About me

Contact me

Carried Away

Doing Science

Adventures in interdisciplinarity

My name is Dan Rogger, and I am a development economist (you can find out more about me at my About me page of my main web site). This page presents occasional missives on my engagement with other disciplines. My research is rooted in economics but proper study of the delivery of public services requires myriad frameworks of analysis. If you have any thoughts on the below, please don't hesitate to contact me.

Doing science
4th January 2013

The Magic Circle is "the world's premier magic society". One evening recently I was treated to an evening of magic there. We saw hundreds of coloured handkerchiefs be drawn from an empty box, superb mindreading and a goldfish appear from thin air. The magic was that which had enthralled me since I was a child.

As I have become increasingly embedded in research, the wonder of science has become as enchanting as magic. There are many affinities between the two. Both draw you in and reward you for your singular attention on a precise yet relatable topic. Sometimes the result is what you expected and somtimes it is a surprise, but it is always satisfying to see it done well. The panache of execution is important, but nothing beats a good idea.

For me however, there is a big difference between the two. Amongst the applause, the evening at the Magic Circle was a night of mixed emotions. The tricks seemed more illusory than ever before. In science, the magic is in the doing. It is in the observability of the careful building of structure. Science conjures its magic from the bottom up. With illusion, the preparation is necessarily smoke and mirrors. It is built on hidden mechanisms so that the audience will be amazed by the resulting outcome rather than be indifferent to a clockwork response.

The doing of science has overtaken my love of magical theatre. Whilst I still delight in the simple card tricks my Dad performs when guests come round, it is nothing like the creation of new knowledge. The thrill of measurement, analysis and argument.

Yet we still present so much of science as magical theatre. We shy away from presenting the process of our work. Last year I thought of tracking an entire paper of mine from start to finish. Then I realised this was an idea whose time had already come. A few brave researchers have already presented us with an exposition of the creative process. David Roodman opened up a blog that tracked the writing of his book, 'Due Diligence: An impertinent inquiry into microfinance'. Ben Fry has illustrated the many editions of Darwin's 'On the Origin of the Species'. In the sphere of fiction, Alan Yentob followed Ian Rankin as he wrote a novel in BBC One's 'Ian Rankin and the Case of the Disappearing Detective'.

Exploring and comparing these efforts allows us to better understand the creative process. Numerous postings could be made on what is shared across disciplines in that creative process. I will focus on one: the iterative nature of doing science. In all of the examples above the researchers used the process of building an argument to refine it and better understand their topic. In no case did the completed result present itself in a first attempt. Rather, the creative process across the disciplines is characterised by toil and response as much as a brilliant initial idea. As Ben Fry states about Darwin,

We often think of scientific ideas, such as Darwin's theory of evolution, as fixed notions that are accepted as finished. In fact, Darwin's On the Origin of Species evolved over the course of several editions he wrote, edited, and updated during his lifetime.

Looking at only one discipline, it would be hard to determine that this was a common theme across research. Seeing that it is, it feels much easier to get up and move on when you are felled. It's what all the great scientists, in every field, have done.

The commonalities of the creative process across disciplines will surely be matched by differences I have not discussed. One commonality seems robust. The nature of doing science may vary across disciplines, but the magic that arises from its being done well is universal.

Back to top

Culture club
14th December 2012

There are many criticisms of economics. One I have often heard is that we abstract from the social environment and particulars of the local culture. Recently I have started worrying about that.

The last couple of years have seen an increasing number of economics researchers working in Nigeria. That is an excellent change, as Nigeria has long been under-researched in most fields. However, since the country is so close to my heart, observing the way they do research is a window into the frustrations other fields have with economics.

Some of these researchers have found themselves working on Nigerian data without much knowledge of the country, its context, or its many cultures. One could argue you could never know a country as diverse as Nigeria, but such a discussion would come down to one of degrees of knowledge. You can gain degrees of knowledge of Nigeria.

Confronted with this lack of contextual understanding, many of these researchers seem to then expend little effort to understand the environment in which their data was collected, nor the economic context in which their respondents live. One group asked me whether their study area 'was mainly rural, or more urban'. They didn't ask me very much more than that, nor looked anywhere else for information.

One can definitely forgive researchers who are wary of a country with the reputation of Nigeria. But how can you do good science when you have such a limited understanding of the institutional context? Making decisions over the smallest things, like what a respondent might perceive a survey question to be about, or what matters in community decision making, all requires knowledge of a country's context and the cultures within it. Are the anthropologists right? Do economists do economics at their peril?

A respected American (economics) academic gave me a fantastic piece of advice just when I was starting my own academic career. 'Learn what makes the people you are studying tick. Learn what their preferences are and how they would respond to different scenarios. Spend a lot of time getting to know as many of them as possible. Then you will understand the strengths and weaknesses of the particular measurements you have in your data, and what correlations between these measurements is most likely to imply.'

I believe the standard models and theories of economics are powerful unified frameworks for understanding the motivations behind human behaviour. However, they are blunt and sometimes dangerous tools unless measured against and embedded in the cultural, social, and institutional context of the study.

This is a diluted form of the criticism I have heard from many other fields about economics in the past. Today I have empathy for them. However, this is a criticism of doing economics the wrong way. Doing economics the right way (by doing what the professor advised me) is the stuff of brilliant science. The professor in question's research is a great example of a body of work that follows that principle religiously. As a result, it's awesome.

