into UK hospitals, it's all part of the National Health Service. So if you can talk to the right person, and you pull the right lever, you can make it happen. So anybody coming into a UK hospital with COVID, the consent would be sought either from the patient or their family to sign them up for a randomized trial, a very simple randomized trial, and they would receive a promising treatment chosen at random.

And because of the number of people coming into hospital and because of how quickly the disease progresses, it was just a matter of weeks before we had answers. So we found out, for example, hydroxychloroquine: a lot of talk about it, some promising initial results turns out it doesn't work. Dexamethasone, a steroid, works extremely well and it's super cheap, you can afford it almost anywhere in the world. You can afford dexamethasone; it is easy to make. These are findings that saved probably a million lives in the first year of the pandemic, probably three, four, five million since then. And it was just a case of thinking, how do we get systematic about collecting data in an awful hurry? They did that. They got the numbers. And I just think that's just a fantastic example to us all. If you're even a few weeks ahead of the curve, you can really make a difference to the quality of decision-making.


Why did you write The Data Detective?

Satyen Sangani: (16:10) Yeah, and I think for that reason, I would — look, I think you don't necessarily, as an analyst or as a practitioner with data, have the ability to control whether there's an environmental crisis or not, but you certainly have the ability to choose how you react. And I think in a lot of these moments, these can be opportunities for people who are enterprising or want to drive some form of change and use the sort of tactics of a crisis to invest in the infrastructure that is harder to gather. You mentioned The Data Detective, and tell us about what motivated you to write the book and how you came up with the idea.

Tim Harford: (16:50) Sure, and you correctly gave it two titles at the beginning of the podcast, so I should just clarify, so in the US and Canada, and also Guam, as it happens, it is called The Data Detective. Elsewhere, so the UK, Ireland, Australia, India, South Africa, it's called How to Make the World Add Up. The Data Detective, How to Make the World Add Up, same book. So please don't buy both copies — or if you do buy both copies, thanks very much, but please don't get cross with me that you have paid for the same book twice. That's the book.

It's a book about how to think about the world using numbers. A friend of mine said, "Oh, you've written this book about how to think about numbers," and I said, "No, it's a book about how to think. Numbers are just tools that help you think." And there are really two things that I wanted to do. One was to build a positive case for numbers. I know this sounds strange, but actually when you think of the number of statistics books that are published, popular statistics books that are basically all about statistical mistakes, and so I thought that's telling, but there's a perception people have of my radio show More or Less “Oh! That show where you debunk all those false statements by politicians.”

Tim Harford: (18:09) Yeah, we do that, but also we are trying to explain the world. So that was one thing, a constructive view that numbers can be used to understand the world and to help make better decisions and to help communicate ideas. They are not just a vector for misinformation. They can be a vector for misinformation, but that is not the only thing that they are. The second thing that I wanted to do was to be psychologically realistic about this. I can give you all the technical advice in the world about correlation and causation and p-values and Bayesian updating and this and that. It won't do you any good if your head's not right, if your attitude is not right. We are incredibly good at fooling ourselves for all kinds of reasons. And that's why one of the first stories in the book is about an art forgery that fooled the world's greatest art critic. And it's not really a story about statistics at all, it's a story about how expertise is no defense against self-deception. So you need to get your head clear, you need to understand those biases that you have, those filters that are getting in the way of clear thinking. And after you've done that, then I can maybe give you some useful advice as to how to get something really constructive out of these numbers that surround us.

Satyen Sangani: (19:28) So the motivation was to spread the idea of numeracy. And I guess, what did you learn in that journey? What were the lessons that you took away? One you just mentioned, which was, of course, that you had to remove your ego or prior biases from the initial lens of looking at a problem. But what else?

Tim Harford: (19:49) One of the things that really struck me was how much of the advice is actually not about the numbers, it's about the stuff that surrounds numbers. So things like useful comparisons. So how does this number compare to some other number that we understand? If I'm saying that the deficit is whatever, trillion dollars or whatever. That doesn't help to say that the deficit's a trillion dollars. What I need to know is, things like, "What's the deficit relative to what it was last year or five years ago? What's the deficit relative to, say, the total size of the economy? And what about other deficits in other countries relative to their economy?"

