Andrew Leigh

JEFF BORLAND SPEECH LAUNCHING RANDOMISTAS

MELBOURNE LAW SCHOOL

THURSDAY, 8 MARCH 2018

I’d also like to begin in our usual way by acknowledging the traditional owners of the land on which this event is taking place, and their elders past and present.

It is a great pleasure to be able welcome Andrew Leigh back to the University of Melbourne.

Andrew is MHR for Fenner and Shadow Assistant Treasurer.  If you watch TV news, you may have picked up that his other main political role is as the Canberra running partner for Bill Shorten.

Every now and then the media or Labor Party decide it’s a good idea to interview Bill Shorten at the end of a run, at which point you can usually see Andrew standing several metres deferentially behind.  Why do I say deferentially?  Some of you may know that a few years ago Andrew took up marathon running, and last time I asked him, he had a best time of 2:42.  I wanted to give you some idea of how impressive that is.  I was going to guestimate some numbers to do that, but then I realised that Andrew would know I was making them up, so thought I better do some proper research.  So I asked one of my friends who is a Professor of Sports Medicine at UBC how I could translate Andrew being able to run 2:42 in his mid forties into the time he would have run at his peak.  He sent me three sites which do this, which gave me times from 2:30 to 2:33.  If Andrew had run 2:33 at the Australian Marathon Championships in Sydney last year, he would have finished 11^th in the men’s category.  So if you have formed the same opinion of Bill Shorten’s running talent as I have, you’ll now understand why I used the word deferentially.

But it’s not Andrew Leigh the politician or runner who we are welcoming tonight, the Andrew Leigh we are welcoming is the public intellectual.  Those of you who, like me, are academics, won’t be surprised that since Andrew left academia and became a member of Parliament, he has had a lot more spare time.  And he has made the most of it.  Since 2013 he has published 5 books – Battlers and Billionaires, The Economics of Just About Everything, The Luck of Politics, Choosing Openness and now Randomistas.  These books are making a major contribution to the discourse on public policy in Australia.  In Battlers and Billionaires Andrew made a highly influential contribution to on-going discussion about inequality in Australia, no doubt the best general survey on the topic there has been.  In Choosing Openness, Andrew provided a comprehensive, scholarly and powerful yet balanced account of why Australia should remain open to trade, migration and capital flows.   

Now Andrew has presented us with Randomistas.  Being asked to launch Randomistas did give me some moments of introspection.  I know that when someone is asked to do a book launch their objective should be to have the audience sprinting to the exits to buy a copy of the book.  And having read the book, I do think it would be a good idea if you were to do exactly that.  But how to convince you?  Well, that’s where the introspection came in.  I thought about a decision-making process that I have gone through hundreds (perhaps thousands) of times, but never thought about consciously – buying a book.  Once I thought a bit about it, I realised that my decision-making process involves three steps: 

1] The topic – Is it interesting and important enough for me to want to read the book?

2] The author – Is the author going to be able to tell me something I want to read about this topic?  Are they an expert on the topic?  Are they likely to have done a lot of work for the book?

3] The details – If a book passes the first two tests, then the next thing I would do is to look into the book a bit more.  If I’m at a bookstore I would open it up to have a look.  Do the contents match with what I expected?  Does it seem to be well-written?  Does it seem to be well-researched?  Is there value added compared to other books on this topic that I have read?

I decided that what I should do is to use my decision-making process, these three steps, as a way to describe to you why you need to buy a copy of Andrew’s book.

First, the topic.  I assume that everyone here tonight is interested in good policy-making.  In that case you will definitely find the topic of Randomistas interesting and important.  Better policy making relies on improved evidence.  Randomised Control Trials (RCTs), which this book is about, as you will no doubt know, long applied in fields such as medicine, are increasingly being applied as a source of evidence about what works in areas relating to social and education policy, business decision-making and philanthropy. 

Second, the author.  This is also easy to address.  Since he returned to Australia in the early 2000s, Andrew has been Australia’s Ambassador for Randomisation.  He has been a consistent and strong advocate for greater use of RCTs to evaluate policy in Australia, and he has done a broad array of research using RCTs (a field experiment to investigate discrimination against minorities in hiring in Australia); and what is known as quasi-experimental analysis (impact of the Australian gun buy-back to study the impact of gun availability on firearms deaths; changes in compulsory school ages to study return to education; and of course, most famously, the impacts of the 2004 Baby Bonus and 1979 abolition of Death duties in Australia). 

