10 Frontier Technologies for International Development

November 2, 2016

Is Good Advocacy a Science or an Art (or just luck), and how can we sharpen it?

November 2, 2016

Why aren’t ‘Diaries of the Poor’ a standard research tool?

November 2, 2016
empty image
empty image

I’ve been having lots of buzzy conversations about diaries recently. Not my own (haven’t done that since I was a dear-diaryteenager), but diaries as a research method. The initial idea came from one of my all-time favourite bits of bottom-up research, the book Portfolios of the Poor. Here are the relevant paras from my review:

‘A financial fly-on-the-wall account of how poor people manage money. To find out, the researchers set up ‘financial diaries’ with 250 households in selected communities in 3 countries (Bangladesh, India and South Africa). For a year, researchers visited every fortnight and picked over people’s financial affairs. The book then assimilates the findings, and intersperses them with unmistakably real-life examples from among the 250 households (‘Pumza is a sheep intestine seller living in the crowded urban hostels of Cape Town……’)

The first and perhaps most striking finding is the sheer complexity, scale and variety of poor people’s financial activity. People living in poverty need financial skills more than the better off. Just to get by from day to day, they borrow, save, and exchange cash with a huge variety of friends, family, neighbours and institutions, both formal and informal. These last include savings clubs, savings-and-loan clubs, insurance clubs, microfinance institutions, and banks. ‘At any one time, the average poor household has a fistful of financial relationships on the go.’ Every one of the 250 households had both savings and debt of some sort, and no household used fewer than four types of financial instrument over the course of the year. Rural households have turnovers (i.e. total cash flows in or out) between 10 portfolios-of-the-poor-coverand 30 times greater than their asset value at the end of the year.’

That was 2009, and the book’s been at the back of my mind ever since, but has recently come up when discussing entirely different issues. Why not use the same approach – deep, fine-grained observation of how poor people actually lead their lives, with researchers building up relationships of trust through repeated visits and conversations – on, say, how poor people access justice, or experience the state, or the environment? Or how they understand health and deal with people getting sick?

The new research consortium on empowerment and accountability in fragile and conflict settings has picked this up, so it looks like we can run some pilots to develop a methodology – v exciting.

To do it, you need sharp, respectful researchers able to go back to the same households repeatedly over a year or two. Aka students – a Masters student could easily use this as a topic for a dissertation, and receive a stipend that would help them through college. So the approach would lend itself to a partnership with a local university, which can find, say, 10 students who take on 10 families each.

But it could also go further than Portfolios of the Poor in terms of agency. Rather than the researcher coming for their fortnightly download over a cup of tea, the family itself could be asked to record their impressions and experiences – perhaps with a video diary on a phone or flip cam, and then the researcher could come and triangulate that material with their own conversation.

My colleague Irene Guijt was unconvinced by this (and coined the wonderful acronym IIMBSD – If I May Be So Dutch in her reply). What’s the benefit of all this research she asked, either for improving the actions of outsiders or to the people themselves? Provisional answers:

For aid agencies and governments: Portfolios of the Poor was designed to identify where poor people already had all

Diary of a Victorian Child Worker

Diary of a Victorian Child Worker

the financial products they needed, and where new market-based mechanisms might help (it turned out people were fine on short-term savings, but had a gap on long-term provision on things like pensions). The governance equivalent would be to identify what institutions (formal and informal) actually matter to poor families – a pretty good starting point for thinking who to partner with, for example.

Poor people would benefit because the research, especially if it focuses on agency, can be empowering in itself – being listened to respectfully, the act of recording and analysing your own interactions with authority. The chair of the research consortium, John Gaventa, made his name doing exactly this kind of thing in the Appalachians in the 1970s – working with poor families to research who owned their land, with transformative impacts on social organization. Looks like we may be coming full circle.

So I’ll end with the usual questions – what am I missing? Where has it happened already? What do you think?


  1. Hi Duncan,
    If you haven’t already, you should get in touch with Stuart Rutherford – one of main authors of Portfolios.
    There’s a fascinating couple of blogs he’s just written up on his latest portfolio’s research over at the Global Development Institute:
    …and a bit of additional background on how the original idea was hatched with David Hulme: http://blog.gdi.manchester.ac.uk/want-know-poor-people-really-manage-money/
    Stuart’s email is at the bottom of his posts.