Or maybe, "What's the deficit relative to all government spending?" Or maybe, "What's the deficit per person?" You can just tell it to me in dollars per person in the country. So these are all different ways to provide some context for this number that doesn't really make any sense. So, useful comparisons, that's one bit of context. Simple things like, "Is the number going up or down?" That's important. What's the source of the number? That's important. But then one piece of advice that I think I would really emphasize to people who are already quite comfortable with numbers, because this is where they trip up, is what is the actual definition?

Tim Harford: (21:05) What's the label attached to this number? What was the process that generated it? You have this phrase in the book: premature enumeration. And I argue that a lot of people who are comfortable with statistics suffer from premature enumeration. Which is when you grab hold of the numbers because you see the numbers and you're like, "Great, we're gonna stick this into a statistical package or we'll plot a graph or we'll just compute some simple ratios." Or whatever it is. "We'll plot a trendline, calculate a key value." And you do all this without stopping to say, "What was actually being measured?" Or, "What was actually being counted?" And you can really lead yourself astray by doing that.

And to give you a specific example, if you think back 15 years to the financial crisis. That's got many causes, but one of the causes was people betting very large sums of money on the output of calculations about risk. So saying, "Oh, this product has this particular level of risk." Or, "This product has this particular level of risk because there's some cross-correlation with this other thing, or we've sliced and diced it." Or, "90% of this product has to go bad before the 10% that we've bought will go bad."

Tim Harford: (22:22) And all of these things spit out these calculations that say, "This is incredibly safe, there's no way that this will ever go bad." But all of those calculations were based on clever mathematics applied to a measure of risk. And actually, you can't measure risk. You can measure other things, you can mention for example, historical variability, that's fine. But that's not risk, it's not possible. It's just not possible to measure future risk. All you can do is measure some proxy. And so what was happening was, huge bets on the basis of what looked like very solid calculations, but in the end, the whole thing was just built on quicksand. Because those initial numbers that we're purporting to measure risk, never measured risk because you can't measure risk, they measured something else. And everything that came afterwards was basically logically fallacious, you couldn't rely on it. And there you go, magic up one financial crisis. And it's all because people didn't ask, "Where did those initial numbers actually come from and what did they actually mean?"

Satyen Sangani: (23:34) So this is a little hard and maybe discouraging, because you're telling me that, first of all, I've gotta get my ego out of the way. Which means for people who are hearing that, that they have to take their experience, which is their learned lifetime accumulated knowledge about the world, and they have to sort of sublimate that. And then, how do I approach the...

Tim Harford: (23:56) I would say they have to... Sorry, but they have to use that. That... Your experience, should tell you that actually there are lots of ways in which you can be wrong. But if you're...

Satyen Sangani: (24:04) It ought to.

Tim Harford: (24:05) So it should do, and it can do, but you've got... It's partly just the case of stopping and thinking. It really is partly just the case of pausing, because we... When we are very expert, you spot patterns very quickly and maybe just a bit too quickly and count to three and go back and have another look, and you might reach a different conclusion. And at that point, your expertise is helping. Sorry, I interrupted.


How do you convince the data-resistant?

Satyen Sangani: (24:28) No, I was gonna say, so there is certainly the bias and ego question. And then there is this reality that you're pointing out, which is, science is hard. Christie Aschwanden who was a famous science journalist, and she talks a lot about how... It's just, getting positive results is hard. Because the data isn't always conclusive and you can't always prove your hypothesis, and that might be because you don't have the right data, or it might not be. Maybe because the data doesn't actually support what you're thinking. So how do you convince people... So, as you've gone through The Data Detective then. How do you then take those problems and those challenges, and then convince people to still believe?

