Problem-driven Design: Zooming in for a Ten Millimeter View

T4DThis blog post is the first in a new Transparency for Development series “T4D: Views from the Field,” written to highlight what members of the T4D team have observed in launching a co-designed intervention in Tanzania and Indonesia that seeks to empower citizens to improve maternal and newborn health in their communities. 

In the first post from this series, Courtney Tolmie shares take aways from observing the intervention in five villages in Indonesia earlier this year, focusing on problem-driven design and what this theoretical approach looks like on the ground for T4D.

Read other posts from the T4D project here.


By Courtney Tolmie

Transparency and accountability stakeholders spend a lot of time differentiating between the way things used to be – and the way things are now.  We used to talk about accountability triangles, but now we know it is really all about ecosystems.  People once thought information was sufficient to create participation, but we have wised up and now know that is not the case.  And in the good old days, we were into tools.  But we have learned the error of our ways: problem-driven design is the way of the future.

And unlike fads like slap bracelets and (I hope) Pokemon Go, I think this one is here to stay because (unlike slap bracelets and Pokemon Go) taking a problem-driven approach makes sense.  Tools tell us to do an absenteeism study or to track expenditures, while problem-driven design asks us “But what if those aren’t the biggest problems in the sector?  What if they aren’t the only problems in the sector?”

The downside of choosing problem-driven design over a tool is … well, it isn’t a tool.  It is a concept that may make sense in the abstract, but it is harder to think about as something that can be designed and implemented.  In 200 villages.  By people who are not elected officials or professors or policy wonks.  In other words, the ten-thousand foot view makes a lot of sense, but the ten millimeter view?  Not so much.

My colleague Jessica and I recently visited several of our project sites in Indonesia, giving us an opportunity to observe firsthand how the intervention was working – including a deeper dive into one way to concretize problem-driven design.

This post is one a series reflecting on the recent site visits, starting with some big questions about problem-driven design:  What does a problem-driven approach on the ground actually look like?  And what can practitioners and researchers alike take from observing it in action?


Photo credit: Transparency for Development/Courtney Tolmie


Problem-driven means story-driven, not just data-driven.

The first thing to understand about problem-driven design in practice is that it is not just what is doing the driving (as the name suggests, it should be the problem) – but also who is driving.  In social accountability initiatives like the one in T4D, the goal is to have communities themselves in the driver’s seat.  So what do you do if the problem alone is not enough to catalyze community participation?

In observing meetings in Indonesia, two things about our focus problem – maternal and newborn health (MNH) – became clear.  First, not all problem-focused messaging is created equal.

We observed facilitators and community representatives talk about maternal and newborn health in a myriad of ways – citing national level maternal mortality statistics, outlining results from a community scorecard that the T4D facilitators undertook, and sharing experiences of family members and neighbors and the community representatives themselves.  And the level of enthusiasm and participation shifted over the course of the meeting.

Second, without exception, the discussion about maternal and newborn health that got people most excited was the stories, not the data.

There was a pivotal moment in one of the meetings I observed in Indonesia in which energy was fading and it was becoming harder and harder for the facilitators to draw ideas from the community representatives, when one man in the back stood up.  He shared that he had three children: one healthy, one who had died very young of a mysterious illness, and a third whose health is declining.

More than any statistic, this story was able to bring attention back to the overarching problem as well as lead into a discussion of what may have happened that the communities could work to change.


Photo credit: Transparency for Development/Courtney Tolmie


Problem-driven is not only – or even primarily – about identifying the root problem.

There is a natural tendency to think of a problem-driven approach as something that is focused on the early phases of work.  We sit in front a whiteboard, we map out what the problem is, and then we spend the next month or year or decade tackling that problem.

But in reality, solving problems happens iteratively.  People learn that their original approach didn’t work, that someone they thought would act does not, that the thing they thought was the root cause of the problem was not in fact.  And building this iteration directly into a problem-driven approach can make the difference between success and failure.

One of the intervention meetings we observed was the final T4D-organized meeting in a village, meaning it was the last opportunity for the facilitator to walk the community through a discussion of how they had improved health services – and what they had left to do.

The original problem diagnosis had happened more than 3 months earlier, but the facilitator Munir asked the community representatives to return back to their original diagnostic work.

They talked through every single problem with maternal and newborn health that they had originally identified – cleanliness of the health facility, difficulty accessing this midwife, and so on.  They celebrated their successes, for example working with the village head to ensure that the midwife moved into the village so that she was more readily available in the case of emergencies.

And equally important, they discussed what hadn’t been fixed – and what they needed to continue to monitor.  This simple step of revisiting the problem is one that may ensure that the action keeps going – and that those missteps in the path toward fixing the problem are corrected, setting the community back on the right path.


What if the problem is not actually the problem?

Sitting in the car on the way back to Makassar after observing one meeting, we asked Munir to compare the different communities he had facilitated – which ones seemed to be most engaged?  Which ones fizzled?  And what factors did he think made the difference?

Without hesitation, Munir explained that the community representatives that stayed engaged and successfully solved some of the problems that they identified were the ones that cared most about maternal and newborn health.

Doing problem-driven design in a conference room in Washington DC or Boston, it is easy to forget that the focus problem has to be not only one that exists – but also one that potential beneficiaries perceive to be big enough and important enough to fix.

Our baseline survey may have shown that there are real breakdowns in maternal and newborn health in every one of the 200 villages in which we are working.  But our observations showed that, in a few villages, MNH just wasn’t a major concern.

This became clear when we observed one group of community representatives that talked about wanting to shift attention to environmental issues.  From an evaluation perspective, this is tough to swallow.  We want to see community members gathering around an issue that they want to fix, but we want that issue to be something that affects maternal and newborn health – so that we can see if this intervention can actually improve health outcomes.

But from a governance perspective, this raises some questions about whether community empowerment can be transferred from one problem to another.  Even if a community does not stay focused on improving maternal and newborn health (maybe because they see more pressing problems around them), can interventions like this one build participation, empowerment, and skills that community members have to push them to continue transparency and accountability work – but on water and sanitation, or education, or HIV/AIDS?

In the evaluation, we will be trying to learn more about this question too, but for now, it is an important point for practitioners and donors to consider when zooming in from the ten-thousand foot view of problem-driven design.


Courtney Tolmie (@CourtneyTolmie) is a Principal Investigator on the Transparency for Development (T4D) Initiative and a Senior Program Director for the Governance and Transparency Program at Results for Development Institute (R4D). T4D is led by the Harvard Kennedy School and Results for Development Institute, in partnership with Pattiro, CHAI, JPAL Southeast Asia, the University of Washington, and coordinated by the Transparency and Accountability Initiative. T4D is supported by The Bill & Melinda Gates Foundation, the Hewlett Foundations, and DfID.

1 Comment

Filed under Transparency for Development

One Response to Problem-driven Design: Zooming in for a Ten Millimeter View

  1. Pingback: Links for 01/11/16 – Jaya Jung Mahat

Leave a Reply

Your email address will not be published. Required fields are marked *