  2. Perhaps because sadly all too often in development especially the research side it is very much about experts making sense of other peoples lives. Often with very good intention and some good results but actually handing the keys over to the poor…?

  3. Thanks Duncan. I use financial diaries and similar methods regularly in my research and teaching. I find them important for a few reasons. Yes, the act of participating in these discussions can be interesting for participants but I think the more important element is supporting them in using this knowledge to make informed choices about options. For example, gold jewelry is a woman’s own asset but gold bars are more liquid with higher returns. The second reason is that these methods show us the sheer complexity and sophistication with which households of limited resources use financial instruments, formal and informal. Multiple instruments and in dynamic ways. Understanding these choices that are often choices (even distress sale of assets), and not a lack of financial literacy, help programs design supports in ways that don’t bias particular choices. For example, we often see that poor households choose to retain informal instruments even while they access formal finance. The last reason relates to evidence-based research and policy influence. These are great tools to add to our mixed methods to try to get at complexity. We need methods like these that get right under the hood to help round out, speak to and, at times, challenge what might be coming out of “big data.” Critical so that our policy doesn’t take normative stances about “better” paths for people and households based on growth and formalization.

  4. Hi Duncan – you may be interested to look at Growing up on the Streets research project, which has worked with street children and youth in three African cities over the last 4 years. While not using a written diary (to overcome literacy barriers and safety issues) the young people met with a project worker once every week for 3 years to talk about their experiences of growing up. These ‘diary’ accounts have been transcribed for analysis to provide a very detailed picture of everyday life. Additionally, the young people were trained in research skills to extend these conversations through social networks and increase participation in the project. The early findings and research tools can be found via the project website link http://www.streetinvest.org/guots

  5. Hey Duncan,

    You are talking about listening. As you point out, something longer and deeper than, say, high-touch constituent feedback. There might be a connection with your thoughts and Dayna Brown’s 2012 work at CDA. http://cdacollaborative.org/publication/time-to-listen-hearing-people-on-the-receiving-end-of-international-aid-a-presentation/

    I assume you’ve read “Time to Listen” but there are others doing similar work – and usually it is around empowering local actors or supporting local initiative. Peace Direct uses some similar ideas in their work as well.

  6. In conjunction with a cohort study in rural Malawi (1998-2012) , we commissioned ethnographic journals in order to learn what people said to each other about AIDS in public spaces in their communities (e.g. in a bar, waiting at the borehole, watching a fight in a market between a wife and her husband’s girlfriend). The voices of the poor are strikingly different in the journals than in formal settings such as a survey interview, a qualitative interview or a focus group, or what is said in the media. The 22 participant observers were asked to simply listen as they went about their daily business, and then in the evening to write what they had heard in a common school notebook. Between 1999 and the present, they produced more than 1300 journals, each about 20 typed pages; most journals consisted of several, sometimes many, conversations. The most frequent were about marriage, (e.g. fidelity issues, divorce), but they also talked about prices for basic commodities, work (e.g. farming, casual labor, doing business), religion, politics and sport. So far, we know of 20 publications in peer-reviewed journals that have used data from the journals, as well as a forthcoming (Feb 2017 book: “A Fraught Embrace: The Romance and Reality of AIDS Altruism in Africa”, by Ann Swidler and Susan Watkins), The journals have been anonymized and are available for public use at


  7. I lead a financial diaries project in Kenya with about 300 families, based on the methods of the Portfolios of the Poor authors. (http://fsdkenya.org/financial-diaries/)

    I’ve actually heard from lots of researchers trying to modify this kind of methodology to look at other things like energy usage and livestock decision making. So maybe it is becoming more of a standard tool, when it’s the right tool for the job.

    I think there are a couple of barriers that keep diaries from being embraced more broadly:
    1. It can be hard to get right. Building relationships with respondents means investing lots of time to let them warm up, getting researchers who are committed to a long term project, and making sure researchers have the right incentives to really listen and capture key qualitative insights.
    2. People sometimes struggle with how to report this kind of data since it’s not representative and some of the best stuff is qualitative.
    3. Some think it’s not “efficient.” But in reality repeat interviews can be done on many scales. Some are hoping technology will offer a new way to collect diaries data, though I’m more of a skeptic there. Self-reporting can be a lot to ask of respondents, and you miss chances to build relationships, which are key for data quality, getting good insights, and making the process meaningful for respondents. Diaries are in some ways inherently costly to conduct, even if you just think of time alone, so to make them more “efficient” you have to get more out of the data you collect.