With his co-author Josh Gans, by comparing the births in the week before and week after the introduction of the Baby Bonus in Australia on July 1 2004 to the same weeks in the preceding 30 years, Andrew was able to show that 1000 births had been shifted from before to after July 1 by the policy.  While this is the better known study, my favourite is actually the study where Andrew and Josh showed that 50 deaths appeared to have been shifted from the week before to the week after the abolition of Death Duties in Australia in mid-1979.  I can understand, indeed Andrew and Josh use data to show, how births can be shifted by delaying inducement and caesareans.  But the result on deaths has always puzzled me.  How do you change the date of death?  I have this image of a son, daughter-in-law and grandkids going round to visit Grandpa in the week before the policy change.  Grandpa is very still in his favourite armchair, which prompts the kids to say – Gee Dad, Grandpa isn’t moving, do you think he is OK?  And Dad replying – yeh, sure, he’s tired and just sleeping.  We’ll come back in a few days (after July 1) to see him again.  And when they return and Grandpa is in the same place, Dad says ‘Well, what do you know…’

It is worthwhile saying that the campaigning for more RCTs in Australia is starting to pay off.  State governments have established nudge units, and BETA within Prime Minister and Cabinet.  Also TTL initiative by Department of Social Services.

Third, about the details of the book. Andrew faced a bit of a challenge in writing this book.  Most of his previous books have had a focus on Australia, but as I have alluded to, if you tried to do the same thing for RCTs, it would be a short book.  So Andrew has done something much more wide-ranging and valuable with this book.  He has written what I think is – and I say this comparing across all books that are available internationally - the best and most comprehensive introduction to the ways that RCTs are currently being used around the world – in all the fields where they matter:  medicine, development economics, social policy, education, business, politics and philanthropy.  But Andrew has also given particular emphasis to Australian RCTs where they are available. 

You’ll learn about topics such as sham surgery, how a thousand research studies on learning have been based on Sesame Street, how microcredit in poor countries may not be all it’s claimed to be, whether to wear compression socks after running a marathon, how to experiment on google to find the best title for your book, and whether deep canvassing works – as well as about Australian trials such as Journeys to Social Inclusion and the Head Injury Retrieval Trial.  Andrew uses these examples to make what I think is his main argument for RCTs - that without an evaluation we can never really know if a program or policy is working. Andrew says – and I think this is a great way to put it - that the real value of RCTs is when they surprise us.

As well as learning about the research studies, Andrew tells the fascinating stories of some of the pioneers and characters of RCT history – Archie Cochrane, Judith Gueron, Eshter Duflo and Frances Kelsey.  Importantly, you’ll also learn plenty about the method of RCTs – how to go about doing an RCT that will make a valuable contribution to evidence, and ways to overcome potential problems and objections to using RCTs to study policy. 

So it’s a story book that you will learn heaps from.  I already knew a fair bit about RCTs from having being involved in several of them, and I was already a convert to their value.  But, in addition to enjoying seeing the story put together so skilfully and finding out about some research studies I didn’t previously know of, I did feel I learned one big lesson from the book.  I had always thought about RCTs that – well, they’re good, but probably a bit limited in their application.  The examples in the book have made me think that I was wrong in this perception – that with a creative approach and some hard thinking the range of applications of RCTs can be much larger than I had thought.  A nice illustration which Andrew uses at the start of the book is about parachute experiments.

There’s two final things to say about the book.  One is that it’s incredibly well-researched.  Andrew has gone roamed fare and wide, and discovered a treasure trove of academic and government studies that he uses as the basis for the book.  The other is that it’s wonderfully written.  The book is engaging and fluent throughout, and written for a general audience of readers interested in public policy.

In the end, this book is about making a case for the value, and for making greater use, of RCTs.  Those of us with a commitment to improving evidence for decision-making should feel greatly indebted to Andrew for having made that case in Randomistas in such an entertaining, wide-ranging and powerful way.