Lessons on fostering collaboration: learning so far

27 July 2015
SERIES Articles and Opinions

The first phase of the On Think Tanks Exchange (OTTE) comes to an end on September 2015. We are planning a meeting in Rio for the week of the 14th, hosted by CEBRI. This will be an opportunity for the teams to report on their projects and for the organisers to report on their effort to learn about collaboration. It is also a chance to put forward possible future ‘phases’ of the initiative.

In this post we share some of the lessons we have learned with the intention of opening up a discussion with the participants and other interested parties.

Over the course of the project we have learned a number of lessons, which can be taken into account in thinking about a possible second phase. These lessons include:

What to bond over?

  • It has been easier for participants to “bond” over Organisational Development (OD) issues, and all the teams formed around such projects. This is probably because their interests in policy issues were too diverse to begin with. It might have been better to focus the call for applications on one or two substantive policy themes in order to improve the chances of teams forming around projects on policy issues instead of on OD issues.
  • Organisational development issues did, however, prove to be a very useful subject to bond over. The participants themselves recognised this: since they were all drawn from the “research” side of think tanks, the projects gave them an opportunity to examine their home institutions from a different perspective. This can not only make a useful contribution to their current work, but also provide them with a valuable perspective should they move into more senior positions in their think tanks.
  • The ‘stress’ of putting together a proposal in an area where they were not specialists helped with the ‘bonding’ process. However, the process can benefit from more structure and guidance from the facilitators.
  • Match-making is a process and it needs to be given the space to happen more organically and allow for progress to be based on learning. Two or more phases could be employed so that learning is built into the process.

Changes, changes, changes

  • Changes in staff and organisations are inevitable in a two-year project, particularly when the participants are young researchers rather than senior staff (which was the case for most of them). They are more likely to move from one organisation to another. This turnover must be built into the design so that it does not disrupt relations between The Exchange and the think tanks or between the participants themselves.

People, people, people …

  • The participant’s personalities as well as cultural differences and ways of working have had a significant impact on the projects. These need to be “flushed out “early on during the matchmaking stage but, equally, they need to be carefully addressed throughout the project: it is unrealistic to expect the participants to deal with such issues on their own.
  • “Multilateral” (as opposed to bilateral) teams have made it easier to manage these unexpected changes and personalities but have increased the transaction costs involved in developing trust and launching the projects.
  • Larger teams made it possible to spread some of the risk in taking on a project on a topic in which no one was a specialist, and this may have encouraged the participants to form such “multilateral” teams.

Organisational linkages

  • The participants were relatively mid-level to junior staff within their organisations, and more researchers than managers. This has many advantages over the longer run, but if the aim is to engage the institutions more deeply by involving their leaders, this is likely to prove difficult for small projects like the ones supported by The Exchange – even if these are focused on OD. Leaders need other ‘excuses’ to get involved in the collaboration.


  • A reasonably long period of time is needed to ‘build trust’, but two years may be too long from the point of view of staff turnover. This is also a lengthy commitment for a researcher. It may be possible to build trust among the participants more quickly (e.g. within one year project or less) by building in more frequent and intensive engagement during the initial stages.


  • Administrative and logistical support for the teams has been limited to the meetings but has been of great help. It could be extended to the projects themselves through a more intensive mentoring or coaching process.