Tim Harford: (25:03) I think the final message of the book is to approach the world and approach the numbers with a spirit of curiosity. Which might sound fairly straight forward. "Oh yeah, curiosity. We all like curiosity." But very often we don't actually... We don't behave like we like curiosity. We often like simple answers, or we often like to just win arguments, we like things to go our way. And so we very often grab hold of the numbers thinking, "Oh yeah, this is gonna help me prove my point. This is good supporting evidence." As like legal evidence, rather than scientific evidence. 'This will help me convince somebody." Rather than to say, "The world is a confusing place. The world is a fascinating place. I have questions, and maybe the data can help me answer those questions." If that's the spirit with which you approach the data, then obstacles become intriguing mysteries, they become satisfying puzzles. Arguments turn into constructive exploratory questions. And what might have seemed like hard work, I think just... It's just a lot more fun.

As I quote Orson Welles, towards the end of The Data Detective, where he says, "As a film director, we worry about audiences not understanding things, but we shouldn't worry about that. That's not the problem. The problem is to get them interested. If they're interested, they'll understand anything."

Tim Harford: (26:39) And so, when we are looking at these complex questions, the problem is to interest ourselves. Can you actually summon the curiosity to go, "I don't actually understand this, I want to know the answer." If you have that spirit, then the rest of what you describe as hard work, yeah, it's hard work in a way, but it's satisfying work as well, and I think that keeps people going.

Satyen Sangani: (26:58) Do you have recommendations or practices around fostering that sense of curiosity? In the simplest setting, we have an executive staff that runs the business, and often people come into the room and there's some problem that's agendized on the table that we're all there to solve. And everybody's got views and issues and ideas and solutions. And often the best answers come when you have a spirit of curiosity, but it's hard to get there because of that ego, and the agenda, and the eyes that you come in with. How... Are there practices that you can recommend where you can foster that spirit, either amongst groups or individuals?

Tim Harford: (27:41) Yeah, I hope so. When you're communicating an idea, I think the way to foster a spirit of curiosity is to tell the story, then we wanna know what happens next. If you're trying to communicate, say, something about climate change, rather than starting with the science, whatever that means, or some very scary message. Maybe tell a story about a scientist making some discovery, or having some... That, I think can be a much more compelling way to get people interested in what would otherwise be this very... It's a very fraud-loaded subject. So that's one... And if your aim is to communicate that the story-telling mode is a good one, to awaken curiosity.

Another mode is to ask questions, and to ask genuine questions. Questions to which you think you don't know the answer or you might not know the... And if you're asking genuine questions of other people, you're treating them with respect. They may realize they don't know the answer to their own questions, which tends to defuse conflict. Or maybe they do know the answer to the question, in which case you're gonna get smarter. So that question-asking is another useful mode. And one final idea I would suggest is just ideas from design thinking. So rather than fixating on the first solution to say, "Okay, let's generate parallel solutions. Let's generate a second solution, or a third solution."

Tim Harford: (29:00) I'm not a design expert, there are lots of good books on this subject. But that's a third way to help generate the... This sort of parallel thinking and lots and lots of different ideas and getting people curious and exploring, rather than converging too early.

Satyen Sangani: (29:16) Yeah, these are all super interesting practices, and it reminds me a little bit of a... Especially in the group setting, it reminds me a little bit of a customer meeting I once had. We were going to a rail company in the Midwest, and they started every single meeting, even the ones in the office conference rooms with, "We are gonna do a safety check in this meeting. And we're gonna look around and see if there's any pinch points. And then we're gonna look around and see if there are fall hazards."

And I looked around, and this is quite unusual because we're just in an office building, not in a rail yard. And so at the time, it seemed a little ridiculous. But it clearly was a culture setting for them. And in this environment, I've often thought, "How can we do that without feeling... Making people feel like this is some hokey odd place?" But having that set of rules, or having that intention, or setting that intention seems like a really strong way. And remembering these rules in some way.

