How to Make the World Add Up – by Tim Harford
Tim Harford is one of Britain’s best social scientific communicators. For almost two decades, his books have put into simple terms complex phenomena, and brought otherwise esoteric research in economics, psychology and beyond into the domain of the lay reader. On More or Less, Radio 4’s programme about the statistics around us in everyday life, he offers a calm and reasonable voice in the face of both accidental and wilful misuses of statistics.
His most recent book, “How to Make the World Add Up”, aims to correct what Harford views as a misconception in the world; that statistics, and their use in public discourse, cannot be trusted, as it is too easy to ‘lie with statistics’.
Harford disagrees. In fact, he says, statistics are the only tool we have for understanding some aspects of the world that we live in. Without statistics, how could we have learned that smoking causes cancer? Without statistics, how would we learn how to stop cancer’s spread through the body? How would we know what government was up to in our name, or how serious global wealth inequality really is? Instead of taking the popular view that there are “lies, damned lies, and statistics”, Harford proposes ten simple rules for understanding statistics, learning to interpret them, and working out whether a given statistic can be trusted.
To many, there is not much new here. Harford covers ground that most people nerdy enough to be reading this review will also have been nerdy enough to have already come across. From the reproducibility crisis in psychology, to the life of John Maynard Keynes, via Greek national statistics, you might have heard of many of these things before.
What makes Harford so different, however, is the skill with which he tells his story in colourful detail, and draws together seemingly disparate historical anecdotes to create a meticulous argument in favour of statistics and statisticians, allowing him to present the national statisticians of Canada as courageous heroes – not language often reserved for people in their position, but accurate nonetheless.
Part of this comes down to his ability to craft a story that is so calm and reasonable. Much of the debate around the reproducibility crisis in psychology suffers from the fact that its main participants (on both sides), come across as pugilistic, antagonistic, and more than a little smug. For people close to the conflict, I can see how this is a natural or even tempting way to behave. But, if we want to convince the wider world to be less convinced by claims of extra-sensory perception, or the power of power posing, then being seen to sneer and scoff probably doesn’t convince many people who didn’t already agree with us. Harford, by contrast, is convincing, because he does not appear to be emotionally attached to one hypothesis or another – following, in fact, the first of his own ten rules.
One of my slight disagreements with Harford in this book is about his assertion that “misinformation can be beautiful too” – a warning against being too drawn in by infographics and their seductive beauty. This chapter is structured around the lesser known tale of Florence Nightingale; Statistical Pioneer, which is certainly one of the stories I had not heard before. But I wonder if it might better be thought of as a call to action for people trying to ‘do statistics right’. The baddies will make use of whatever tools they can to mislead people, and so it is not enough for us to be right – we must also present our findings beautifully.
Overall though, Harford’s book is a masterpiece of statistical communication. People who follow his golden rule – to “stay curious” – should read this and subscribe to its lessons. Anyone who tries to communicate with statistics should read it, and learn from his example.
Dr Michael Sanders is Chief Executive of What Works for Children’s Social Care, and a Reader in Public Policy at the Policy Institute, King’s College London.