How can aid workers study without giving up the day job? Your thoughts please.

February 4, 2014

My first trip to Central Asia. First impressions of Tajikistan, world’s most remittance-dependent country (and a very big flagpole)

February 4, 2014

Voices of the Hungry; killer indicators, and how to measure the social determinants of health. New thinking on measurement with Gallup Inc.

February 4, 2014
empty image
empty image

About once a year, I head off for the plush, Thames-side offices of Gallup Inc, for a fascinating update on what they’re up to on maizedevelopment-related topics. In terms of measurement, they often seem way ahead of the aid people, for example, developing a rigorous annual measurement of well-being across 147 countries. Not quite sure why they talk to me – maybe as part of the wilder shores of their business development – they know they won’t get much business out of it, but some useful ideas might come out of the discussion. This time, Katherine Trebeck, Oxfam’s wellbeing guru (only she prefers to call it ‘collective prosperity’ for some reason) and developer of the Humankind Index, was there too, which added some actual knowledge to our side of the exchange.

First up was Gallup’s partnership with the FAO on their ‘Voices of the Hungry’ project, aimed in part at correcting the alarming weakness of the numbers on hunger (see Richard King’s 2011 post on that). After pilots in Angola, Ethiopia, Malawi and Niger, in part supported by the Government of Belgium, FAO has now got DFID funding to go global, initially for two years. Through ‘Voices of the Hungry’, FAO has developed the Food Insecurity Experience Scale (FIES), modelled on the 15-item Latin American and Caribbean Food Security Scale. This uses interviews to place people along a spectrum from worried about food to seriously hungry.

Food insecurity spectrumWhen it goes global, the FIES will be included in the Gallup World Poll, which means the FAO will be able to obtain national measures of food insecurity in all countries covered by the GWP,  but, very importantly, also to compare the prevalence of food insecurity with issues such as jobs, corruption and migration. As the poll is with individuals rather than households, it will also allow an exploration of gender differences in food security. Down the line, Gallup is wondering if something similar could work on water and sanitation, which are also dogged by dodgy stats. Watch this space.

Next, some intriguing thinking from Katherine. The world is now awash with alternative indicators, many trying to go ‘beyond GDP’, add to thinking about sustainability and measure wellbeing, progress, aspects of poverty and so on. Most sink without trace with policy makers, and barely register with the public. new economics foundation (with its annoying, Martin Lukes-style lower case) is trying to understand which indicators get traction with policy makers through the Bringing Alternative Indicators into Policy, or BRAINPOoL project (what is it with these guys and upper/lower case?), which is presenting its final conclusions at a conference in Paris in March.

Katherine is thinking more about the public communication angle. She wants to identify ‘totemic’, ‘gateway’ or ‘killer’ indicators that resonate with what people say is important to them, on their terms and in their language, but which when unpacked, contain relevant information for policy makers (hence ‘gateway). Things like ‘how many kids go to school by bicycle?’, ‘do you know your neighbours?’, or ‘do you feel safe walking your street at night?’. Andrew Rzepa, the Gallup nerd in the room, labelled them ‘sexy compound indicators’.

A few questions emerged in the ensuing discussion:

–          Do they have to be comparable across countries (can’t see too many Delhi parents wanting their kids to go anywhere by bike – I’ve tried it, and it’s scary)? Or is the main issue the time series within each country, to help governments better understand what matters to their people?

–          Do we want to first develop new SCIs (sexy compound indicators, please keep up) or test the dozens of existing indicators, whether mainstream or whacky, and see which ones resonate with different sections of the public?

–          Is the search restricted to identifying killer indicators that communicate what we think is important (health, education, decent work, wellbeing)? Or are we willing to engage with a more complex range of topics that is genuinely bottom up but may not have any obvious policy implications (how my favourite soccer team is doing, how much time I get to spend partying, watching TV or shoe shopping, can I go to Church/mosque as regularly as I would like)?

dilbert auditingFinally, Gallup is keen to work with Michael Marmot to develop metrics for measuring the social determinants of health. If you accept the argument that much of the variation in health outcomes is determined not by access to health care and medicines, but by things like poverty, stressful working conditions, poor housing and unequal access to education, then a widely accepted measure of that enabling environment could really help galvanise policy makers to take it seriously. They are currently looking for some seed funding to get this work started, so if you’re interested, get in touch.

This may all seem rather arcane/nerdy, but I’m increasingly convinced that the measurement agenda (if you can’t measure it, it probably doesn’t exist, won’t get funded and is unlikely to get much attention from policy makers) is here to stay. So we really need to make sure we count what counts, not just what is easily counted. Gallup looks likely to be a significant player in that effort.


  1. Hi Duncan,

    I might be nit-picking here, but I’m a little concerned about your interpretation of what the “measurement agenda” entails. In particular:

    “If you accept the argument that much of the variation in health outcomes is determined not by access to health care and medicines, but by things like poverty, stressful working conditions, poor housing and unequal access to education, then a widely accepted measure of that enabling environment could really help galvanise policy makers to take it seriously.”

    This puts the cart firmly before the horse. The time to get serious about measurement is in the premise, which is itself a measurable hypothesis. You use measurement of these `social determinants’ to determine a (preferably causal) relationship between them and health outcomes, then you use that to galvanize policy makers. Accepting the premise without rigorous evidence makes any subsequent measurement a bit pointless.

  2. Hi Duncan,
    thank you for the nice and informative post.

    As is Matt, I am also particular keen towards rigourous measurement (which, in the area of social sciences, sounds almost like an oxymoron)

    I gave a seminar yesterday at the FAO, talking of the analytics behind Voices of the Hungry.
    The webcast can be seen at and the presentation is available through our website

    Perhaps Matt and others among the more analytically inclined readers may find it useful. I look forward to receive comments and suggestions.


Leave a comment

Translate »