Tim Harford: (30:08) Yeah, I love that. And it reminds me of, years ago, I visited a nuclear power station because I was trying to figure out whether there was anything we could learn from safety in nuclear power stations, that could teach us about safety in the financial system. And there's this idea that these are... I know it sounds strange, but these are very complex, these are tightly coupled systems, there are some parallels that are potentially useful. And I had a sandwich lunch with... Or was gonna have a sandwich lunch with the head of safety and took off my trainers and put on these steel-capped boots to go and visit the turbine hall. And then the lady came in with a trolley with sandwiches and tea and coffee. And so she's like... She's the catering staff, and I'm talking to the head of safety. And the first thing she did, she stopped the trolley, and she pointed at my running shoes, and she said, "That's a tripping hazard." And she didn't talk to me, she talked... She was talking to the director of safety at the power plant and said, "That's a tripping hazard. They need to be moved immediately." And I thought, "Oh, okay, these guys are actually serious about safety." And from the point of view of curiosity, it's not the same thing. But what are the practices that you could have regularly that remind people? So those practices are: we always think about safety, we're always looking for hazards, we work as a team together to spot problems.

Tim Harford: (31:32) So what are the practices that with a group of analysts, you can say, remind people: we are always curious, we're always asking questions, we don't wanna converge on the wrong answer without exploring other possibilities. So what are the practices that we can have together that remind people that it's a safe space to raise those questions? Yeah, I love it, I love it.

Satyen Sangani: (31:52) Yeah, and I think it extends to... And I think this is what the brilliance of that setting was, which was that it extends not just to the people doing the analytical work, but to everybody, because it's the core of what they do. And in so many information enterprises, whether it's a bank or a technology company or the like, you have similar dynamics apply, where you're assembling knowledge workers to do something that's... There's no deterministic right answer.

Tim Harford: (32:21) Yeah.

Satyen Sangani: (32:21) I think this idea of... because I do think there's a lot of books, and yours is a great one in particular, that sort of convince people and teach people how to think. And it's not always, to your point, about p-values and conditional probabilities and distributions. But rather, about just being calm in the face of information. But then there's the organizational dynamics that really make it hard in a group setting to foster that religion.

Tim Harford: (32:51) And that's one of the things that I do find fascinating about the recovery trial, the fact that they managed to get a whole organization to buy in at least enough to make it happen. And they didn't have to get every doctor and every nurse in the National Health Service to constantly pay attention to the data. They managed to persuade them to install the system and use the system enough to get the data that we needed. And that saved millions of lives. It matters.

Satyen Sangani: (33:24) Yeah, and convincing people, you would think, "Oh, that's so obvious to do. We're in an emergency setting, so let's just convince these people to gather some data because it'll help us." But, while they're fighting this war, probably any change was quite hard to do.

Tim Harford: (33:34) Yeah. The people are always busy but never been busier. But they... Yeah, they managed to do it, and it really did make a difference.


The most interesting economist in the world

Satyen Sangani: (33:44) Yeah, for sure. So now, I'd love to go back to one of your earliest works because it's the one that sort of, I think, in some sense, launched your... I think it is, maybe it wasn't, but you can tell me about your history. Which is The Undercover Economist. And you've often referred to yourself as an undercover economist. And I think economics is certainly interesting, our first episode was with David Epstein. And he talked — or at least one of our first episodes — and he talked about how... He has this book, Range, and he talks about being a generalist and said very, very few disciplines prepare you for this “range.” In fact, all the specialized knowledge mostly gets forgotten in his telling except for economics, which I love to hear, because I had economics training as well. Tell us about that.

Tim Harford: (34:23) That's possibly slightly flattering to ask an economist, isn't it? But that's...

Satyen Sangani: (34:27) I love it.

Tim Harford: (34:28) It's how we like to think of ourselves, as this classic Keynes Essay, which is actually an obituary of Alfred Marshall, in Keynes’ wisdom. Alfred Marshall, great economist, Keynes, great economist. And Keynes is writing this obituary of Marshall and he describes the qualities of a good economist. And I'm afraid he refers to this economist as a “he” — which is a problem economists have, we need more female economists, and we've got some great ones these days. But Keynes says, "He must be mathematician, historian, statesman, philosopher in some degree. He must understand symbols and speak in words. He must contemplate the particular, in terms of the general, and touch abstract and concrete in the same flight of thought. No part of man's nature or his institutions must lie entirely outside his regard. He must be purposeful and disinterested in a simultaneous mood, as aloof and incorruptible as an artist. Yet sometimes as near the earth as a politician." So that's Keynes on the quality of a great economist. And he clearly has himself in mind. So yeah, I think economists of Keynes' generation had that range; I think economists today don't. We're much pure math guys.

