The test of a capacity-building partnership is not what was produced during it.
It is what still functions five years after the funding ends — and by that measure a great many partnerships build nothing.
What genuine support builds
- People who can now do things they could not before, including train others.
- Methods that continue to be used locally.
- Infrastructure that is maintained, not just delivered.
- Institutional processes that outlast the project.
- Relationships that persist without funding.
Common failure modes
- Extracting data with nothing transferred.
- Equipment delivered without training, consumables or maintenance.
- Training with no positions for trained people to occupy.
- Local partners as recruitment sites rather than intellectual partners.
- Everything ending when the grant does.
Designing it properly
- Local partners in question-setting, not only delivery.
- Training as a funded component, not an incidental benefit.
- Local first-authorship on suitable outputs.
- Budget allocated to local institutions, including indirect costs.
- Maintenance and running costs included, not just acquisition.
- Handover planned from the beginning.
Transferring methods properly
- Documentation that works without the original team.
- Training more than one person — key-person dependency defeats the purpose.
- Local ability to troubleshoot, not only to operate.
- Access to protocols and code afterwards.
- Test it by having local staff run it unassisted before you leave.
Evaluating a partnership
- What can local partners do now that they could not before?
- Who leads outputs?
- Does the work continue without external funding?
- Are trained people still in the system?
- Publication counts answer none of these.
For local partners
- Negotiate capability transfer explicitly, in the agreement.
- Ask what remains afterwards.
- Ensure your people are trained, not only employed.
- Retain access to data and methods.
- Seek to lead at least one output.
One thing worth remembering
Ask what will still be working five years after the funding ends.
If the honest answer is nothing, the partnership produced outputs rather than capacity — and those are not the same thing, however the final report describes them.
Câu hỏi thường gặp
What does genuine capacity support build?
People who can do things they could not before including training others, methods that continue locally, maintained infrastructure, institutional processes, and relationships that persist without funding.
What are the common failure modes?
Extracting data with nothing transferred, equipment without training or maintenance, training with no positions to occupy, local partners as recruitment sites, and everything ending with the grant.
How should capacity support be designed?
Local partners in question-setting, training as a funded component, local first-authorship, budget to local institutions including indirect costs, and handover planned from the beginning.
How are methods transferred properly?
Documentation that works without the original team, training more than one person, local troubleshooting ability, continued access to protocols, and testing unassisted operation before leaving.
What is the test of a capacity partnership?
What will still be working five years after funding ends — if nothing, it produced outputs rather than capacity.