Economics does itself no favours at all when it produces research that is so clearly in line with its fiercest critics. And so if anyone would like assistance on getting to know Nigeria, I am very much available to point you to some resources I have found useful. Whilst I don't believe it will solve the problem, you are also welcome to borrow my Yinka Ayefele CD.

Back to top

9th March 2010

This afternoon I met a self-proclaimed charlatan of the academic world, Professor Murray Last. It was an extraordinary experience just talking to the man. The lunch was topped off by his final words being, "Don't believe a word I've said! Only believe Allah." He was the first person to receive a PhD from a sub-Saharan university, as he was coincidentally graduating from the University of Ibadan just as it was becoming independent from the University of London. He was one of the founders of medical anthropology, invented the term 'the Sokoto Caliphate', and for the last 35 years or so, has spent time each year living in a rural farmhouse in Nigeria, gaining a unique insight into the lives of the farmers and their families there.

"This longevity of contact," he told me, "is the most credible approach to understanding these communities. Otherwise, how do you know what is going on?" For every answer I tried, he had an anecdote. On household surveys, he told me that he had once surveyed all 300 households in the local village regarding their health status. "The most prominent pattern in my data was that as the interview time got closer to lunch, people got healthier! They just wanted to get me out the door." Theoretical models could be manipulated. Data could be tortured to say anything. And who could rely on it? "The only way I really knew how many children had died that year was to go and count the graves myself."

His message was this: the most reliable way of finding out about people was to sleep in the same hut as they did. You had to be with them day and night. "Most of the interesting stuff happens at night." Without this, you would miss the important stuff.

Of course there are arguments that go the other way. "Oh yes, people lied to me all the time. They were probably lying to themselves a bit as well. But you learn to look out for that." There is little you can say to a man so experienced that he hasn't heard before. As an economist, I build stories, and then test for them in data. I believe there is tremendous power in doing this. There is probably also tremendous power in sitting and talking with people about what the treatment means to them, and why they responded in the way they did.

I'm a charlatan, he kept telling me. But a charlatan with a point.

Back to top

22nd January 2010

I began reading the engineering literature a couple of months ago, and it was painful. The language used was alien. The structure of the papers unfamiliar. And where was the data?! After months of slowly nudging my brain into action, I had a eureka moment - I understood something. Then I put it in practice, and I think it worked. And even better, I think there are a few ideas in economics that would fit nicely into the engineering literature.

There are three things worth highlighting here. One, the entry costs seem high at the start. Two, it turns out they aren't that high. Three, talking to people really helps. The academics I talked to helped me get a shape of the literature much faster than my reading was.

Back to top

2nd September 2009

A colleague and I are writing a research proposal that crosses service delivery and sanitation. To do this properly, we need an epidemiologist. We met a couple. They're brilliant. And they have so many tips and tricks! Ways to compute things, ways to analyse things, computing shortcuts, that are just great. Why can't we have more cross-discipline sharing of the sweet little stuff that makes life easier? This is a call for a book on the tricks and tips of the academic world - no disciplines barred.

Back to top

Let's not get carried away
31st October 2008

Having become involved in a couple of UCL Institute for Global Health activities, the Institute asked me to write a piece on 'why interdiscplinary?' You can read it here. However, I wanted to highlight the final section of the article. After 3 pages of extolling the vitures of interdisciplinarity, I wrote:

I don’t believe interdisciplinarity is a silver bullet ... There is no substitute for rigorous thought in one’s own field. The clarity of thought in interdisciplinary work will critically rest upon the clarity of thought in those involved.

Interdisciplinary work can also be prohibitively costly. As Nicholas Saunders of UCL, who has written on interdisciplinary work, argues, “it requires hard thinking, difficult choices and doesn't always work.” Scholars from distinct disciplines often find there is a ‘language barrier’ through which misconception and miscommunication can flow. Interdisciplinarians may find it difficult to agree on an integrated methodology and to satisfy their own discipline's demands for accountability and publication.

The challenges may go even deeper than that. Diana Rhoten, of the National Academy of Sciences, writes of “the incompatibility of university incentive and reward structures with interdisciplinary practices.” Her study of interdisciplinary work found only a gradual move towards interdisciplinary work because of this weakness.

Back to top

12th February 2008

The Institute for Global Health, a fantastic new platform at UCL, held a syposium this evening entitled 'What's justice got to do with it?'. The aim of the event was to discuss the role of justice in the study of global health. I found myself sitting there frustrated. 'We economists have the answers to these questions,' I thought to myself. 'Why are other disciplines still grappling with them'. This was perfect preparation for Anthony Costello's interjection, half way through the meeting. Prof. Costello, director of the Institute, turned to the audience and said, "I bet you're all sitting there thinking, 'my discipline has all the answers to these questions; why can't everyone else realise we're right?'"

Nice one. His comments reflected that my frustration wasn't to do with economics being superior, but that I couldn't see the breadth of the question. It implied many of the researchers in the room faced the same challenge. It is difficult to evaluate the content of a question from a perspective that isn't ones own. So the answers given by those from other disciplines can seem to be missing the point. Of course, sometimes they are. There is no substitute for clear, analytical thought. However, there are also great complementarities between the distinct frameworks we might use to assess the challenge of global health. So if I don't get why someone is discussing the impact of a treatment on the freedoms of the treated, maybe I should try harder. Knowledge is just elusive, so any one who takes a serious stab at gaining it should be respected.

Back to top