Satyen Sangani: (35:46) It makes me think of that Dos Equis commercial of “the most interesting man in the world.”

Tim Harford: (35:51) Yeah, it does, it does. Yeah. Keynes, I think, clearly thought of himself as the most interesting man in the world. And he wasn't completely wrong, he was a very interesting man, but I suspect you, he had an even higher opinion of himself than everyone else had of him. So yes, there is something of that range in economics as it should be, anyway, not necessarily in economics as it is. But I also studied philosophy, I studied philosophy and economics at university, and before I went to university, we specialized very early in the UK. And I specialized in maths, physics, and English literature. So there's that constant fence-sitting of being interested in communication and poetry and writing, while also getting geeky and being interested in the differential equations and the numbers.

Satyen Sangani: (36:40) Yeah. But I think the... So besides the self-effacing view of an economist, there is something useful there I think, and I think it might... And it's a decision science. And in some sense, helping people to make better decisions is a lot of — or at least think better — is a lot of your work. Tell us about this idea of being an undercover economist. What's the ethos of that? Do you describe yourself that way? And I think it's a really helpful way for people to... Maybe that description of how you think about yourself is also a way in which many people might want to think about themselves. So I think it's great to hear it.

Tim Harford: (37:15) Yeah, so it was... You're right, it was my first book, it was the book that launched my career. And as the book was published, I had persuaded the FT to start a column also called “The Undercover Economist.” So I've been writing that column for 17 years, and the book was published 17 years ago, and they've gone hand in hand. But the original idea was that economists live in the real world, and so while you as the reader of my book, dear reader, might be looking around the café you're in, or wherever you are, whatever environment you're in, there might be a person next to you who's an economist, and she is seeing the world in a very different way. She is seeing things that you're not seeing. She doesn't look any different to you, but she thinks differently to you, and she sees different things because she's an economist. And that was the way I began that book: "Don the X-ray spectacles and see the hidden patterns in the world." So that was the idea. And yeah, economists do see the world differently. I think over the last 20 years, I've maybe slightly wound my neck in about the brilliance of economics and started to realize that some of the weird ways economists see the world are just weird, and are not necessarily helpful, but they are different, and there's a value in that difference. And some economic ideas I think are incredibly powerful.

Satyen Sangani: (38:35) Yeah, I think so too. And it's a really romantic way of defining a profession that I think is more and more useful in a world that's super complex. So super exciting and thoughtful and what a great conversation, what a great pleasure to have you on the show. I think our audience will really enjoy it, so thank you for taking the time.

Tim Harford: (38:58) It's my pleasure, it's very kind of you to have invited me, I'm glad. We've been waiting a while to do it, so I'm glad we were finally able to do it.

Satyen Sangani: (39:04) Yeah, me too.

But how do we foster curiosity and humility? We need to tell compelling stories, and ask thoughtful questions. We should encourage creativity and open-mindedness, and we must admit that we don't know everything. A mind is kind of like a parachute: It works best when it's open. So as we explore new avenues and study the topics that intrigue us, we'll start to see the world differently. As Tim said, you could become an undercover economist seeing the world completely differently than the person next to you.

Consider the world's greatest artists: Picasso, Dalí, Van Gogh, and Frida Kahlo. All of them interacted with the same world, and yet, they expressed it so differently in their art. The world would be a dull place without their unique visions, and the same goes for yours. So be curious, ask questions, delight in being mistaken, and then keep learning. Think like an undercover economist, and you'll grow as a data radical. Thank you for listening to this episode, and thank you, Tim, for joining. I'm your host, Satyen Sangani, CEO of Alation. And data radicals, stay the course, keep learning, and sharing. Until next time.

Producer: (40:37) This podcast is brought to you by Alation. Alation achieved 8 top rankings and 11 leading positions in 2 different peer groups in the latest edition of the Data Management Survey 23, conducted by BARC, the Business Application Research Center. Read the report at Alation.com/BARC23. That's A-L-A-T-I-O-N.com/B-A-R-C, then the number 23.