Collaboration happens when institutions find a genuine, shared stake, and when the programme gives them the time, resources, and structure to act on it.

Over the past two years, I have helped coordinate a partnership among four leading Bangladeshi think tanks: BRAC Institute of Governance and Development (BIGD), Centre for Policy Dialogue (CPD), Policy Research Institute (PRI), and Research and Policy Integration for Development (RAPID). This partnership happened under the UK Foreign, Commonwealth & Development Office (FCDO)’s Transformative Economic Policy Programme (TEPP).

After wrapping up the first phase of the initiative, I reflected on what actually makes a consortium work. From my vantage point in the coordination engine room, one lesson stands out: bringing institutions together on paper does not automatically translate into collaboration in practice. 

A shared logframe is not a shared incentive

As the convenor, BIGD set out with a clear theory of change. We wanted to bring these institutions closer through shared platforms, joint research, and cross-learning. Bangladesh’s knowledge ecosystem is often fragmented, with predominantly short-term projects and limited coordination among think tanks and academic institutions. The consortium then offered an opportunity to pool our different strengths.

Each institution had its own goals and funding streams, along with a few joint outcomes, and all four partners reported jointly against a shared programme framework and logframe. The assumption was that if institutions share objectives and report against the same framework, collaboration should follow. In practice, it was not that simple.

The four institutions brought complementary strengths, but their research pipelines did not always intersect. Because we often worked at different altitudes, albeit in the same economy, our day-to-day research pipelines rarely intersected enough to co-author a paper, which we thought would be a natural next step.

The bigger challenge was resources. Each organisation was already operating at full capacity, delivering on its own mandates and existing commitments. Without a dedicated mechanism or budget for joint work, collaborating essentially meant asking researchers to carve out unbilled hours. That taught us an important lesson: a shared logframe is a reporting mechanism, not a collaboration mechanism.

Collaboration sparked by timing

If collaboration does not emerge naturally from institutional structures, can it be created? Our experience suggests that sometimes an external event can provide the catalyst.

Access to reliable public data was a structural challenge each of the consortium partners had navigated independently, but what turned that shared problem into a shared output was timing. Following a July 2024 uprising, Bangladesh entered a period of transition. The newly formed interim government established a number of reform commissions and actively sought input on how state institutions should function. Suddenly, the consortium had both a shared problem and a shared opportunity: a documented systemic challenge that mattered to all of us, and a policy environment that was receptive to new ideas.

We finalised the report, jointly signed by BIGD, CPD, and RAPID, in early 2025. A few months later, when the government formed a national task force to overhaul the statistics bureau, leaders from consortium partner institutions were appointed to it, and experts from the consortium were brought in to contribute to the consultation workshops. 

When a critical policy window opened, the consortium gave institutions a way to respond with a collective voice rather than as fragmented organisations. Collaboration did not emerge because our research agendas suddenly became identical. It emerged because an important problem created an intersection between them.

Collaboration sustained by design

If external events can spark collaboration, the next question is how to sustain it in everyday practice. For us, the answer was to identify something that every institution genuinely had a stake in: developing the next generation of researchers.

Maintaining a strong talent pipeline is everyone’s business. Regardless of thematic focus, every think tank needs young researchers who can develop rigorous research skills, gain practical experience, and build professional networks. This became the basis for Envisioneers, a platform we set up to connect and support young researchers across think tanks and academic institutions in Bangladesh. Through the platform, we arranged learning initiatives on areas such as survey design, research communication, and coding, bringing together young researchers from the consortium and beyond. 

These initiatives eventually culminated in our jointly designed call for proposals for the Research Initiative on Development, Governance, and Economy (RIDGE). It was designed to support the next generation of social science researchers through funding, training, and mentorship. Rather than asking young researchers to submit polished proposals at the outset, we invited short expressions of interest. Shortlisted research teams were then paired with expert mentors from across the four institutions and beyond. Mentors and applicants worked together to develop their ideas into full proposals, with 10 grantees ultimately selected collectively by the consortium’s senior leadership.

RIDGE worked beautifully because it cracked the code on two fronts. First, it tapped into a universal shared stake: every institution in the consortium, regardless of its thematic focus, needs a strong talent pipeline of junior researchers. We did not need our research and methodological agendas to perfectly align; we just needed a goal everyone genuinely cared about. Second, the collaboration was resourced from the outset. The resources for collaboration were built directly into the design, allowing partners to contribute their expertise without worrying about the financial and operational logistics.

Looking Ahead

Our experience has left me with two rules for thinking about institutional partnerships.

First, optimise for intersections, not overlaps. Consortium members do not need identical mandates, research agendas, or methodological approaches. Instead, look for the cross-cutting needs that genuinely matter to everyone, whether that is developing young researchers, improving access to data, strengthening research methods, or creating platforms for policy engagement. These shared needs can become the connective points between otherwise distinct institutions.

Second, treat the connective tissue as its own deliverable. If collaboration matters, it needs to be designed and budgeted for explicitly. A shared framework can establish common objectives. Meetings can create opportunities for interaction. But neither necessarily creates the time or incentives needed for people to work together. Goodwill is abundant in the development sector, but it does not pay for a researcher’s time.

For funders and programme designers, the implication is straightforward. If collaboration is an intended outcome, it needs to be treated as an activity that requires resources, not as a by-product of placing institutions under the same umbrella.

Our experience in Bangladesh suggests that the most sustainable collaborations may not come from forcing institutions to overlap. They may instead come from identifying where their interests intersect, and giving those intersections enough structure and resources to become something more.

RELATED POSTS