    We’re putting a lot of emphasis on that in our project. We use money as a lens, but that also gives us an important view into other aspects of life, like health, education, gender, citizenship, coming of age, social relationships, etc. In Kenya, we’re doing analyses on these kinds of topics, trying to get as much value as we can out of the investment our funders, staff, and respondents made in capturing some incredibly rich quantitative data and stories.

  8. In my experience it is important to work with low-income and disadvantaged groups when doing research on poverty and poverty reduction. Research questions and methodologies can then be agreed. At IIED Human Settlements Group we have a long experience of doing this. I have recently been collaborating with the Ugandan affiliate of Shack/Slum Dwellers International in research for the ESID research centre at the University of Manchester. Interested readers can learn more about it here:
    The key point is that appropriate methodologies are not decided just by academics but are determined in collaboration with low-income and disadvantaged groups themselves. And community members are active as researchers.

  9. Fascinating stuff! It also resonates with the PhD research I have been doing with informal settlement dwellers (also part of SDI) in South Africa and Zambia. Although it was not quite having the methodologies decided by the people in communities, as suggested by Diana Mitlin, it focused on people using cameras and diaries to record “how change happens” for them, which we then analysed together. The research highlighted the strong agency of some people in informal settlements, as different kinds of leaders as well as the detailed and diverse strategies they used to support their change-making efforts. They also had wonderful advice for other communities seeking change, advice which could be applied into lots of Australian communities. I am still in the laborious process of writing up the research for my dissertation, but in the meantime, people can look at my three minutes thesis https://www.youtube.com/watch?v=_Dz8S6eOZM8

  10. To answer your question: Because for years and decades economists and political scientist in particular have successfully convinced donors that qualitative methods like diaries do not produce good, relevant and reliable data; surveys and nowadays experiential methods are the only way to collect and analyze-far away from the field on your computer with STATA software…

  11. Hi Duncan and those leaving comments,
    Yes, diaries as a research methodology have the potential to impact profoundly on a number of fields of expertise and are as relevant for justice issues, energy issues, agric and medical work. At L-IFT we say that diaries are an appropriate methodology for behaviour that fluctuates and/or has trends over time and that is complex. Moreover, it is particularly good for issues that require trust, i.e. those issues people are not comfortable talking about to outsiders as is common in a survey. L-IFT is a research company that I founded in January 2015 and as a company we fully focus on diaries research. We are further developing the methodology to make it self-financing and to make the research process so useful to all involved that it more than justifies the time invested. Our researchers are local (rural) high-school graduates who conduct the research as a professional training preparing them for “customer centric” jobs (e.g. mobile money agents, solar energy agents, loan officers…), we put efforts into making the research directly beneficial to the respondents who tell us that these question sessions help them getting better grip on their lives, their decisions and help them see ways of changing their behaviour that may lead to better outcomes. The research is partly steered by the respondents who can direct some of what we research. At the end of the research period we give the participants feedback on the findings and give the individuals details on how their behaviour compares to others in the sample. While we appreciate the above comments on qualitative research, we also aspire to building this movement to such a scale that we have data with statistical power and reasonable representation to ensure that policy makers will take our data seriously.
    We are looking for collaborations with a number of organisations and individuals. All are welcome to use our data, we actively promote that a range of organisations request the data that they need, we are also looking for organisations who are interested in hiring our researchers after they complete their 8 months of experience/training.

  12. Hello Duncan
    Thank you for your blog on Diaries as a research tool – it is great to scroll through and see how diaries are being used. I used the qualitative diary research method in my ethnography of an NGO in my doctoral research. My interest was in gender and organisational change and getting beyond the formal commitments to gender mainstreaming and finding out how gender shapes organisational culture. The diaries in my research were kept by the staff of the NGO. I was using participatory methods ( diaries and photo-voice). As in the comments above, diaries produce a depth of data that is beyond the scope of an interview. In organisational research it is likely for example that the diary writers will be reflecting on the same organisational incident and this produces multiple perspectives and nuanced understandings. The diary method is an effective tool for breaking the silence in organisations around organisational culture. I am a strong advocate of diaries for research and writing as a tool for personal reflection. I use the Personal Learning Journal in my practice as an OD practitioner – journal writing enables personal reflection and can build confidence to share personal stories in a group. I use diaries and journaling for data collection and intervention.

Leave a comment

Translate »