A research system can expand student numbers quickly. It cannot expand the number of people able to supervise them quickly — and that mismatch is where most capacity-building efforts stall.
The components
- People — researchers, and crucially those able to train others.
- Training — a pipeline that produces capable independent researchers.
- Infrastructure — equipment, literature access, data facilities.
- Support functions — research administration, technical staff, statisticians.
- Funding that is stable enough for multi-year work.
- Weakness in any one constrains all the others.
The supervision bottleneck
- Supervision capacity grows slowly — it takes years to produce a supervisor.
- Expanding student numbers without it produces poor completion rates.
- Overloaded supervisors supervise badly, which reduces future capacity further.
- Count supervision capacity before expanding intake.
Retention
- Training people who then leave the system builds nothing.
- Retention depends on position availability, funding continuity and working conditions.
- Return schemes without positions to return to do not work.
- Measure how many remain active five years after training — this indicator is rarely collected and highly informative.
The support functions nobody counts
- Research administration determines what researchers can attempt.
- Technical staff determine what can be run.
- Statistical and methodological support determines quality.
- These are consistently underinvested because their absence shows up as researcher underperformance rather than as a staffing gap.
What can be measured
- Active researchers and career-stage distribution.
- Supervision capacity relative to student numbers.
- Completion rates and time to completion.
- Five-year retention.
- Domestic first-authorship in international collaborations.
- These describe capacity; output counts describe activity.
Sequencing
- Supervision capacity before student expansion.
- Retention before recruitment — a leaking system does not fill.
- Support functions alongside research posts, not after.
- Infrastructure that can be maintained, not only acquired.
One thing worth remembering
Count supervision capacity before expanding student intake.
Students without adequate supervision do not become researchers — they become non-completions, and the effort produces neither output nor capacity.
Câu hỏi thường gặp
What are the components of research capacity?
People including those able to train others, a training pipeline, infrastructure, support functions like administration and technical staff, and funding stable enough for multi-year work.
Why is supervision the bottleneck?
Because supervision capacity grows slowly since it takes years to produce a supervisor, and expanding student numbers without it produces poor completion rates.
Why does retention matter?
Because training people who then leave builds nothing — retention depends on position availability, funding continuity and working conditions.
Which support functions are underinvested?
Research administration, technical staff, and statistical support — consistently underinvested because their absence appears as researcher underperformance rather than a staffing gap.
What should be counted before expanding intake?
Supervision capacity — students without adequate supervision become non-completions, producing neither output nor capacity.
